Content uploaded by Paolo Giorgi Rossi
Author content
All content in this area was uploaded by Paolo Giorgi Rossi on Jan 09, 2016
Content may be subject to copyright.
Pubblicazione bimestrale Registrazione
del Tribunale di Milano
n. 239/1977 Spedizione in AP - 45% - art. 2 comma 20b legge
662/96 - Milano.
Iscrizione al Registro degli Operatori
di Comunicazione (ROC) n. 11747.
Una copia: 13,50 euro.
Abbonamento annuo: informazioni e condizioni sul sito
www.epiprev.it
Gestione abbonamenti: ufficio abbonamenti
tel. 02 48702283, fax 02 48706089.
I dati necessari per l’invio della rivista sono trattati elettronica-
mente e utilizzati dall’editore Inferenze scarl per la spedizione
della presente pubblicazione e di altro materiale medico-scien-
tifico. Ai sensi dell’art.13 Legge 675/96 è possibile in qualsiasi
momento e gratuitamente consultare, modificare e cancellare i
dati, o semplicemente opporsi al loro utilizzo scrivendo a: Infe-
renze scarl, responsabile dati, via Ricciarelli 29, 20148 Milano.
IVA assolta dall’editore ai sensi dell’art. 74 lettera C del DPR
26/10/1972 n.633 e successive modificazioni e integrazioni non-
ché ai sensi del DM 29/12/1989. Non si rilasciano quindi fattu-
re (art. 1 c. 5 DM 29/12/1989).
Stampa
Arti grafiche Ancora srl - Milano
Anno 39 (3) 2015 - Suppl. 1
Rivista fondata da Giulio A. Maccacaro
Epidemiologia & Prevenzione
è indicizzata in Medline,
Science Citation Index Expanded,
Journal Citation Reports/Science Edition
EPIDEMIOLOGIA
& PREVENZIONE
Testata associata
A.N.E.S.
ASSOCIAZIONE NAZIONAL E
EDITORIA PERIODICA SPECIALIZZATA
via Ricciarelli 29, 20148 Milano
segreteria@inferenze.it
Direttore scientifico
Eugenio Paci
Vicedirettore scientifico
Francesco Forastiere
Past director
Benedetto Terracini
Direttrice responsabile
Maria Luisa Clementi
Segreteria scientifica
Liliana Cori
Redazione
Marco Crespi, Cinzia Tromba, Maria Cristina Porro
Segreteria di redazione
via Giusti 4, 21053 Castellanza (VA)
e-mail: epiprev@inferenze.it
Impaginazione
Giovanna Smiriglia
Comitato di direzione
Fabio Barbone, Annibale Biggeri, Dolores Catelan, Dario Consonni, Emanue-
le Crocetti, Marina Davoli, Paolo Giorgi Rossi, Chiara Marinacci, Andrea Ran-
zi, Lorenzo Richiardi, Antonia Stazi, Giuseppe Traversa.
Comitato editoriale
Alessandro Barchielli, Antonella Bena, Carla Bietta, Achille Cernigliaro,
Fabrizio Faggiano, Emilio Gianicolo, Ursula Kirchmayer, Paola Michelozzi,
Alessio Petrelli (AIE); Franco Berrino, Annibale Biggeri, Pietro Comba, Gemma
Gatta, Luigi Mara, Alberto Martinelli, Enzo Merler, Franco Merletti, Salvatore
Panico, Silvano Piffer (Società per l’epidemiologia e la prevenzione Giulio A.
Maccacaro Impresa sociale);
Fabio Barbone, Pier Alberto Bertazzi, Fabrizio Bianchi, Piero Borgia, Silvia
Candela, Franco Carnevale, Ugo Fedeli, Stefano Ferretti, Alba Finarelli, Livia
Giordano, Roberto Grilli, David Kriebel, Andrea Micheli, Roberta Pirastu, Renato
Pizzuti, Walter Ricciardi, Roberto Romizi, Stefania Salmaso, Rodolfo Saracci,
Salvatore Scondotto, Paolo Vineis, Marco Zappa (membri invitati dalla Direzione
Scientifica, non in rappresentanza della proprietà).
Modalità di abbonamento
Pagamento con carta di credito (American Express, Carta Sì, VISA, Eurocard,
Master Card) telefonando allo 02-48702283 dal lunedì al venerdì dalle 9 alle
13 oppure utilizzando il servizio PayPal sul sito web della rivista www.epiprev.it
Versamento su conto corrente postale n. 55195440 intestato a Inferenze scarl,
via Ricciarelli n. 29, 20148 Milano (segnalare la causale del versamento).
Accredito tramite c/c bancario presso: UNIPOL BANCA P.za Wagner n. 8,
20145 Milano, IBAN: IT53P 03127 01600 000000003681 intestato all’im-
presa editoriale Inferenze scarl, via Ricciarelli n. 29, 20148 Milano.
Si ringrazia la Fondazione IRCCS Istituto nazionale dei tumori di Milano
che ospita l’Impresa sociale.
© Inferenze scarl, Milano
EPIDEMIOLOGIA
& PREVENZIONE
WWW.EPIPREV.IT
Editors / A cura di: Marco Zappa, National centre for screening monitoring (ONS)
Francesca Carozzi, Italian group for cervical screening (GISCi)
Livia Giordano, Italian group for mammographic screening (GISMa)
Romano Sassatelli, Italian group for colorectal screening (GISCoR)
Antonio Federici, Ministry of Health,Prevention Department
THE NATIONAL CENTRE
FOR SCREENING MONITORING
Eleventh Report
OSSERVATORIO NAZIONALE
SCREENING
Undicesimo Rapporto
GISCi
Gruppo Italiano
Screening
Cervicocarcinoma
GISCoR
Gruppo Italiano
Screening
Colorettale
GISMa
Gruppo Italiano
Screening
Mammografico
Contents/Indice
WWW.EPIPREV.IT
2
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
The diffusion of screening programmes in Italy, years 2011-2012 . . . . . . . . . . . 5
La diffusione dei programmi di screening in Italia, anni 2011-2012
Marco Zappa, Francesca Maria Carozzi, Livia Giordano, Romano Sassatelli, Antonio Federici
Cancer screening uptake: association with individual characteristics, . . . . . . . . 9
geographic distribution, and time trends in Italy
La copertura dei test di screening: caratteristiche, distribuzione geografica
e trend temporali
Giuliano Carrozzi, Letizia Sampaolo, Lara Bolognesi, Laura Sardonini, Nicoletta Bertozzi,
Paolo Giorgi Rossi, Marco Zappa, Sandro Baldissera, Stefano Campostrini, Gianluigi Ferrante,
Maria Masocco, Valentina Minardi, Angelo D’Argenzio, Pirous Fateh Moghadam, Elisa Quarchioni,
Mauro Ramini, Massimo Oddone Trinito, Stefania Salmaso for the regional and local PASSI coordinators
BREAST CANCER SCREENING
Glossary/Glossario ................................................ 20
Mammographic breast cancer screening in Italy: 2011-2012 survey . . . . . . . . . . 21
Lo screening mammografico in Italia: survey 2011-2012
Leonardo Ventura, Daniela Giorgi, Livia Giordano, Alfonso Frigerio, Paola Mantellini, Marco Zappa
and the Italian breast cancer screening survey group
Breast cancer screening in Italy: evaluating key performance indicators . . . . . . 30
for time trends and activity volumes
Lo screening mammografico in Italia: valutazione degli indicatori
di performance per trend temporali e volumi di attività
Livia Giordano, Roberta Castagno, Daniela Giorgi, Cistiano Piccinelli, LeonardoVentura,
Nereo Segnan, Marco Zappa
Audit system on Quality of breast cancer diagnosis and Treatment (QT): . . . . . . 40
results of quality indicators on screen-detected lesions in Italy, 2011-2012
Il “progetto SQTM” sulla qualità della diagnosi e della terapia entro
i programmi di screening in Italia: risultati 2011-2012
Antonio Ponti, Maria Piera Mano, Mariano Tomatis, Diego Baiocchi, Alessandra Barca, Rosa Berti,
Denise Casella, Enrico D’Ambrosio, Erika Delos, Giovanni Donati, Fabio Falcini, Brunella Frammartino,
Alfonso Frigerio, Fabiola Giudici, Paola Mantellini, Carlo Naldoni, Carlo Olla Atzeni,
Lorenzo Orzalesi, Giovanni Pagano, Francesca Pietribiasi, Sabina Pitarella, Alessandra Ravaioli,
Anna Silvestri, Mario Taffurelli, Enrica Tidone, Fabrizio Zanconati, Nereo Segnan
Information provided by Italian breast cancer screening programmes: . . . . . 48
a comparison between 2001 and 2014
Informazioni fornite dai programmi di screening mammografico in Italia:
un confronto tra il 2001 e il 2014
Roberta Castagno, Debora Canuti , Marco Petrella, Lauro Bucchi, Chiara Fedato,
Francesca Garena, Livia Giordano
Problems, solutions, and perspectives in the evaluation of interval cancers . . . 52
in Italian mammography screening programmes: a position paper
from the Italian group for mammography screening (GISMa)
Problemi, soluzioni e prospettive nella valutazione dei cancri d’intervallo
nei programmi italiani di screening mammografico: position paper
del Gruppo italiano screening mammografico (GISMa)
Lauro Bucchi, Alfonso Frigerio, Manuel Zorzi, Chiara Fedato, Giovanni Angiolucci, Daniela Bernardi,
Cinzia Campari, Emanuele Crocetti, Stefano Ferretti, Daniela Giorgi, Francesca Marchisio,
Doralba Morrone, Carlo Naldoni, Marco Petrella, Antonio Ponti, Alessandra Ravaioli, Gianni Saguatti,
Dolores Santini, Priscilla Sassoli de Bianchi, Monica Serafini, Viviana Vergini, Livia Giordano
anno 39 (3) maggio-giugno 2015
e
&
p
3
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
CERVICAL CANCER SCREENING
Glossary/Glossario ................................................ 60
Extension of organized cervical cancer screening programmes in Italy . . . . . . . 61
and their process indicators, 2011-2012 activity
Estensione dei programmi organizzati di screening del cancro cervicale
in Italia e loro indicatori di processo
Guglielmo Ronco, Pamela Giubilato, Francesca Carozzi, Giovanni Maina, Paolo-Giorgi-Rossi,
Marco Zappa and the Cancer screening survey working group
A first survey of HPV-based screening in routine cervical cancer screening . . . . 77
in Italy
Prima survey sull’utilizzo routinario del test HPV nello screening cervicale
in Italia
Guglielmo Ronco, Paolo Giorgi-Rossi, Pamela Giubilato, Annarosa Del Mistro, Marco Zappa,
Francesca Carozzi and the HPV screening survey working group
hr-HPV testing in the management of women with ASC-US+ . . . . . . . . . . . . . . . . 84
and in the follow-up of women with cytological abnormalities and negative
colposcopy. Recommendations of the Italian group for cervical cancer screening
(GISCi) Test hr-HPV nella gestione delle donne con citologia ASC-US+
e nel follow-up delle donne con citologia anormale e colposcopia negativa:
raccomandazioni del Gruppo italiano per lo screening del carcinoma della cervice
uterina (GISCi)
Francesca Maria Carozzi, Anna Iossa, Aurora Scalisi, Mario Sideri,†Karin Louise Andersson,
Massimo Confortini, Annarosa Del Mistro, Giovanni Maina, Guglielmo Ronco, Patrizio Raggi,
Maria Luisa Schiboni, Marco Zappa, Paolo Giorgi Rossi
COLORECTAL CANCER SCREENING
Glossary/Glossario ................................................ 92
Screening for colorectal cancer in Italy: 2011-2012 survey . . . . . . . . . . . . . . . . . . 93
Screening dei tumori del colon retto in Italia: survey 2011-2012
Manuel Zorzi, Filippo Da Re, Paola Mantellini, Carlo Naldoni, Priscilla Sassoli de’ Bianchi, Carlo Senore,
Anna Turrin, Carmen BeatrizVisioli, Marco Zappa and the Italian colorectal cancer screening survey group
Characteristics of the colorectal cancers diagnosed in the early 2000s in Italy . . . 108
Figures from the IMPATTO study on colorectal cancer screening
Caratteristiche dei tumori del colon retto diagnosticati in Italia nei primi anni Duemila.
Dati dello studio IMPATTO dello screening colorettale
Manuel Zorzi, Lucia Mangone, Emanuela Anghinoni, Susanna Baracco, Elisabetta Borciani,
Adele Caldarella, Fabio Falcini, Anna Clara Fanetti, Stefano Ferretti, Paolo Giorgi Rossi,
Maria Michiara, Giorgia Randi, Fabrizio Stracci, Massimo Vicentini, Antonella Zucchetto,
Marco Zappa and IMPATTO COLONRETTO working group
Incidence trends of colorectal cancer in the early 2000s in Italy. . . . . . . . . . . . . . . 115
Figures from the IMPATTO study on colorectal cancer screening
Trend di incidenza tumori del colon retto nei primi anni Duemila in Italia.
Dati dello studio IMPATTO dello screening colorettale
Manuel Zorzi, Lucia Mangone, Emanuela Anghinoni, Susanna Baracco, Elisabetta Borciani,
Adele Caldarella, Fabio Falcini, Anna Clara Fanetti, Stefano Ferretti, Paolo Giorgi Rossi,
Maria Michiara, Giorgia Randi, Fabrizio Stracci, Massimo Vicentini, Antonella Zucchetto,
Marco Zappa and IMPATTO COLONRETTO working group
WWW.EPIPREV.IT
anno 39 (3) maggio-giugno 2015
e
&
p
4
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Region Males Females Total
Abruzzo 633,941 672,475 1,306,416
Alto Adige (PA Bolzano) 248,407 256,301 504,708
Basilicata 282,546 295,016 577,562
Calabria 953,767 1,004,651 1,958,418
Campania 2,794,720 2,969,704 5,764,424
Emilia-Romagna 2,094,766 2,246,474 4,341,240
Friuli-Venezia Giulia 587,449 630,331 1,217,780
Lazio 2,635,689 2,864,333 5,500,022
Liguria 740,458 826,881 1,567,339
Lombardia 4,711,292 4,989,589 9,700,881
Marche 745,469 795,219 1,540,688
Molise 152,547 160,598 313,145
Piemonte 2,101,852 2,255,811 4,357,663
Puglia 1,962,375 2,087,697 4,050,072
Sardegna 800,451 837,395 1,637,846
Sicilia 2,417,426 2,582,428 4,999,854
Toscana 1,759,289 1,908,491 3,667,780
Trentino (PA Trento) 255,832 269,045 524,877
Umbria 423,559 459,656 883,215
Valle d’Aosta 61,775 64,845 126,620
Veneto 2,362,989 2,490,668 4,853,657
Total 28,726,599 30,667,608 59,394,207
Alto Adige
Valle d’Aosta Trentino
VenetoLombardia
Piemonte
Liguria
Toscana Marche
Umbria
Abruzzo
Molise
PugliaPuglia
Basilicata
Campania
Sardegna
Sicilia
Calabria
Lazio
Emilia-Romagna
Friuli-Venezia Giulia
Table. Italian population by
sex and region, year 2012
(www.demo.istat.it).
Tabella. Popolazione ita-
liana nell’anno 2012, suddi-
visa per sesso e per Regione
(www.demo.istat.it).
anno 39 (3) maggio-giugno 2015
e
&
p
Introduction
The diffusion of screening programmes
in Italy, years 2011-2012
La diffusione dei programmi di screening
in Italia, anni 2011-2012
Marco Zappa,1Francesca Maria Carozzi,2Livia Giordano,3Romano Sassatelli,4Antonio Federici5
In this report, we present the results of cancer screening programmes in Italy for the years 2011-
2012. This report is produced by the National centre for screening monitoring (ONS), together with
the Italian professional multidisciplinary screening groups: GISMa (Italian group for mammographic
screening), GISCor (Italian group for colorectal screening), and GISCi (Italian group for cervical
screening). Since 2004, ONS has been monitoring and supporting Italian screening programmes,
in accordance with a decree issued by the Ministry of Health. Multidisciplinary groups work with
ONS and provide the know-how required to promote the quality of public health programmes.
The following is a brief outline of the Italian screening programme setting:
■
screening programmes (cervical, mammographic, colorectal) have been a Basic Healthcare Param-
eter (livello essenziale di assistenza, LEA) since 2001;
■
guidelines are provided by the Ministry of Health’s Department of Prevention in agreement with
regional governments;
■
regional governments are responsible for the organization, management, and quality assurance of
screening programmes;
■
since 2004, ONS has been responsible for monitoring and promoting screening programmes na-
tionwide;
■
the results of the screening programmes of each region are evaluated annually by the Ministry of
Health in terms of coverage and impact.
The
main characteristics of protocols of of mammographic, cervical and colorectal screening pro-
grammes are summarized in
table 1
(p. 7).
Overall, in 2011-2012 almost 20 million people were invited to undergo a screening examination
(7,419,295; 5,271,248 and 7,744,295 for cervical, breast, and colorectal cancer, respectively). As com-
pared to the previous years, an increase was observed for all the screening programmes. Almost 10 mil-
lion actually complied to the invitation (3,051,852; 2,959,329 and 3,556,486 for cervical, breast, and
colorectal cancer, respectively). Unfortunately, in the observed increase in invitation and participation
inequality persisted and grew between Centre, North, and South of Italy.
The screening activity has already produced a remarkable impact on the epidemiology of these three
cancers in Italy. Changes have been documented in several papers.1-5
CERVICAL CANCER SCREENING
Taking a closer look at the data (and adopting the same criteria for each year), we can observe that
the actual extension of cervical cancer screening (i.e., how many 25-64 year-old women regularly re-
ceived an invitation letter to perform a Pap smear every three years) in 2011-2012 was close to 70%
(69.5%).This does not mean that 30% of the target population did not receive an invitation to screen-
ing. In some cases, it is possible that invitations were issued but the interval was longer than 3 years.
1National centre
for screening monitoring
(ONS) –Istituto
per lo studio
e la prevenzione
oncologica (ISPO), Firenze
2Italian group for cervical
screening (GISCi) –Istituto
per lo studio
e la prevenzione
oncologica (ISPO), Firenze
3Italian group
for mammographic
screening (GISMa) –
Centro per la prevenzione
oncologica (CPO),
Piemonte, Torino
4Italian group for colorectal
screening (GISCor) –
Azienda ospedaliera
di Reggio Emilia,
Emilia-Romagna
5Ministry of Health,
Prevention Department,
Dipartimento prevenzione,
Ministero della salute,
Roma
Corresponding author
Marco Zappa
m.zappa@ispo.toscana.it
WWW.EPIPREV.IT
anno 39 (3) maggio-giugno 2015
e
&
p
5
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Introduction
6NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
2004-2006* 2007-2009 2010-2012
%
60
70
80
90
50
40
30
20
10
0
Figure 1. Actual extension of
cervical screening programmes
by geographical area. Years
2004-2012.
Figura 1. Estensione effettiva
dei programmi di screening
cervicale per area geografica.
Anni 2004-2012.
51,8
40.6
69.2
52.9
62.6
79.5
53.8
63.0
69.7 69.5
83.5
61.2
North
Centre
South/Islands
Italy
2005-2006 2007-2008 2009-2010 2011-2012
%
60
70
80
90
100
50
40
30
20
10
0
Figure 2. Actual extension of
mammographic screening pro-
grammes by geographical area.
Years 2005-2012.
Figura 2. Estensione effettiva
dei programmi di screening
mammografici per area geo-
grafica. Anni 2005-2012.
72.5
66.8
20.8
54.3
86.0
75.2
31.6
65.5
89.0
79.1
36.9
69.2
93.9
86.1
39.6
North
Centre
South/Islands
Italy
Extension in 2010-2012 was greater than in 2004-2006
(51.8%) and 2007-2009 (63%) (figure 1). This increase con-
cerns all three Italian macro-areas (North, Centre, South),
with a low heterogeneity among them, unlike what was ob-
served in the other two types of screening. Unfortunately, this
is partly due to the fact that the largest Italian regions in
northern Italy did not implement a cervical screening pro-
gramme throughout the entire region.
A crucial innovation for cervical screening policy is currently
taking place. Italy is one of the first countries in Europe to
move towards the use of DNA HPV test as a primary test. As
reported by Ronco et al. in this issue,6in 2012, 19 Italian pro-
grammes from 10 regions invited women for HPV-based
screening. During 2012, more than 300,000 (8% of the tar-
get population) women were invited to HPV testing and more
than 130,000 accepted. As far as we know, this is one of the
first reports in Europe on the performances of HPV-based
screening programmes.
BREAST CANCER SCREENING
Regarding mammography screening, actual extension from
2005 to 2012 (percentage of 50-69 years old women regularly
receiving a letter of invitation every two years) is reported in
figure 2. In the biennium 2011-2012, almost 3 out of 4
women were invited (73.2%). Unfortunately, screening diffu-
sion is still heterogeneous, with a higher distribution in north-
ern/central Italy (nearing or over 90%), compared with south-
ern/insular Italy (only 40%). Even though we observed a stable
increase from 2005-2006 in all three areas (on average, each
area showed twenty percentage points less in 2005-2006), this
trend does not allow us to be fully optimistic. Due to the dif-
ficulties in spreading organized screening activity in southern
Italy, the goal of assuring complete breast screening coverage
in Italy remains uncertain.
It is worth mentioning that in 2011-2012, 227,00 women
older than 69 (13.6% of the target population) were invited to
continue screening till 74 years of age. Furthermore, two re-
73.3
gions (Emilia-Romagna and Piemonte) also included younger
women (ages 45-49) among those to be invited. In 2011-
2012, almost 380,000 women in this age class were invited an-
nually (7.9% of the Italian target population of 45-49 year-old
women). The latter figure shows a small increase in compari-
son with the previous two years.
COLORECTAL CANCER SCREENING
Concerning colorectal cancer screening, in the period 2011-
2012 we continued to observe an increase in the actual exten-
sion for the whole country (extension was 53% of the target
population: men and women aged 50-69). Actual extension
was almost double compared to the biennium 2005-2006
(29.7%). This is very encouraging, since colorectal cancer
screening was only introduced recently (2005) in Italy. Un-
fortunately, once again, differences between North and South
are evident and become increasingly greater, with 82%, 59%,
and 12% actual extension in the North, Centre, and South, re-
spectively. Even more worrisome is the fact that in the South
we did not observe any relevant increase till 2012.
DISCUSSION
In conclusion, we observed an increase in the actual extension
of all three screening programmes, although the differences be-
tween Centre, North, and South remained relevant, especially
for breast and colorectal cancer screening.
Our data are consistent with the PASSI survey reported on in
this issue by Carrozzi et al.7PASSI is a national telephone sur-
Introduction
7NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
2005-2006 2007-2008 2009-2010 2011-2012
%
60
70
80
90
50
40
30
20
10
0
Figure 3. Actual extension of
colorectal screening pro-
grammes by geographical area.
Years 2005-2012.
Figura 3. Estensione effettiva
dei programmi di screening
colorettali per area geogra-
fica. Anni 2005-2012.
32.9
19.4
4.3
20.7
61.8
29.8
3.5
35.9
72.4
40.0
7.1
43.7
82.5
58.9
12.2
North
Centre
South/Islands
Italy
53.1
Mammographic screening
Target population women aged 50-69 (some regions have extended the age target from 45 to 74)
Primary test 2 views, doubling reading mammographic test
Screening interval 2 years
Cervical screening
Target population women aged 25-64
Primary test Pap smear
Screening interval 3 years
Some programs have moved towards HPV testing as primary test:
Target population HPV: women aged 30/35-64
Pap smear: women aged 25-30/35
Primary test HPV
Screening interval 5 years
Colorectal screening
Primary test fecal immunochemical test (FIT)
Target population subjects aged 50-69 (some regions have extended the age target to 74 or 75 years)
Screening interval 2 years
Primary test flexible sigmoidoscopy (FS) + FIT
Target population subjects aged 58 or 60 (FS); subjects aged 59-69 (FIT)
Screening interval flexible sigmoidoscopy once in a lifetime and FIT every 2 years for non-responders to FS
Table 1. Main characteristics of pro-
tocols of mammographic, cervical and
colorectal screening programmes.
Tabella 1. Caratteristiche principali
dei programmi di screening mammo-
grafico, cervicale e colorettale.
veillance system that continuously collects information about
behavioural health risk factors and the diffusion of preventive
health interventions. PASSI collects information both on or-
ganized screening programmes and spontaneous public and
private screening. The PASSI survey reports that from 2010 to
2013 coverage increased for all types of screening and the in-
crease was mostly due to the tests performed within organized
programmes. All three screening types show a decreasing
North-South trend in coverage. The gap between Centre-
North and South is mainly due to organized screening.
A screening programme is not limited to the administration of
a test. It is the construction of a process which takes care of the
invited person from the primary test to (if necessary) the as-
sessment phase, treatment, and follow-up of the detected le-
sions. Each of these phases requires a standardized protocol and
a monitoring system in order to maintain high quality assur-
ance. In the present issue, we present examples of the effort we
are making in that direction.
Ponti et al.8reports on the audit system on Quality of breast
cancer diagnosis and treatment (QT). QT is a voluntary qual-
ity assurance programme concerning screen-detected breast
cancer care and it has been running in Italy since 1997. Dur-
ing the period 2000-2012, about 40,000 lesions in thirteen
Italian regions were documented in QT.
Castagno et al.9deal with the quality and completeness of the
information provided to women by Italian breast screening
programmes. It reports the results of a survey promoted by the
Italian group for mammography screening (GISMa) in the
spring of 2014. Aim of the study was to compare information
provided by invitation letters and leaflets of Italian breast
screening programmes in 2001 and nowadays, and to verify
whether there has been an evolution in the type of information
provided, and, if so, of what type.
Bucchi et al.10 report the position paper on interval cancers by
the Italian group for mammography screening. In particular,
the paper outlines problems and solutions with respect to ap-
propriate assessment of the frequency of interval cancers in re-
lation to expected incidence (proportional incidence).
Carozzi et al.11 describe the HPV-based follow-up protocol
for cervical lesions proposed by the Italian group for cervical
screening (GISCi). Aim of the protocol is to improve follow-
up appropriateness (eliminating too frequent check-ups) by
using HPV testing. To date, screening programmes in Italy
lack any clearly defined follow-up protocol after an abnormal
Pap smear and negative colposcopy, or any uniform indica-
tions.
In the two papers by Zorzi et al.4,5 the early impact of imple-
mentation of screening programmes on stage distribution at di-
agnosis and incidence of colorectal cancer is reported. Despite
the brief time since programme implementation, clear changes
have nevertheless been evident in the epidemiology of col-
orectal screening.
Introduction
8NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Zucchetto A, Ronco G, Giorgi Rossi P et al. Screening patterns
within organized programs and survival of Italian women with in-
vasive cervical cancer. Prev Med 2013;57(3):220-26.
2. IMPACT Working group. Epidemiological changes in breast tu-
mours in Italy: the IMPACT study on mammographic screening pro-
grammes. Pathologica 2011;103(5):290-93.
3. Foca F, Mancini S, Bucchi L et al. Decreasing incidence of late-stage
breast cancer after the introduction of organized mammography
screening in Italy. Cancer 2013;119(11):2022-28.
4. Zorzi M, Mangone L, Anghinoni E et al. Characteristics of the col-
orectal cancers diagnosed in the early 2000s in Italy. Figures from
the IMPACT study on colorectal cancer screening. Epidemiol Prev
2015;3(Suppl 1):108-14.
5. Zorzi M, Mangone L, Sassatelli R et al. Incidence trends of col-
orectal cancer in the early 2000s in Italy. Figures from the IMPACT
study on colorectal cancer screening. Epidemiol Prev 2015;3(Suppl
1):115-25.
6. Ronco G, Giorgi Rossi P, Giubilato P et al. A first survey of HPV-
based cervical cancer screening in routine activity in Italy. Epi-
demiol Prev 2015;3(Suppl 1):77-83.
7. Carrozzi G, Sampaolo S, Bolognesi L et al. Cancer screening uptake:
association with individual characteristics, geographic distribution,
and time trends in Italy. Epidemiol Prev 2015;3(Suppl 1):9-18.
8. Ponti A, Mano MP, Tomatis M et al. Audit system on Quality of
breast cancer diagnosis and Treatment (QT): results of quality in-
dicators on screen-detected lesions in Italy, 2011-2012. Epidemiol
Prev 2015;3(Suppl 1):40-47.
9. Castagno R, Canuti D, Petrella M. Information provided by Italian
breast cancer screening programmes: a comparison between 2001
and 2014. Epidemiol Prev 2015;3(Suppl 1):48-51.
10. Bucchi L, Frigerio A, Zorzi M et al. Problems, solutions, and per-
spectives in the evaluation of interval cancers in Italian mammog-
raphy screening programmes: a position paper from the Italian
Group for Mammography Screening (GISMa). Epidemiol Prev
2015;3(Suppl 1):52-57.
11. Carozzi FM, Iossa A, Scalisi A et al. hr-HPV testing in the man-
agement of women with ASC-US+ and in the follow-up of women
with cytological abnormalities and negative colposcopy: Recom-
mendations of the Italian Group for Cervical Cancer Screening
(GISCi). Epidemiol Prev 2015;3(Suppl 1):84-90.
9
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Cancer screening uptake: association
with individual characteristics, geographic
distribution, and time trends in Italy
La copertura dei test di screening: caratteristiche,
distribuzione geografica e trend temporali
Giuliano Carrozzi,1Letizia Sampaolo,2Lara Bolognesi,1Laura Sardonini,3Nicoletta Bertozzi,3
Paolo Giorgi Rossi,4,5 Marco Zappa,6Sandro Baldissera,7Stefano Campostrini,8Gianluigi Ferrante,7
Maria Masocco,7Valentina Minardi,7Angelo D’Argenzio,9Pirous Fateh Moghadam,10 Elisa Quarchioni,7
Mauro Ramigni,11 Massimo Oddone Trinito,12 Stefania Salmaso7for the regional and local PASSI
coordinators
Abstract
Background. In Italy, organized screening programmes invite the vast majority of the population for
cervical and breast cancer, and about one half of the population for colorectal cancer. Programme ac-
tivity and quality are closely monitored. Nevertheless, there is a vast spontaneous activity, both pub-
lic and private, for which information on service and coverage is missing. To estimate actual population
coverage for the three types of screening the extent of spontaneous screening needs to be known.
Methods. PASSI is a national telephone-interview surveillance system that continuously collects infor-
mation about behavioural health risk factors and the diffusion of preventive health interventions. From
2010 to 2013, more than 151,000 18- to 69-year-olds were interviewed. During 2013, 136 out of 147
Italian local health authorities participated in the survey. Information about screening includes: test up-
take (Pap smear, HPV, mammography, faecal occult blood test, colonoscopy), date of the last test,
provider of the last test (whether paid or for free, proxy of the organized screening programme), rea-
son for not participating in screening, and screening promotion/recommendation received. Individual
information on socio-economic characteristics is available.
Results. Seventy-seven percent of the 25-64 year-old women interviewed said they had undergone a
Pap smear or HPV test in the three years before the interview, 40% within the screening programme,
37% spontaneously and paying. Seventy percent of the 50-69 year-old women interviewed reported
having had a mammography in the two years before the interview, 51% within the screening pro-
gramme, 19% spontaneously and paying. Thirty-eight percent of the 50-69 year olds interviewed re-
ported having undergone colorectal screening in the two years before the interview, 31% within the
screening programme, 7% spontaneously and paying.
All three screening programmes showed a decreasing North-South trend in coverage. From 2010 to
2013, coverage increased for all types of screening; the trend was stronger in the South; the increase
was mostly due to the tests performed within the organized programmes. People with low education,
economic problems, and immigrants from high migration pressure countries had lower coverage lev-
els. In regions with well-implemented organized screening programmes, test coverage was higher
and differences for socio-economic factors were smaller than in regions with incomplete programme
activation.
Epidemiol Prev 2015; 39(3) Suppl 1: 9-18)
Keywords: breast cancer, cervical cancer, colorectal cancer, mass screening, opportunistic/spontaneous screening, Italy
1Dipartimento di sanità
pubblica, AUSL Modena
2Dipartimento di sanità
pubblica, AUSL Modena
e Università Ca’ Foscari,
Venezia
3Dipartimento di sanità
pubblica, AUSL
della Romagna
4Servizio interaziendale
di epidemiologia, AUSL
Reggio Emilia
5IRCCS-Arcispedale
S. Maria Nuova,
Reggio Emilia
6Istituto per lo studio
e la prevenzione
oncologica, Osservatorio
nazionale screening
7Centro nazionale
di epidemiologia,
sorveglianza
e promozione della salute,
Istituto superiore di sanità
8Dipartimento
di economia, Università
Ca’ Foscari, Venezia
9Servizio epidemiologia
e prevenzione, ASL
Caserta
10Dipartimento salute
e solidarietà sociale,
Provincia autonoma
di Trento
11Dipartimento
di prevenzione, Azienda
ULSS 9 Treviso
12Dipartimento
di prevenzione,
ASL Roma C
Corresponding author
Giuliano Carrozzi
g.carrozzi@ausl.mo.it
WWW.EPIPREV.IT
anno 39 (3) maggio-giugno 2015
e
&
p
Cancer screening uptake – PASSI data
10
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUZIONE
In Italy, in accordance with the European Commission’s 2003
Recommendation,1the Italian Ministry of Health guidelines rec-
ommend the implementation of organized screening pro-
grammes for cervical, colorectal, and breast cancer.2These pro-
grammes involve active invitation of the entire target population,
free testing and treatment, quality assurance in all stages of the
process, and process and early outcome monitoring system.
Activation of screening programmes is not complete and uni-
form throughout Italy.3-5 Furthermore, cervical and breast
cancer screening programmes started when Pap smears and
mammography were already in widespread use in the popula-
tion. For these reasons, in Italy, there is a strong opportunis-
tic/spontaneous uptake of both mammography and, in par-
ticular, Pap smears, both in the public and private sector. The
spontaneous activity is not precisely measurable, it is not mon-
itored, and its target population is not defined. Any attempt
to measure the spontaneous activity through routine or ad-
ministrative data failed due to strong under-reporting of pre-
ventive tests in these databases.6,7
In order to estimate the actual population coverage for the three
types of screening it is necessary to know the spontaneous up-
take of preventive tests. To date, the most reliable source of in-
formation for spontaneous screening are population interviews.8
Until 2007, the only national survey estimating mammography
and Pap smear coverage was the National Health interview,
which is repeated every five years.9Starting from 2007, the
PASSI surveillance has monitored cervical, colorectal, and breast
cancer screening coverage with a continuous survey.10
Aim of this paper is to present the coverage estimates for the
three types of screening, their geographical differences, their
association with individual socio-economic factors, and their
time trends.
METHODS
PASSI is a National surveillance system that continuously
collects information via phone calls about behavioural health
risk factors and the diffusion of preventive health care services.
From 2010 to 2013, more than 151,000 18-69 year-old peo-
ple were interviewed. During 2013, 136 out of 147 Italian lo-
cal health units participated in the survey.
The sampling and survey methodologies are described in de-
tail elsewhere.11 Briefly, the surveillance system is based on
a random sample of people resident in the area and registered
in the list of each Local Health Authority. Samples are strat-
ified by gender and age to respect the proportion of the
population (18-34, 35-49, 50-69). Eligibility criteria are: age
18-69, residence, ability to understand and answer the ques-
tions in Italian, and not being in a residential institution (hos-
pital, nursing home, military barracks, prison).12 The Local
Riassunto
Introduzione. In Italia sono attivi programmi di screening organizzati per il carcinoma della cervice uterina, della mammella e del
colon-retto, la cui attività è dettagliatamente monitorata. Ciononostante esiste una intensa attività di screening spontanea, sia nel
privato sia nel pubblico, di cui non si conosce il dettaglio delle prestazioni e della popolazione target. Per stimare la reale copertura
della popolazione per i tre screening è dunque necessario conoscere il ricorso da parte della popolazione allo screening spontaneo.
Metodi. PASSI è un sistema di sorveglianza nazionale che raccoglie in continuo, tramite interviste telefoniche, informazioni sui fat-
tori comportamentali di rischio per la salute e sulla diffusione degli interventi di prevenzione messi in campo dalle aziende sanita-
rie nei confronti delle persone tra i 18 e i 69 anni. Dal 2010 al 2013 sono state intervistate oltre 151.000 persone. Nel 2013 hanno
partecipato al sistema 136 su 147 ASL italiane. Tra i vari temi indagati ci sono: l’effettuazione dei test di screening (Pap-test e test
HPV, mammografia, sangue occulto e colonscopia), la data dell’ultimo test, il setting in cui è stato fatto (a pagamento o meno, proxy
del programma di screening organizzato), i motivi di non adesione al programma di screening e gli interventi di promozione (let-
tera ASL, consiglio sanitario, campagna informativa). Sono raccolte, inoltre, informazioni sociodemografiche individuali.
Risultati. Il 77% delle donne di 25-64 anni intervistate ha eseguito un test di screening cervicale (Pap-test o test Hpv) nei tre anni
precedenti l’intervista, il 40% all’interno di programmi organizzati dalle ASL e il 37% su iniziativa personale. Il 70% delle donne
intervistate di 50-69 anni ha eseguito una mammografia a scopo preventivo nel corso dei due anni precedenti l’intervista, il 51%
all’interno dei programmi organizzati e il 19% su iniziativa personale. Il 38% delle persone intervistate di 50-69 anni ha ese-
guito esami per la diagnosi precoce dei tumori colon-rettali, il 31% all’interno dei programmi di screening, il 7% su iniziativa
personale.
La copertura di tutti i tre test mostra un gradiente Nord-Sud. Nel periodo 2008-2013 le coperture risultano complessivamente
in crescita, andamento più evidente nelle regioni meridionali; aumentano soprattutto gli esami eseguiti all’interno dei programmi
organizzati. La copertura mostra differenziali per livello di istruzione e difficoltà economiche; è inoltre più alta tra le persone con
cittadinanza italiana o provenienti da altri Paesi a sviluppo avanzato (PSA) rispetto agli stranieri provenienti da Paesi a forte pres-
sione migratoria (PFPM).
Nelle Regioni con programmi di screening organizzati con buona estensione e adeguatamente funzionanti l’esecuzione dei test di
screening è significativamente più alta e le disuguaglianze socioeconomiche nella copertura sono minori.
(Epidemiol Prev 2015; 39(3) Suppl 1: 9-18)
Parole chiave: cancro del seno, cancro della cervice uterina, cancro del colon-retto, programmi di screening, screening opportunistico/spontaneo, Italia
Cancer screening uptake – PASSI data
11
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Health Authority (LHA) alerts all sampled people with a let-
ter informing them about the interviews, the privacy condi-
tions, and the way to opt out and deny consent to being con-
tacted by phone. The LHA also contacts GPs, asking them to
help contact sampled people and explain the scope and aim
of the interview. Interviews are conducted mostly by health
personnel specifically trained with classroom and online
courses. The interview takes about twenty minutes and is ei-
ther a CATI (Computer AssistedTelephone Interview) or reg-
istered on paper with back office data entry. Interviews are
stored anonymously in a national database. The questionnaire
has closed questions on perceived health status, symptoms, de-
pression, prevalence of chronic diseases and conditions, preva-
lence of behavioural risk factors, received preventive and
health promotion screening interventions, vaccines, and safety
on the road and at home.13,14
Information about screening includes: test uptake (Pap smear,
HPV, mammography, faecal occult blood test, colonoscopy),
date of the last test, provider of the last test (paying or for free,
proxy of the organized screening programme), reason for not
participating in screening, and screening promotion/recom-
mendation received. Individual information on socio-eco-
nomic characteristics is available.
Analysis
Coverage was defined as the proportion of people in the tar-
get population who had a test within the recommended time:
women aged 25-64 who had a Pap smear or HPV test within
three years before the interview for cervical cancer; women aged
50-69 who had a bilateral mammography within two years be-
fore the interview for breast cancer; women and men aged 50-
69 who had a faecal occult blood test within two years before
the interview or a colonoscopy/sigmoidoscopy within five
years before the interview for colorectal cancer. For the region
of Piemonte, where the screening programme adopts a strat-
egy of a once-in-a-lifetime sigmoidoscopy at the age of 58, the
target population was restricted to ages 58-69, and subjects
were considered covered if they had had a colonoscopy/sig-
moidoscopy in their life.
Regions were classified as having a well-implemented screen-
ing programme if more than 75% of the interviewed target
population declared they had received the invitation letter.
Associations between coverage and individual characteristics
were tested with logistic regression models; time trends were
evaluated through Cochrane-Orcutt linear regression models.
Time trends for colorectal cancer screening test coverage are
limited to the period 2010-2013 because the questions in the
questionnaire were changed at the end of 2009.
RESULTS
Cervical cancer screening test coverage
Overall, 77% of the 25-64 year-old women had a Pap smear
or HPV test in the three years before the interview. There was
a decreasing North-South trend (85% in the North, 84% in
the Centre, and 65% in the South and Islands).
Forty percent of the women performed the test within a screen-
ing programme for free and 37% performed the test sponta-
neously paying it entirely or in part.
In northern regions, the proportion of women who performed
the test within a screening programme was higher than in
southern regions, where spontaneous testing was predomi-
nant (figure 1), with the exception of the province of Bolzano
%
60
80
100
40
20
0
Figure 1. Cervical cancer screening test coverage.Proportion of 25-64 year-old women who had a Pap smear or HPV test in the three years before the interview,within
screening programmes or spontaneously,by region. Nationwide pooled data, PASSI 2010-2013.
Figura 1. Copertura di un test per la prevenzione dei tumori del collo dell’utero. Proporzione di donne di età 25-64 anni che hanno avuto un Pap test o un test HPV
negli ultimi tre anni, all’interno dei programmi di screening o spontaneamente. Pool, PASSI 2010-2013.
40
37
within the screening programme spontaneous screening
nationwide pooled data
North
45
39
Centre
30
35
South
73
15
Valle d’Aosta
60
26
Piemonte
14
68
Liguria
20
57
Lombardia*
55
27
Trento
33
57
Bolzano
50
37
Veneto
59
28
Friuli-Venezia Giulia
63
25
Emilia-Romagna
64
24
Toscana
62
22
Umbria
51
29
Marche
29
53
Lazio
37
38
Abruzzo
34
35
Molise
21
40
Campania
28
40
Puglia
58
8
Basilicata
33
23
Calabria*
32
33
Sicilia
46
22
Sardegna*
47
77% natiowide pooled data
38
*Regions without complete survey coverage
Cancer screening uptake – PASSI data
12
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
(Alto Adige), , Lombardia and Liguria, northern regions with
a low proportion of tests performed within the programmes,
and Basilicata, which among southern regions has a low pro-
portion of spontaneous screening.
From 2008 to 2013 coverage increased (p <0.001). The trend
was appreciable in all three geographic areas, but was stronger
in the South. The trend was entirely due to the increase in
women who had a test within screening programmes (p <0.001),
while the coverage due to spontaneous screening showed a
slight decrease (p=0.052) (figure 2).
Coverage was higher in 35-49 year-old women, married or
with a stable partner, with a medium or high educational level,
without economic problems, and who are Italian or come from
industrialized countries (compared to immigrants from high
migration pressure countries).
Women 50-64 years old, married or with a stable partner,
with low education, and who are immigrants from high mi-
gration pressure countries more frequently performed the
test within the screening programmes. On the contrary,
women aged 25-34, highly educated, without economic
problems, with Italian nationality or coming from industri-
alized countries, more frequently performed the test sponta-
neously (figure 3,table 1). Multivariate analysis confirmed all
the associations found (table 1).
2008 2009 2010 2011 2012 2013 2014
coverage %
20
0
40
60
80
100 Figure 2. Time series of cervi-
cal cancer screening coverage
in 25-64 year-old women
within screening programmes
or spontaneous screening.Na-
tionwide pooled data, PASSI
2010-2013.
Figura 2. Andamento tempo-
raledella coperturadello scree-
ning cervicale (organizzato o
spontaneo),tra le donne di 25-
64 anni. Pool, PASSI 2010-
2013.
overall
within the screening programme
spontaneus screening
Within a screening programme Spontaneous screening
OR 95%CI p-value OR 95%CI p-value
Age
25-34 1.00 1.00
35-49 1.25 1.18 1.32 0.000 1.24 1.17 1.32 0.000
50-64 1.65 1.55 1.75 0.000 0.77 0.72 0.82 0.000
Married/with stable partner
yes 1.00 1.00
no 0.77 0.73 0.80 0.000 0.80 0.76 0.84 0.000
Education level
none/elementary 1.00 1.00
middle school 1.21 1.11 1.32 0.000 1.49 1.34 1.65 0.000
secondary school 1.17 1.07 1.27 0.000 2.04 1.84 2.26 0.000
academic degree 1.03 0.94 1.14 0.478 2.41 2.16 2.70 0.000
Economic difficulties
major 1.00 1.00
minor 1.10 1.03 1.18 0.004 1.19 1.10 1.28 0.000
none 1.12 1.05 1.20 0.001 1.35 1.26 1.46 0.000
Nationality
Italian 1.00 1.00
foreign 1.47 1.35 1.60 0.000 0.52 0.47 0.57 0.000
Table 1. Logistic regression model to analyze the characteristics associated with cervical cancer screening coverage. Nationwide pooled data, PASSI 2010-2013.
Tabella 1. Copertura di un test per la prevenzione dei tumori del collo dell’utero negli ultimi tre anni. Pool,PASSI 2010-2013.
Cancer screening uptake – PASSI data
13
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
In those regions with well-implemented screening programmes,
i.e., in which at least 75% of the target population declared they
had received the invitation letter, coverage was higher than in
those with incomplete programme activation, i.e., 87% vs
72% (p <0.001). Furthermore, in regions with well-imple-
mented programmes the difference in coverage between women
with a degree and women with lower education was 16% and
the difference between women with major economic difficul-
ties and with no economic problems was 11%; in regions with
incomplete programme activation these differences were 38%
and 20%, respectively.
Breast cancer screening test coverage
Overall, 70% of the 50-69 year-old women had a mammog-
raphy in the two years before the interview. There was a de-
creasing North-South trend (81% in the North, 77% in the
Centre, and 54% in the South and Islands).
Fifty-one percent of the women performed the test within a
screening programme for free and 19% performed the test
spontaneously paying it entirely or in part.The coverage due to
spontaneous testing was similar in the three geographic areas,
while the part due to organized screening varied (figure 4, p. 14).
From 2008 to 2013 mammography coverage slightly increased
(p=0.060).The increase was present in all three geographic ar-
eas and both in organized programmes and spontaneous
screening, but was stronger in the South and in spontaneous
activity (figure 5, p. 14).
Coverage was higher in 50-59 year-old women, married or
with a stable partner, with high education, without economic
problems, and who are Italian or come from industrialized
countries (compared to immigrants from high migration pres-
sure countries). Women 60-69 years old, with poor education,
without economic problems, and who are immigrants from
high migration pressure countries more frequently performed
the test within the screening programmes. On the contrary,
women 50-59 years old, with a degree, and who are Italian or
come from industrialized countries, more frequently per-
formed the test spontaneously (figure 6,table 2, p. 15). Mul-
tivariate analysis confirmed all the associations found (table 2).
In regions with well-implemented breast cancer screening
programmes, i.e., in which at least 75% of the target popula-
tion declared they had received the invitation letter, coverage
was higher than in regions with incomplete programme acti-
vation, i.e., 81% vs 60% (p <0.001). Furthermore, in regions
with well-implemented programmes the difference in cover-
age between women with a degree and women with lower ed-
ucation was 8% and the difference between women with ma-
jor economic difficulties and those with no economic
problems was 13%; in regions with incomplete programme ac-
tivation the difference was 37% in both cases.
Figure 3. Proportion of 25-64 year-old women who had a Pap smear or HPV test in the three years before the interview, within screening programmes or spontaneously,
according to socio-economic characteristics. Nationwide pooled data, PASSI 2010-2013.
Figura 3. Proporzione di donne di età 25-64 anni che hanno effettuato un Pap test o un test HPV negli ultimi tre anni, all’interno dei programmi di screening o spon-
taneamente, secondo lo stato socioeconomico. Pool, PASSI 2010-2013.
WITHIN A SCREENING PROGRAMME
AGE
25-34
35-49
50-64
MARRIED/WITH STABLE PARTNER
yes
no
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign 47
0% 20% 40% 60% 80% 100%
40
40
40
40
40
46
43
43
41
41
38
36
35
33
SPONTANEOUS SCREENING
AGE
25-34
35-49
50-64
MARRIED/WITH STABLE PARTNER
yes
no
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign 24
0% 20% 40% 60% 80% 100%
41
40
37
29
29
32
43
36
21
45
34
38
38
38
* Italian: Italian or people coming from industrialized countries –foreign: immigrants from high migration pressure countries
Cancer screening uptake – PASSI data
14
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Colorectal cancer screening test coverage
Overall, 38% of people aged 50-69 years were covered for col-
orectal cancer screening; 33% had a faecal occult blood test in
the two years before and 13% a colonoscopy five years before
the interview (these data do not include the region of
Piemonte). There was a decreasing North-South trend (59%
in the North, 41% in the Centre, and 17% in the South and
Islands).
Thirty-one percent performed the test within a screening pro-
gramme for free and 7% performed the test, mainly a
colonoscopy, spontaneously, paying it entirely or in part. The
vast majority of occult blood tests was performed within screen-
ing programmes, while about half of the colonoscopies or sig-
moidoscopies were preformed in spontaneous testing settings.
The coverage due to spontaneous testing was similar in the
three geographic areas, while the part due to organized screen-
ing varied (figure 7, p. 16).
From 2010 to 2013, colorectal cancer screening test coverage
rapidly increased (p <0.001). The increase was present in all
three geographic areas and both in organized programmes and
%
60
80
100
40
20
0
Figure 4. Breast cancer screening test coverage. Proportion of 50-69 year-old women who had a mammography in the two years before the interview, within screen-
ing programmes or spontaneously, by region. Nationwide pooled data, PASSI 2010-2013.
Figura 4. Copertura dello screening mammografico.Proporzione di donne di età 50-69 anni che hanno eseguito una mammografia negli ultimi due anni, all’interno dei
programmi di screening o spontaneamente. Pool, PASSI 2010-2013.
51
19
within the screening programme spontaneous screening
nationwide pooled data
North
56
21
Centre
33
21
South
66
5
Valle d’Aosta
59
13
Piemonte
44
32
Liguria
67
18
Lombardia*
75
8
Trento
55
19
Bolzano
62
21
Veneto
67
17
Friuli-Venezia Giulia
73
13
Emilia-Romagna
71
9
Toscana
67
12
Umbria
49
29
Marche
46
28
Lazio
36
20
Abruzzo
53
15
Molise
21
23
Campania
38
24
Puglia
61
5
Basilicata
28
17
Calabria*
34
19
Sicilia
40
20
Sardegna*
64
70% natiowide pooled data
17
2008 2009 2010 2011 2012 2013 2014
coverage %
20
0
40
60
80
100 Figure 5. Time series of breast
cancer screening coverage in
50-69 year-old women. with-
in screening programmes or
spontaneous screening. Na-
tionwide pooled data, PASSI
2010-2013.
Figura 5. Andamentodella co-
pertura dello screening mam-
mografico (organizzato o spon-
taneo), tra le donne di 50-69
anni. Pool, PASSI 2010-2013.
overall
within the screening programme
spontaneus screening
*Regions without complete survey coverage
Cancer screening uptake – PASSI data
15
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
spontaneous screening, but was stronger in the northern and
central regions. The increase was totally due to tests performed
within the organized screening programmes (figure 8, p. 16).
Coverage was higher in 60-69 year-olds men, without eco-
nomic problems, and Italian or coming from industrialized
countries (compared to immigrants from high migration pres-
sure countries).
People aged 60-69, with poor education and without economic
problems more frequently performed the test within the screen-
ing programmes. On the contrary, those with higher education,
Figure 6. Proportion of 50-69 year-old women who had a mammography in the two years before the interview, within screening programs or spontaneously, accord-
ing to socio-economic characteristics. Nationwide pooled data, PASSI 2010-2013.
Figura 6. Proporzione di donne di età 50-69 anni che hanno eseguito una mammografia negli ultimi sue anni, all’interno dei programmi di screening o spontaneamente,
secondo lo stato socioeconomico. Pool, PASSI 2010-2013.
WITHIN A SCREENING PROGRAMME
AGE
50-59
60-69
MARRIED/WITH STABLE PARTNER
yes
no
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign
0% 20% 40% 60% 80% 100%
55
51
51
53
51
52
49
49
52
52
43
46
47
48
SPONTANEOUS SCREENING
AGE
50-59
60-69
MARRIED/WITH STABLE PARTNER
yes
no
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign
0% 20% 40% 60% 80% 100%
23
25
19
15
14
18
12
12
30
17
17
24
20
38
* Italian: Italian or people coming from industrialized countries –foreign: immigrants from high migration pressure countries
Within a screening programme Spontaneous screening
OR 95%CI p-value OR 95%CI p-value
Age
50-59 1.00 1.00
60-69 1.18 1.11 1.26 0.000 0.61 0.56 0.66 0.000
Married/with stable partner
yes 1.00 1.00
no 0.84 0.78 0.90 0.000 0.83 0.76 0.92 0.000
Education level
none/elementary 1.00 1.00
middle school 0.99 0.91 1.08 0.851 1.28 1.13 1.45 0.000
secondary school 0.90 0.82 0.98 0.018 1.93 1.70 2.19 0.000
academic degree 0.72 0.63 0.81 0.000 2.55 2.18 2.98 0.000
Economic difficulties
major 1.00 1.00
minor 1.31 1.19 1.44 0.000 1.18 1.03 1.35 0.016
none 1.74 1.58 1.92 0.000 1.30 1.14 1.49 0.000
Nationality
Italian 1.00 1.00
foreign 1.14 0.94 1.38 0.171 0.47 0.35 0.64 0.000
Table 2. Logistic regression model to analyze the characteristics associated with breast cancer screening coverage. Nationwide pooled data, PASSI 2010-2013.
Tabella 2. Modello di regressione logistica per la copertura di una mammografia preventiva entro gli ultimi due anni. Pool, PASSI 2010-2013.
Cancer screening uptake – PASSI data
16
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Italian nationality or coming from industrialized countries, and
without economic problems more frequently performed the
test spontaneously (figure 9,table 3). Multivariate analysis con-
firmed all the associations found (table 3).
In regions with well-implemented colorectal cancer screening
programs, i.e., in which at least 75% of the target population
declared they had received the invitation letter, coverage is
higher than in regions with incomplete programme activation:
59% vs 14% (p <0.001). Furthermore, whereas in regions with
well-implemented programmes those with a low level of edu-
cation have higher coverage than people who hold a degree
(+8%), the situation is exactly the opposite in regions where
programmes are not well-implemented, where people with an
academic degree have higher coverage (+24%) than those
with a lower level of education. There is also a reduction in the
difference in coverage for the economically disadvantaged:
29% vs 41%.
CONCLUSIONS
About three fourths of the female target populations are cov-
%
60
80
100
40
20
0
Figure 7. Colorectal cancer screening test coverage. Proportion of 50-69 year-old people who had a faecal occult blood test in the two years before the interview or
colonoscopy/sigmoidoscopy in the five years before the interview, within screening programmes or spontaneously,by region. Nationwide pooled data, PASSI 2010-2013.
Figura 7. Copertura dello screening colorettale.Proporzione di persone di età 50-69 anni che hanno eseguito un test SOF (sangue occulto fecale) negli ultimi due anni
o una colonscopia negli ultimi cinque anni, all’interno dei programmi di screening o spontaneamente. Pool, PASSI 2010-2013.
31
7
within the screening programme spontaneous screening
nationwide pooled data
North
31
10
Centre
11
6
South
64
6
Valle d’Aosta
31
6
Piemonte°
17
8
Liguria
59
5
Lombardia*
56
10
Trento
11
15
Bolzano
58
7
Veneto
49
8
Friuli-Venezia Giulia
65
5
Emilia-Romagna
56
6
Toscana
53
6
Umbria
26
13
Marche
12
12
Lazio
14
12
Abruzzo
40
8
Molise
10
6
Campania
6
7
Puglia
27
4
Basilicata
8
7
Calabria*
8
4
Sicilia
24
6
Sardegna*
52
38% natiowide pooled data
7
2008 2009 2010 2011 2012 2013 2014
coverage (%)
20
0
40
60
80 Figure 8. Time series of col-
orectal cancer screening cov-
erage in 50-69 year-old people,
within screening programmes
or spontaneous screening. Na-
tionwide pooled data, PASSI
2010-2013.
Figura 8. Andamento della
copertura dello screening colo-
rettale (organizzato o sponta-
neo), tra le persone di 50-69
anni. Pool, PASSI 2010-2013.
* Regions without complete survey coverage
° Piemonte region adopted a different screening strategy based on sygmoidoscopy once in a life at the age of 58
change questionnaire
overall
within the screening programme
spontaneus screening
Cancer screening uptake – PASSI data
17
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Within a screening programme Spontaneous screening
OR 95%CI p-value OR 95%CI p-value
Age
50-59 1.00 1.00
60-69 1.36 1.30 1.43 0.000 1.00 0.92 1.09 0.996
Gender
male 1.00 1.00
female 1.01 0.97 1.05 0.634 0.79 0.73 0.86 0.000
Education level
none/elementary 1.00 1.00
middle school 0.98 0.92 1.04 0.527 1.20 1.06 1.36 0.004
secondary school 0.88 0.82 0.94 0.000 1.59 1.40 1.80 0.000
academic degree 0.63 0.58 0.70 0.000 2.17 1.86 2.52 0.000
Economic difficulties
major 1.00 1.00
minor 1.37 1.27 1.48 0.000 1.05 0.91 1.20 0.498
none 2.58 2.39 2.78 0.000 1.13 0.99 1.29 0.071
Nationality
Italian 1.00 1.00
foreign 1.12 0.94 1.33 0.192 0.48 0.32 0.71 0.000
ered by cervical and breast cancer screening, although there are
significant differences between northern-central and southern
Italy. Colorectal cancer screening coverage is still below 40%.
The role of spontaneous screening is relevant for female can-
cer screening, in particular cervical cancer screening, but the
presence of well-implemented organized programmes makes it
possible to reach high coverage levels and reduce inequalities
in the access to evidence-based screening.
Conflicts of interests: none declared
Table 3. Logistic regression model to analyze the characteristics associated with colorectal cancer screening coverage. Nationwide pooled data, PASSI 2010-2013.
Tabella 3. Modello di regressione logistica per la copertura di un esame preventivo per la diagnosi precoce dei tumori colorettali entro i tempi raccomandati. Pool, PASSI
2010-2013.
Figure 9. Proportion of 50-69 year-old people who had a foecal occult test in the two years before the interview or a colonoscopy in the five years before the interview,
within screening programs or spontaneously,according to socio-economic characteristics. Nationwide pooled data, PASSI 2010-2013.
Figura 9. Proporzione di persone di età 50-60 anni che hanno eseguito un test SOF (sangue occulto fecale) negli ultimi due anni o una colonscopia negli ultimi cinque
anni, all’interno dei programmi di screening o spontaneamente, secondo lo stato socioeconomico. Pool,PASSI 2010-2013.
WITHIN A SCREENING PROGRAMME
AGE
50-59
60-69
GENDER
male
female
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign
0% 20% 40% 60% 80%
39
35
31
31
31
31
31
32
32
22
28
29
27
27
SPONTANEOUS SCREENING
AGE
50-59
60-69
GENDER
male
female
EDUCATION LEVEL
none/elementary
middle school
secondary school
academic degree
ECONOMIC DIFFICULTIES
major
minor
none
NATIONALITY*
Italian
foreign
0% 20% 40% 60% 80%
7
7
7
7
5
4
11
6
6
6
8
8
8
8
* Italian: Italian or people coming from industrialized countries –foreign: immigrants from high migration pressure countries
References/Bibliografia
1. Council Recommendation of 2 December 2003 on cancer screen-
ing. 2003/878/EC.
2. Ministero della salute, Direzione generale della prevenzione. Rac-
comandazioni per la pianificazione e l’esecuzione degli screening di
popolazione per la prevenzione del cancro della mammella, del can-
cro della cervice uterina e del cancro del colon retto. Ministero della
salute, Roma 2005. [www.osservatorionazionalescreening.it/ons/
documentazione/raccomandazioni/screening_vers_completa.pdf].
3. Ronco G, Giubilato P, Carozzi F et al. and the Cancer Screening Sur-
vey working group. Extension of organized cervical cancer screen-
ing programmes in Italy and their process indicators, 2011-2012
activity. Epidemiol Prev 2015;3(Suppl 1):61-76.
4. Ventura L, Giorgi D, Giordano L et al. Mammographic breast can-
cer screening in Italy: 2011-2012 survey. Epidemiol Prev
2015;3(Suppl 1):21-29.
5. Zorzi M, Da Re F, Mantellini P, Naldoni C, Sassoli de’ Bianchi P, Se-
nore C, Turrin A, Visioli CB, Zappa M and the Italian Colorectal
Cancer Screening Survey group. Screening for colorectal cancer
in Italy, 2011-2012 survey. Epidemiol Prev 2015;3(Suppl 1):93-
107.
6. Osservatorio nazionale screening. Gli screening sul campo: Lom-
bardia. [www.osservatorionazionalescreening.it/node/80].
7. Giorgi Rossi P, Segnan N. Studi sul confronto di efficacia dei percorsi
assistenziali negli screening oncologici. ONS GISCi Conference,
Venice, 2010. [www.gisci.it/documenti/convegni/venezia2010/
20100526/giorgi_rossi_workshop_ons_20100526.pdf].
8. Chini F, Mancini E, Cogo C et al. Le indagini campionarie sulla “co-
pertura” del Pap-test: appunti e spunti metodologici. Florence: GI-
SCi, 2007. [www.gisci.it/documenti/documenti_gisci/copertura.pdf].
9. Istituto nazionale di statistica (Istat). La prevenzione dei tumori fem-
minili in Italia; il ricorso a Pap test e mammografia: anni 2004-2005.
Published online in 2006. [www.istat.it/salastampa/comunicati/
non_calendario/20061204_00/testointegrale.pdf].
10. Gruppo tecnico nazionale PASSI. Sistema di sorveglianza PASSI
(Progressi delle Aziende Sanitarie per la Salute in Italia): risultati 2007.
Rapporto Istisan, 9/31. Istituto superiore di sanità, Roma 2009.
11. Gruppo tecnico nazionale PASSI. Sistema di sorveglianza PASSI
(Progressi delle Aziende Sanitarie per la Salute in Italia). Rapporto
Istisan 7/30. Istituto superiore di sanità, Roma 2007.
12. Baldissera S, Ferrante G, Quarchioni E et al. Field substitution of
nonresponders can maintain sample size and structure without al-
tering survey estimates –the experience of the Italian behavioral
risk factors surveillance system (PASSI). Ann Epidemiol 2014;
24(4):241-45.
13. Baldissera S, Campostrini S, Binkin N et al. and the PASSI Coordi-
nating Group. Features and initial assessment of the Italian be-
havioral risk factor surveillance system (PASSI), 2007-2008. Prev
Chronic Dis 2011;8(1):1-8.
14. Binkin N, Gigantesco A, Ferrante G, Baldissera S. Depressive symp-
toms among adults 18-69 years in Italy: results from the Italian be-
havioural risk factor surveillance system, 2007. Int J Public Health
2009;55(5),479-88.
NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 18
Cancer screening uptake – PASSI data anno 39 (3) maggio-giugno 2015
e
&
p
Breast cancer
screening
Breast cancer screening: 2011-2012 survey
20
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Theoretical or potential or nominal extension: percentage of women involved in a screening programme out of the
total female population in the 50-69 age range resident in the area covered by an organized screening programme.
Actual extension or Extension of invitations: percentage of women involved in a screening programme out of the
total female population in the 50-69 age range who actually received an invitation to screening during the analyzed
period.
Compliance with invitation or Crude attendance: number of respondents out of the total number of invited women
excluding undelivered invitations.
Adjusted attendance (compliance): number of respondents out of the total number of invited women excluding
undelivered invitations and women with a recent mammography (undergone during the last 12 months).
Recall rate: percentage of women recalled for further assessments out of the total number of women attending.
Total detection rate: number of women with screen-detected cancer out of 1,000 screened women.
Benign-malignant ratio: ratio between benign and malignant histological diagnosis, independently of the procedure
of diagnosis.
Detection rate for cancers ≤10mm: number of women with screen-detected cancers smaller or equal to 10 mm out
of 1,000 screened women.
Percentage of ductal carcinoma in situ: percentage of ductal carcinoma in situ out of the total number of screen-
detected cancers.
Piemonte: Nereo Segnan, Centro prevenzione oncologia – CPO
Piemonte, Torino
Provincia Autonoma di Bolzano (Alto Adige): Antonio Fanolla,
Assessorato alla sanità, Osservatorio epidemiologico, Provincia
autonoma di Bolzano, Bolzano
Provincia Autonoma di Trento (Trentino): Sivano Piffer, Giovanni
De Pretis, Osservatorio epidemiologico, Azienda provinciale
per i servizi sanitari, Trento
Puglia: Vincenzo Pomo, Cinzia Annatea Germinario, Agenzia
regionale sanità, Regione Puglia, Bari
Sardegna: Pierina Thanchis, Assessorato dell’igiene e sanità
e dell’assistenza sociale, Regione Sardegna, Cagliari
Toscana: Paola Mantellini, Istituto per lo studio e la prevenzione
oncologica, Regione Toscana, Firenze
Umbria: Mariadonata Giaimo, Direzione regionale salute, coesione
sociale e società della conoscenza, Regione Umbria, Perugia
Valle D’Aosta: Gabriella Furfaro, Servizio dipendenze patologiche,
salute mentale e promozione della salute, Aosta
Veneto: Chiara Fedato, Registro tumori del Veneto, Padova
Abruzzo: Tamara Agostini, Direzione politiche della salute, Regione
Abruzzo, Pescara
Basilicata: Vincenzo Barile, Angelo Sigillito, Sergio Schettini, AO
San Carlo, Potenza
Calabria: Liliana Rizzo, Dipartimento Tutela della salute e politiche
sanitarie, Regione Calabria, Catanzaro
Campania: Renato Pizzuti, Osservatorio epidemiologico regionale,
Assessorato alla sanità, Regione Campania, Napoli
Emilia-Romagna: Carlo Naldoni, Assessorato alle politiche
per la salute, Regione Emilia-Romagna, Bologna
Friuli-Venezia Giulia: Nora Coppola, Direzione centrale salute,
integrazione socio sanitaria, politiche sociali e famiglia, Regione
Friuli-Venezia Giulia, Trieste
Lazio: Alessandra Barca, Lazio sanità, Agenzia di sanità pubblica,
Roma
Liguria:Luigina Bonelli, Gabriella Paoli, Istituto nazionale
per la ricerca sul cancro, Genova
Lombardia: Direzione generale salute, Regione Lombardia, Milano
Marche: Lucia Di Furia, Servizio salute, Regione Marche, Ancona
Molise: Ospedale Cardarelli, Regione Molise, Campobasso
Italian breast cancer screening survey group:
Gruppo di studio sullo screening mammografico:
Glossary/Glossario
WWW.EPIPREV.IT
21
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Mammographic breast cancer screening
in Italy: 2011-2012 survey
Lo screening mammografico in Italia: survey
2011-2012
Leonardo Ventura,1Daniela Giorgi,2Livia Giordano,3Alfonso Frigerio,4Paola Mantellini,1Marco Zappa1
and the Italian breast cancer screening survey group
Abstract
This report is an update of a number of papers that have been published by the ONS (Osservatorio
nazionale screening, National centre for screening monitoring) since 2002. Data for the survey come
from several programmes that may have changed over time, and may have different settings of or-
ganization and management.
During 2011-2012, a slight increase in actual extension was recorded compared to the previous years.
Currently, all Italian regions have implemented screening programmes. In 2011-2012, almost
5,300,000 women aged 50-69 years were invited to have a screening mammogram, and almost
3,000,000 were screened. While potential extension was 94.4%, actual extension was 73.3%. An im-
balance in extension is still present when comparing northern and central Italy, that have an actual
screening extension of 94% and 86% respectively, to southern Italy, that has less than 40%.
During the last few years, participation rates have been substantially stable, at around 56% for crude
rate, and 60% for adjusted rate, respectively. Women actually screened during 2011-2012 were
38.9% of the national target population. Referral rates of 9.2% at first screening and 4.7% at repeat
screening were recorded, showing an increasing trend in recent years. Detection rate was 4.8x1,000
at first screening and 4.4x1,000 at repeat screening, while benign to malignant surgical biopsy ratio
for first and repeat screening was 0.2 and 0.1, respectively. Detection rate of small (≤10 mm) invasive
cancers was 1.3x1,000 at first screening and 1.4x1,000 at repeat screening; the proportion of in situ
carcinomas was 13.3% and 12.0% for first and repeat screening, respectively. Indicators by 5-year age
group confirm greater diagnostic problems at younger ages (50-54 years), with higher referral rates
and a substantially lower detection rate as compared to older age groups.
Epidemiol Prev 2015; 39(3) Suppl 1: 21-29)
Keywords: breast cancer screening, breast, survey, Italy
1Istituto per lo studio
e la prevenzione
oncologica (ISPO), Firenze
2SC Epidemiologia
e screening, ASL 2, Lucca
3Unità di epidemiologia
dei tumori, CPO Piemonte,
AOU Città della salute
e della scienza, Torino
4SSCVD Senologia
di screening, AOU Città
della salute e della scienza,
Torino
Corresponding author
Marco Zappa
m.zappa@ispo.toscana.it
Riassunto
Questo rapporto rappresenta un aggiornamento di precedenti pubblicazioni dell’ONS (Osservatorio na-
zionale screening) a partire dal 2002. I dati della survey derivano da programmi anche molto diversi
tra loro, che possono rispecchiare situazioni differenziate, sia per il livello di esperienza sia per i mo-
delli organizzativi e gestionali.
Nel periodo 2011-2012 si registra un lieve incremento dell’estensione teorica rispetto agli anni pre-
cedenti. Allo stato attuale tutte le Regioni italiane hanno implementato programmi di screening.
Nel 2011-2012 quasi 5.300.000 donne di età 50-69 anni sono state invitate a sottoporsi alla mam-
mografia di screening, e circa 3.000.000 sono state esaminate. L’estensione teorica è risultata pari
a 94,4%, mentre quella effettiva è stata del 73,3%. Il confronto tra le Regioni del Nord e del Cen-
tro con quelle del Sud Italia rivela ancora uno squilibrio nell’estensione dello screening: mentre al
anno 39 (3) maggio-giugno 2015
e
&
p
Breast cancer screening: 2011-2012 survey
22
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
In this paper the performances of Italian mammographic
screening programmes for the biennium 2011-2012 are re-
ported. Since the early 1990s, GISMa (Gruppo italiano per lo
screening mammografico, Italian group for mammography
screening) has carried out yearly surveys on the implementa-
tion of screening programmes in Italy. Starting from 2002, the
results of these surveys have been published in the annual re-
ports of the Osservatorio nazionale screening (ONS, National
centre for screening monitoring). Moreover, monitoring, com-
parisons and evaluation activities have led to the publication
of updated operating reports of process indicators for mam-
mography screening.1In Italy, activation of mammography
screening programmes is regulated by the Ministry of Health’s
new guidelines.2These guidelines recommend that women in
the 50-69 year age range be personally invited to undergo
mammography every two years, and require a monitoring sys-
tem and quality evaluation activity for each phase of the pro-
gramme. Recently, two regions (Emilia-Romagna and Pied-
mont) expanded the lower age of invitation to 45 (with an
annual invitation). Several programmes continue the invitation
up to age 74-75 with a two year interval.
This report is an update of previous papers published by the
ONS; it is available on the ONS website (www.osservatorion-
azionalescreening.it).3-12
GUIDELINES FOR DATA INTERPRETATION
Data referring to the 2011-2012 activity are reported, strati-
fied by region, 5-year age groups, and, where applicable, by first
and repeat screening.
It should be considered that these are summarized data, that
may reflect different situations, both as to varying levels of ex-
perience and dissimilar settings of organization and manage-
ment. Therefore, when interpreting these results, it is impor-
tant to bear in mind some critical aspects inherent to the data:
■
not all programmes were able to separate first and repeat screen-
ing tests, so for these programmes results were assigned to the
round that includes the majority of the screened women;
■
a few programmes are not yet able to provide data stratified
by five-year age group, so the age-stratified results provided re-
late to a subset of programmes;
■
an important aspect to consider is the uneven completeness
of the information provided by different programmes. The re-
sult of this is that the denominator of different indicators can
vary within each programme.
EXTENSION
With the term “extension” we define the percentage of women
involved in a screening programme out of the total female pop-
ulation in the 50-69 age range resident in the area.
For a deeper understanding of screening activity we considered
two types of extension:
■
potential extension (or programme extension), referring to
eligible women residing in areas covered by an organized screen-
ing programme;
■
actual extension (or invitation extension), related to women
who were actually sent an invitation to screening during the an-
alyzed period. Actual extension is calculated according to new
rules introduced in 2008, in order to consider undelivered in-
vitations and women excluded before invitation: the former are
subtracted from the numerator (115,812 women, in 2011-2012,
all Italy) and the latter from the denominator (420,830
women, in 2011-2012, all Italy).
In 2011-2012, the total target population was in excess of 7.5
million, and potential and actual extension were 94.4% and
73.3%, respectively.
For some regions (see table 1 and figure1) a discrepancy is ev-
ident between the two figures, indicating a substantial difficulty
in inviting all the target population within the protocol inter-
val of two years. In 2011-2012, about three out of four women
(73.3%) were actively invited to screening: actual extension
showed a slight increase as compared to the previous year
(69.1% in 2010). A strong imbalance in the screening offer still
persists between northern-central and southern Italy. In the
northern and central Italian regions, actual extension is rather
good (93.9% in the North and 86.1% in the Centre). In the
South the value is much lower (39.6%) although slightly
higher compared to 2010 (37.8%). Within the southern area,
two small regions (Basilicata and Molise) showed good results,
comparable to the Centre-North. On the other hand, a very
difficult situation still persists in Abruzzo.
Although some regions show good mean results, a large inter-
Nord e al Centro l’estensione effettiva è rispettivamente del 94% e dell’86%, nel Sud il valore registrato è inferiore al 40%.
Negli ultimi anni i tassi di partecipazione sono rimasti sostanzialmente stabili, intorno al 56% per l’adesione grezza e al 60%
per l’adesione corretta. Le donne esaminate nel 2011-2012 sono state il 38,9% della popolazione obiettivo.
Ai primi esami si è registrato un tasso di richiami del 9,2%, del 4,7% agli esami successivi, rivelando un trend in aumento negli
ultimi anni. Il tasso di identificazione è risultato pari a 4,8x1.000 ai primi esami e 4,4x1.000 agli esami successivi, mentre il rap-
porto benigni/maligni (B/M) registrato è stato 0,2 e 0,1 rispettivamente per i primi e per gli esami successivi. Il tasso di identifi-
cazione dei tumori invasivi ≤10 mm è risultato pari a 1,3x1.000 ai primi esami e 1,4x1.000 ai successivi; la percentuale di
carcinomi duttali in situ è stata del 13,3% e del 12% rispettivamente per i primi esami e per i successivi.
Gli indicatori per fasce di età quinquennali confermano la presenza di maggiori problemi diagnostici nelle donne più giovani (50-54
annni), con tassi di richiamo più elevati e un tasso di identificazione sostanzialmente più basso rispetto ai gruppi di età più anziani.
Epidemiol Prev 2015; 39(3) Suppl 1: 21-29)
Parole chiave: screening mammografico, mammella, survey, Italia
Breast cancer screening: 2011-2012 survey
23
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Regions Target Theoretical N. of invited Actual 10th-90th
population extension (%) women extension (%) percentile
Valle d’Aosta 16,205 100.0 14,456 89.2
Piemonte 585,794 100.0 436,849 74.6 49.8-98.0
Liguria 208,177 100.0 127,519 61.3 35.2-119.7
Lombardia 1,121,416 100.0 1,085,618 96.8 83.0-104.6
Trento 64,183 100.0 68,380 106.5
Bolzano 52,898 100.0 53,866 101.8
Veneto 525,161 100.0 525,775 100.1 91.7-120.2
Friuli-Venezia Giulia 158,663 100.0 163,341 102.9
Emilia-Romagna 480,227 100.0 541,826 112.8 99,9-121,7
North 3,212,721 100.0 3,017,630 93.9 69.2-119.7
Toscana 474,441 100.0 462,567 97.5 93.1-104.8
Umbria 101,751 100.0 104,266 102.5
Marche 171,086 100.0 152,910 89.4 49.6-114.1
Lazio 712,105 94.9 536,607 75.4 35.,3-115.6
Centre 1,459,382 100.0 1,256,350 86.1 41.6-111.9
Abruzzo 167,315 47.6 10,622 6.3 8.0-19.1
Molise 38,503 100.0 32,347 84.0
Campania 674,303 75.8 208,824 31.0 19.4-75.6
Puglia 507,678 100.0 246,351 48.5
Basilicata 71,226 100.0 63,735 89.5
Calabria 234,761 94.0 68,585 29.2 17.5-58.5
Sicilia 609,254 75.2 262,131 43.0 21.2-95.3
Sardegna 217,081 100.0 104,673 48.2 31.1-161.9
South 2,520,119 83.2 997,268 39.6 18.0-92.2
Italy 7,192,221 94.4 5,271,248 73.3 33.2-114.1
Table 1. Potential and actual
extension of Italian mammo-
graphic screening programmes.
For regions with more than 3
local programmes, the tenth
(p10) and the ninetieth (p90)
percentiles of actual exten-
sion are reported.
Tabella 1. Estensione, teorica
e reale, dei programmi di scree-
ning mammografico.Per le Re-
gioni che hanno più di tre pro-
grammi locali sono forniti il 10°
e il 90° percentile dell’esten-
sione reale.
Figure 1. Actual extension (%) of mammography
screening. Years 2011-2012.
Figura 1. Estensione aggiustata (%) dei programmi
di screening mammografico.Anni 2011-2012.
<50%
50-74%
75-94%
>94%
anno 39 (3) maggio-giugno 2015
e
&
p
Breast cancer screening: 2011-2012 survey
24
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
nal variation is evident (among local programmes within each
region). This is the case of Piemonte, Marche, and Lazio,
where the gap between the 10th and 90th percentile varies
from two to four times. The total gap between the 10th and
90th percentile remained unchanged compared to 2010.
During the last few years, screening invitation for women be-
longing to the age groups 45-49 and 70-74 has increased. In
2011-2012, out of a total target population of about 4,800,000
in the 45-49 year age group, 7.9% were invited, corresponding
to 379,701 women. In women over 70 years of age, consider-
ing a target population of 1,671,000 women in the 70-74 year
age range, 13.6% were invited to screening, corresponding to
227,387 women.
ATTENDANCE
The number of women invited and responding to the invita-
tion is reported in table 2. Overall, more than 5 million
women were invited in the biennium 2011-2012. This was an
increase in comparison with the previous biennium when
fewer than 5,000,000 were invited. Of all women invited in
2011-2012, almost 3 million attended. In the same table, the
mean volumes of activity of each programme for 2011-2012
are also reported. Generally speaking, the volume of activity
could be considered as an indirect indicator of the level of ex-
perience of the medical and technical personnel involved.
Most Italian regions (with the exception of the province of
Trento and Lombardia) did not attain, on a regional mean ba-
sis, the desirable level of at least 20,000 examinations per lo-
cal programme (although several local programmes actually
did). In a few cases (Friuli-Venezia Giulia, Umbria, Basilicata,
Puglia) data were collected at a regional level, from local pro-
grammes of limited sizes, as several programmes work at vol-
umes of activity that are too low (below 10,000 or even 5,000
examinations per year) to assure an appropriate level of expe-
rience of the personnel involved. In evaluating these figures,
two (opposite) considerations should be taken into account:
■
in each programme more than one radiological centre can be
present so that the actual number of mammograms is lower;
■
in many cases the radiological screening centre also performs
mammograms on “spontaneous” patients (i.e., non-invited, self-
referred, or clinical patients). In such cases the actual number
of mammograms performed could be much higher than it ap-
pears from the screening files. In some settings, a low volume
of mammograms is justified by the small regional target pop-
ulation (Valle d’Aosta, Molise), but in some regions it is prob-
ably due to management choices that should be re-evaluated.
Table 3 shows the crude and adjusted attendance rates for Italy,
for Italian macro-areas, and for each region. Screening pro-
gramme attendance is one of the main indicators for the impact
of mammography screening and it is also an indirect indicator
Region Total active Invited women Attendees Mean number of tests
programmes by local unit
Valle d’Aosta 1 14,456 10,124 5,062
Piemonte 9 436,849 274,463 15,248
Liguria 5 127,519 68,309 6,831
Lombardia 15 1,085,618 647,254 21,575
Trento 1 68,380 50,358 25,179
Bolzano 1 53,866 30,698 15,349
Veneto 21 525,775 346,562 8,251
Friuli-Venezia Giulia 1 163,341 95,035 47,518
Emilia-Romagna 11 541,826 352,344 16,016
North 65 3,017,630 1,875,147 14,424
Toscana 12 462,567 315,781 13,158
Umbria 1 104,266 69,026 34,513
Marche 5 152,910 76,358 7,636
Lazio 12 536,607 213,936 8,914
Centre 30 1,256,350 675,101 11,252
Abruzzo 2 10,622 5,086 1,272
Molise 1 32,347 15,593 7,797
Campania 12 208,824 59,654 2,486
Puglia 1 246,351 131,000 65,500
Basilicata 1 63,735 34,087 17,044
Calabria 7 68,585 27,303 1,950
Sicilia 9 262,131 91,002 5,056
Sardegna 8 104,673 45,356 2,835
South 41 997,268 409,081 4,989
Italy 136 5,271,248 2,959,329 10,880
Table 2. Mean volume of ac-
tivity by region. Years 2011-
2012.
Tabella 2. Volume medio di
attività per Regione. Anni
2011-2012.
Breast cancer screening: 2011-2012 survey
25
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
of perceived quality of the programme by the invited women.
Adjusted attendance rate (where women reporting a recent
mammogram outside the programme are excluded from the de-
nominator) is more representative of the real response to invi-
tation of the target population. Currently, GISMa recommended
standards are: ≥50% (acceptable) and ≥70% (desirable) for
crude attendance; ≥60% and ≥75% for adjusted attendance.
In the years 2011-2012, crude and adjusted attendance rates
were 56.1% and 60.9%, respectively, showing a slight im-
provement compared to 2010. As already noted in the previ-
ous years, in 2011-2012 participation rates were substantially
stable, placing the 10th and 90th percentiles close to the val-
ues recorded during the year 2010 both for crude rate (32.1%-
74.0%) and adjusted rate (33.3%-80.0%). Furthermore, it is
encouraging that with an increased extension of invitations, the
attendance rate remained stable.
A large variance of participation exists both among regions and
within each region. It is worth noting that in the large and well
performing regions of the Centre-North (namely Veneto,
Emilia-Romagna, Lombardia, Piemonte, and Toscana) we can
observe a difference ranging from 15 to 25 percentage points
between the tenth and the ninetieth percentiles in the distri-
bution of the programmes’ compliance rate.This means there
is large room for improvement.
Women screened during 2011-2012 were 38.9% of the na-
tional target population. A strong imbalance still persists be-
tween the North, Centre, and South of Italy, with 53.1%,
44.1%, and 16%, respectively.
A decreasing trend towards the South of Italy is evident for
these parameters. All regions showing attendance rates below
the minimal standards are concentrated in the South and Is-
lands areas. In 2011-2012, 8 out of 21 regions (more than one
third of the total) were still below the minimum standards for
crude attendance. Only the province of Trento and Valle
d’Aosta were above the desirable level for these parameters.
Adjusted attendance rates reveal problems of completeness of
data registration. It is important to consider that this param-
eter is often underestimated, as previously mentioned, since
many programmes are unable to provide information about
Region Crude 10th-90th Adjusted 10th-90th
attendance (%) percentile attendance (%) percentile
Valle d’Aosta 70.0 70.7
Piemonte 62.8 53.3-75.3 64.8 56.5-76.4
Liguria 53.6 46.1-63.8 63.7 58.7-75.4
Lombardia 59.6 50.3-68.1 68.6 57.8-77.0
Trento 73.6 77.1
Bolzano 57.0 57.9
Veneto 65.9 53.3-78.4 75.2 63.7-83.1
Friuli-Venezia Giulia 58.2 58.2
Emilia-Romagna 65.0 56.6-77.9 71.0 62.2-78.9
North 62.1 53.0-77.1 68.7 58.7-82.2
Toscana 68.3 58.4-75.4 72.6 64.2-80.1
Umbria 66.2 69.6
Marche 49.9 44.9-57.7 50.7 46.2-57.8
Lazio 39.9 30.3-56.4 43.8 33.6-61.4
Centre 53.7 35.9-74.3 57.5 39.6-78.3
Abruzzo 47.9 37.6-53.0 48.2 37.6-53.8
Molise 48.2 48.5
Campania 28.6 19.1-72.5 31.0 19.2-72.5
Puglia 53.2 55.9
Basilicata 53.5 53.7
Calabria 39.8 23.2-77.0 40.7 23.4-79.5
Sicilia 34.7 19.6-49.7 35.1 20.2-49.7
Sardegna 43.3 35.1-54.6 46.4 37.2-58.9
South 41.0 20.7-54.6 42.7 21.0-58.9
Italy 56.1 30.3-74.2 60.9 33.5-80.1
Table 3. Crude and adjusted
attendance rates, with tenth
and ninetieth percentiles (%).
Years 2011-2012.
Tabella 3. Adesione, grezza e
aggiustata, con il 10° e 90°
percentile. Anni 2011-2012.
Age Crude attendance (%) Adjusted attendance (%)
50-54 52.9 59.1
55-59 57.8 62.7
60-64 60.2 64.5
65-69 57.9 61.9
Total 50-69 56.1 60.9
Table 4. Crude and adjusted attendance rates (%) by 5-year age groups. Years
2011-2012.
Tabella 4. Adesione, grezza e aggiustata, per fasce d’età quinquennali. Anni
2011-2012.
anno 39 (3) maggio-giugno 2015
e
&
p
Breast cancer screening: 2011-2012 survey
26
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Region First exams Repeat exams
recall rate 10th-90th recall rate 10th-90th
(%) percentile (%) percentile
Valle d’Aosta 8.9 4.9
Piemonte 6.6 3.3-8.8 3.3 1.8-5.0
Liguria 10.3 5.3-15.0 6.0 1.5-8.7
Lombardia 8.9 6.5-13.7 4.6 3.3-7.2
Trento 8.3 3.2
Bolzano 7.8 3.6
Veneto 9.4 5.5-13.1 4.0 3.0-5.2
Friuli-Venezia Giulia 15.1 5.1
Emilia-Romagna 9.1 5.5-13.1 3.9 2.2-5.2
North 9.1 5.5-13.7 4.1 2.6-6.4
Toscana 12.7 8.3-19.5 6.1 3.6-11.8
Umbria 8.4 3.2
Marche 17.4 4.9-22.5 8.6 2.0-20.3
Lazio 7.6 5.8-15.7 5.1 2.6-12.3
Centre 9.9 6.3-22.4 5.8 2.8-12.6
Abruzzo 13.8 10.2-37.3 5.7
Molise 3.4 2.8
Campania 7.1 1.4-27.4 8.6 2.5-13.3
Puglia NA 5.6
Basilicata 0.0 7.1
Calabria 9.9 2.8-22.3 8.6 3.2-25.0
Sicilia 6.9 1.6-16.4 4.8 3.4-4.8
Sardegna 9.0 4.0-21.5 5.2 0.0-11.1
South 8.2 2.2-21.9 6.0 2.5-12.5
Italy 9.2 4.9-17.3 4.7 2.6-11.1
Table 5. Total crude recall
rates (%) by region, first and
repeat screening.Years 2011-
2012.
Tabella 5. Tasso di richiami
totale grezzo per Regione,
primi esami e successivi. Anni
2011-2012.
women excluded due to recent mammograms. Table 4 reports
attendance rates by 5-year age group. It is interesting to note
that the highest attendance is recorded among women over age
54, i.e., in women invited to screening for several years; con-
sequently they are more likely to participate, as they are aware
of the efficiency and quality of the diagnostic procedures
within an organized screening programme.
DIAGNOSTIC INDICATORS
Referral/recall rates
Referral/recall rate for further assessment is the main indica-
tor of first level screening specificity. It indicates the propor-
tion of screened women referred/recalled for diagnostic as-
sessments. This value needs to be reasonably low, in order to
limit negative psychological impact (anxiety), invasive proce-
dures (cytology, core- or surgical biopsies), as well as costs. Rec-
ommended GISMa standards are: <7% (acceptable) and <5%
(desirable) at first screening; <5% (acceptable) and <3% (de-
sirable) at repeat screening. Table 5 shows crude referral rate,
for first and repeat screening tests.
Considering first tests, rates beyond the maximum acceptable
standard for this indicator persisted in 2011-2012; moreover,
as already observed in previous surveys, an increasing trend was
evident: 7.5% in 2008, 8.0% in 2009, 8.8% in 2010, and
9.2% in 2011-2012. Excessively high rates were recorded
both nationally and (often) regionally: only three regions re-
ported a value within the acceptable standard of 7% and five
regions exceeded 10%.
A more detailed analysis shows that even at the individual, lo-
cal programme level, minimum standards were often ex-
ceeded: only 10% of local programmes were within the de-
sirable standard, while 10% of the programmes had
unacceptably high referral rates (>17%). Again, a large vari-
ability exists within each region.
Repeat tests show better results: the national indicator was still
within the acceptable standard and was rather stable in com-
parison with the previous year (4.7 % in 2011-2012 vs 4.6%
in 2009-2010). Even variability within each region (at least in
absolute numbers) seemed to become more limited, although
in some cases intra-regional variability was still very high.
It is worth noting that the recall rate at repeat screening
tends to be higher in the South as compared to Centre and
North, even though the detection rates (see below) go in the
opposite direction (higher in the North as compared to the
South): as a consequence, positive predictive values are much
lower in the South as compared to the North of Italy.
Detection rates
Table 6 reports the crude detection rates (DR) of carcinomas
(per 1,000 screened women), the crude detection rates of can-
cers≤10 mm, the benign to malignant surgery ratio (B/M),
and the proportion of ductal carcinomas in situ at first and at
Breast cancer screening: 2011-2012 survey
27
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Benign/malignant (B/M) surgical biopsy
ratio
The B/M ratio is determined on women referred to surgery;
it indicates the ratio of benign to malignant (B/M) pathology
outcomes. It is an optimal indicator of the diagnostic speci-
ficity of the programme assessment phase. It should be as low
as possible.
Recommended GISMa standards are: ≤1(acceptable) and
≤0.5 (desirable) at first screening; ≤0.5 (acceptable) and ≤0.25
(desirable) at repeat screening.
Results for this indicator are very satisfactory, well within the
desirable standards both at first and subsequent tests (0.2 and
0.1, respectively); the results are quite homogeneous through-
out Italy (see table 6).
Proportion of carcinomas in situ
It indicates the proportion of ductal carcinomas in situ every
100 total detected cancers, with histological diagnosis.
Recommended GISMa standards are 10% (acceptable) and
10%-20% (desirable) at any screening round.
Overall the results of 2011-2012 are in the desirable range
both at first and repeat tests (13.3% and 12.0%, respectively).
However, five regions reported values higher than 20% at first
subsequent tests. The first two indicators are the most com-
monly referred to indicators of a programme’s diagnostic sen-
sitivity (i.e., the capability of a programme to detect cancers
and to detect cancer at an early stage).These indicators should
be evaluated compared to expected incidence rate in the
screened population, in order to take into consideration the
variability of the baseline risk for breast cancer.
In the years 2011-2012, cancer DRs were 4.8/1,000 and
4.4/1,000 at first and repeat test, respectively, both slightly
lower than in 2010 (5.3 and 4.6 for first and repeat tests, re-
spectively).This decrease is mostly accounted for by the lower
DRs observed in the South of Italy as compared to 2010.
Table 7 (p. 28) reports DRs subdivided by 5-year age groups.
As expected, DRs tended to increase in older age either at first
or at repeat tests.
DRs of small invasive (≤10 mm) carcinomas were sub-
stantially stable as compared to 2010 (1.4/1,000 and
1.5/1,000 for first and repeat tests, respectively). At repeat
tests, DRs of carcinomas ≤10 mm in southern Italy were
very low. The possibility of incomplete registration of data
in the programmes from the South might be considered.
Again (see table 7) we observed an increase in DRs in
older age groups.
Region First exams Repeat exams
total B/M cancer ≤10 mm ductal total B/M cancer ≤10 mm ductal
detection rate ratio detection rate carcinoma in situ detection rate ratio detection rate carcinoma in situ
(x 1,000 (x 1,000 (% of all (x 1,000 (x 1,000 (% of all
screened) screened) malignancies) screened) screened) malignancies)
Valle d’Aosta 3.7 0.0 1.2 33.3 5.5 0.1 1.8 11.8
Piemonte 7.3 0.2 1.3 17.1 5.0 0.1 1.4 14.2
Liguria 3.0 0.2 1.0 7.8 3.6 0.1 1.6 9.6
Lombardia 4.3 0.2 1.4 12.8 4.1 0.1 1.4 10.0
Trento 5.7 0.2 1.3 10.0 4.9 0.1 1.6 16.9
Bolzano 5.6 0.0 1.1 25.0 3.8 0.0 1.2 8.5
Veneto 5.3 0.3 1.4 14.2 4.9 0.1 1.5 10.8
Friuli-Venezia Giulia 8.2 0.1 2.2 15.7 5.1 0.1 1.6 13.6
Emilia-Romagna 7.6 0.2 1.9 24.3 5.6 0.1 1.9 16.8
North 5.3 0.2 1.4 15.9 4.8 0.1 1.5 12.9
Toscana 5.4 0.3 1.6 14.8 5.0 0.1 1.8 13.1
Umbria 4.2 0.1 0.7 19.6 4.0 0.0 1.4 15.7
Marche 5.0 0.1 1.3 12.2 3.3 0.2 0.9 12.4
Lazio 3.3 0.1 0.6 3.9 3.5 0.1 0.9 3.0
Centre 4.2 0.2 1.0 10.3 4.3 0.1 1.4 11.0
Abruzzo 6.9 0.0 3.0 28.0 0.0 0.0
Molise 2.7 0.0 1.3 25.0 2.5 0.2 0.8 23.1
Campania 5.2 0.5 2.5 4.4 2.7 0.2 0.7 1.6
Puglia 3.0 0.1 0.2 6.9
Basilicata 7.2 0.3 2.1 0.0 3.7 0.2 1.0 0.0
Calabria 1.5 0.2 0.2 0.0 1.8 0.2 0.3 21.4
Sicilia 4.5 0.2 0.9 6.4 2.6 0.1 0.4 3.0
Sardegna 3.9 0.1 0.7 4.2 1.5 0.0 0.1 8.7
South 4.1 0.2 1.1 6.4 2.7 0.1 0.3 6.3
Italy 4.8 0.2 1.3 13.3 4.4 0.1 1.4 12.0
Table 6. Diagnostic indicators,first and repeat screening. Years 2011-2012.
Tabella 6. Indicatori diagnostici, primi esami e successivi. Anni 2011-2012.
anno 39 (3) maggio-giugno 2015
e
&
p
Breast cancer screening: 2011-2012 survey
28
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
spontaneous screening, which in some programmes may ac-
count for a substantial proportion of the target population.
Though considering limitations included in the data (as
previously indicated), overall the indicators recorded by Ital-
ian programmes in 2011-2012 appear rather good and com-
ply at a satisfactory level with recommended national stan-
dards. The only critical diagnostic indicator is the recall
rate, which has shown an increase over the past few years,
with a consequent problem of testing overload for health fa-
cilities and stress for patients undergoing further assess-
ments. To some extent, data on recall rates can suggest po-
tential critical aspects for specificity in many programmes.
They likely reflect an attitude that might be described as “de-
fensive medicine”, where diagnosticians tend to protect
themselves from potential litigation. The diffusion of digi-
tal mammography may also have played a role, and it is
worth noting that the increase in referral rate concerns in par-
ticular the first round tests, where previous mammograms are
not available. This high recall rate apparently does not pro-
duce a high rate of unnecessary surgery, as demonstrated by
the good performance of the B/M ratio.
A number of values exceeding acceptable standards may be ex-
plained by the scantiness of cases or by partial data registration.
In any case, further opportunities for discussing observed dif-
ficulties and systematic interventions for quality assurance of
the diagnostic procedures are required in the near future, es-
pecially in the South of Italy, where sensitivity indicators
(such as total detection rate and detection of cancers ≤10mm)
did not attain satisfactory values.
One of the main controversies in cancer screening is related to
overdiagnosis. Usually overdiagnosis refers to detection of in
situ lesions, part of which would never become clinically ap-
parent without screening. The good results recorded for the
percentage of in situ cancers during the years 2011-2012 sug-
gest that the problem of overdiagnosis is contained.
Conflicts of interests: none declared
screening and two at subsequent screening, but this data may
also reflect the relatively small number of cases involved. The
proportion of carcinomas in situ tends to be inversely corre-
lated to age (see table 7).
CONCLUSIONS
This paper presents the performance results of Italian organ-
ized mammographic screening programmes in the biennium
2011-2012. During that period, almost 75% of the national
target population was actively invited by organized screening
programmes, with a slight but constant increase compared to
previous years. Unfortunately, a strong imbalance in mam-
mography screening offer still persists in Italy between the
North-Centre and the South of the country: while almost 90%
of the target population was invited in the northern and cen-
tral areas, only 40% of 50-69 year-old women resident in the
South were invited. In the biennium 2011-2012, there were
only small improvements in this critical issue.
The mean national value of attendance is satisfactory (at least
comparing mammographic screening to cervical and colorec-
tal cancer screening), although a decreasing trend from North
to South is clearly evident. The combination of these two pa-
rameters paints an alarming picture for the South of Italy, for
which we recorded only a small number of implemented pro-
grammes and low participation. A low number of invitations
and low participation rates result in too few women screened
and poor performance, as reflected by the indicator adopted
by the Health Ministry to judge screening programme per-
formance. To some extent, these data may reflect a different
attitude towards prevention in the North and South, as shown
by other national studies, such as Istat’s Multiscopo and the
PASSI survey.
In the biennium 2011-2012, our results confirm that screen-
ing coverage of the target population in Italy was steady, at a
low 38.9%, suggesting the need for further investments and ef-
forts. However, this rate is likely to underestimate the real sit-
uation, as our survey does not include women undergoing
Age First exams Repeat exams
recall total B/M cancers ductal recall total B/M cancers ductal
rate detection ratio ≤10 mm carcinoma rate detection ratio ≤10 mm carcinoma
(%) rate detection in situ (%) rate detection in situ
(x 1,000 rate (% of all (x 1,000 rate (% of all
screened) malignancies) screened) malignancies)
50-54 9.6 4.1 0.3 1.0 15.6 5.6 3.0 0.2 0.8 16.0
55-59 8.6 4.8 0.2 1.2 14.6 4.6 3.6 0.1 1.1 13.4
60-64 8.3 6.2 0.1 1.9 7.6 4.3 4.9 0.1 1.5 11.2
65-69 8.4 8.0 0.1 2.2 10.3 4.3 6.0 0.1 2.0 10.5
Total 50-69 9.2 4.8 0.2 1.3 13.3 4.7 4.4 0.1 1.4 12.0
Table 7. Diagnostic indicators by age group, first and repeat screening.Years 2011-2012.
Tabella 7. Indicatori diagnostici per classe d’età, primi esami e successivi. Anni 2011-2012.
Breast cancer screening: 2011-2012 survey
29
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
References/Bibliografia
1. Giordano L, Giorgi D, Frigerio A e il gruppo GISMa. Indicatori e stan-
dard per la valutazione di processo dei programmi di screening del
cancro della mammella. Epidemiol Prev 2006;2(Suppl.1):1-48.
2. Screening oncologici. Raccomandazioni per la pianificazione e
l’esecuzione degli screening di popolazione per la prevenzione del
cancro della mammella, del cancro della cervice uterina e del can-
cro del colon retto. Ministero della salute, Direzione generale della
prevenzione. Roma 2006.
3. Frigerio A et al. Diffusione dello screening mammografico in Italia. In:
Rosselli Del Turco M, Zappa M (eds). Osservatorio nazionale per la pre-
venzione dei tumori femminili. Primo rapporto. Roma 2002:14-21.
4. Giorgi D et al. Lo screening mammografico in Italia: dati GISMa
2001. In: Rosselli Del Turco M, Zappa M (eds). Osservatorio na-
zionale per la prevenzione dei tumori femminili. Secondo rapporto.
Roma 2003:14-23.
5. Giorgi D et al. Lo screening mammografico in Italia: dati GISMa
2002-2003. In: Rosselli Del Turco M, Zappa M (eds). Osservatorio
nazionale per la prevenzione dei tumori femminili. Terzo rapporto.
Roma 2004:16-25.
6. Giorgi D et al. Lo screening mammografico in Italia: survey 2003-
2004. In: Rosselli Del Turco M, Zappa M (eds). Osservatorio na-
zionale screening. Quarto rapporto. Roma 2005:16-27.
7. Giorgi D et al. Lo screening mammografico in Italia: survey 2004
e dati preliminari 2005. In: Rosselli Del Turco M, Zappa M (eds).
Osservatorio nazionale screening. Quinto rapporto. Inferenze, Mi-
lano 2006:16-31.
8. Giorgi D et al. Lo screening mammografico in Italia: survey 2005
e dati preliminari 2006. In: Rosselli Del Turco M, Zappa M (eds).
Osservatorio nazionale screening. Sesto rapporto. Inferenze, Mi-
lano 2007:20-35.
9. Giorgi D et al. Lo screening mammografico in Italia: survey 2007.
In: Zappa M (ed). Osservatorio nazionale screening. Settimo rap-
porto. Inferenze, Milano 2008:16-31.
10. Giorgi D et al. Lo screening mammografico in Italia: survey 2008.
In: Zappa M (ed). Osservatorio nazionale screening. Ottavo rap-
porto 2009. Tumori, Roma 2010:22-38.
11. Giorgi D, Giordano L, Ventura L et al. Lo screening mammografico
in Italia: survey 2009. Epidemiol Prev 2011;5-6 (Suppl.5):9-27.
12. Giorgi D, Giordano L, Ventura L et al. Lo screening mammografico
in Italia: survey 2010. Epidemiol Prev 2012;6 (Suppl.1):8-27.
Marche: A. Vaccaro, D. Cimini, B. Finocchi, M.G. Volpini,
C. Ragaglia, A. Santone
Lazio: P. Bellardini, G. Baldi, L. Martufi, D. Baiocchi, R. Fulgenzi,
F. Odoardi, S. Marzani, E. Rossi, G. Nicodemo, F. Puddu,
M.G. Acampora, S. Sgricia, M. Arcara, M. Mammola, S. Brezzi,
D. Boninsegna
Abruzzo: M. Muzi, M. Brucchi
Molise: F. Carrozza
Campania: C. Casullo, A. Frieri, M.P. Aversano, G. Capone,
R. Papa, E. Barretta, A. Esposito, C. Maione, U. Scala
Puglia: C. Germinario
Basilicata: V. Barile
Calabria: E. Bova, A. Giorno, T. Landro, A. Bisbano, M.P. Montesi,
M. Viola, M.B. Grasso
Sicilia: S. Malignaggi, R. Candura, P.M. Santino, R. Scillieri, L. Costa,
G. La Perna
Sardegna: S. Tilocca, U. Stochino, A. Onnis, M.A. Atzori, M. Piga,
O. Frongia, F.S. Congiu, L. Paoni
Valle D’Aosta: L. Giordano
Piemonte: L. Giordano, L. Orione
Lombardia: F. Sambo, S. G. Domenighini, L. Cecconami,
G. Magenes, M. Ignone, R. Lucchini, D. Cereda, E. Anghinoni,
G. Marazza, A. Ilardo, M. Dal Soldà, S. Gotti, L. Tessandri,
M. Crisetig, R. Cecchetti, A. Silvestri, M. Montanelli, G. Gola
Alto Adige: A. Fanolla
Trentino: M. Pellegrini, D. Bernardi, M. Gentilini
Veneto: C. Fedato
Friuli-Venezia Giulia: A. Franzo
Liguria:A. Franco, B. Scanu, L. Garibotto, I. Valle, M. Orlando,
F. Maddalo
Emilia-Romagna: P. Baldazzi, C. Imolesi, D. Canuti, G. Benea,
F. Falcini, L. Caprara, R. Negri, M. Zatelli, G. Gatti, M. Serafini,
B. Vitali, A. Cattani, G. Monticelli, C. Debora
Toscana: E. Paci, P. Mantellini, F. D’Elia, F. Mirri, R, Capecchi,
P. Piccini, R. Rosati, C. Maffei, D. Giorgi, P. Vivani, L. Del Chicca,
M. Rapanà, L. Adbelghani, R. Turillazzi, A.A. Scarfantoni
Umbria: S. Prandini
Data for the ONS/GISMa surveys for the year 2011-2012 was provided by:
Hanno fornito i dati per la survey ONS/GISMa 2011-2012:
anno 39 (3) maggio-giugno 2015
e
&
p
30
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
WWW.EPIPREV.IT
Breast cancer screening in Italy:
evaluating key performance indicators
for time trends and activity volumes
Lo screening mammografico in Italia:
valutazione degli indicatori di performance
per trend temporali e volumi di attività
Livia Giordano,1Roberta Castagno,1Daniela Giorgi,2Cristiano Piccinelli,1Leonardo Ventura,3
Nereo Segnan,1Marco Zappa3
Abstract
Together with the
National centre for screening monitoring (ONS)
, GISMa supports annual collection of
data on national breast screening activities. Aggregated data on implementation and performance are
gathered through a standardized form to calculate process and impact indicators. Analyzed data be-
long to 153 local programmes in the period 2006-2011 (2006-2012 for participation rate only).
During the whole period, Italian crude participation rate exceeded GISMa’s acceptable standard (50%),
even though a higher participation in northern and central Italy compared to southern Italy and Islands
was observed. Time trend analysis of diagnostic indicators confirmed in 2011 an adequate quality of
breast screening performance, especially at subsequent screening. Recall rate at initial screening did
not reach the acceptable standard (<7%) and rose slightly over the period. On the contrary, a good
performance was achieved at subsequent screening. The same trend was followed by the overall de-
tection rate and positive predictive value. They both showed a progressive reduction (from 6.2‰ in
2006 to 4.5‰ in 2011 for DR and from 8.0% in 2006 to 5.2% in 2011 for PPV, respectively) at ini-
tial screening and a good, stable trend at subsequent screening.
Activity volume analysis shows that in programmes with greater activity (test/year ≥10,000) RR at both
initial and subsequent screening has a better performance. This is also true for DR and PPV where pro-
grammes with high volumes of activity do better, especially when compared with those that interpret
fewer than 5,000 mammograms per year.
In spite of a few limits, these results are reassuring, and they reward the efforts made by screening
professionals. It is therefore important to continue to monitor screening indicators and suggest, test,
and evaluate new strategies for continuous improvement.
Epidemiol Prev 2015; 39(3) Suppl 1: 30-39)
Keywords: breast cancer screening, time trends, activity volumes, process indicators, Italy
1Unità di epidemiologia
dei tumori, CPO Piemonte,
AOU Città della salute
e della scienza, Torino
2SC Epidemiologia
e screening, ASL 2, Lucca
3Istituto per lo studio
e la prevenzione
oncologica (ISPO), Firenze
Corresponding author
Livia Giordano
livia.giordano@cpo.it
Riassunto
Il GISMa (Gruppo italiano screening mammografico) insieme con l‘Osservatorio nazionale screening
(OMNS) promuove ogni anno la raccolta sistematica dei dati sull’attività dei programmi organizzati di
screening mammografico in Italia. I dati aggregati relativi all’implementazione e alla performance dei
programmi vengono raccolti e registrati su un apposito questionario standard e utilizzati per calcolare
indicatori di processo e precoci di impatto. I dati analizzati si riferiscono a 153 programmi locali attivi
nel periodo 2006-2011 (2006-2012 solo per la parte relativa alla partecipazione).
L’indagine mostra che il tasso di partecipazione grezza raggiunge e mantiene nel tempo lo standard
Breast cancer screening: time trends and activity volumes
INTRODUCTION
To obtain projected benefits and minimize negative outcomes,
breast cancer screening programmes should be implemented
with an organized, population-based approach, with quality as-
surance at all appropriate levels, and in accordance with Euro-
pean guidelines for quality assurance in breast cancer screening and
diagnosis.1According to the IARC Handbook of cancer preven-
tion2an organized screening programme requires the following
six characteristics: a policy specifying target population, screen-
ing methods and interval; a defined target population; a team
responsible for overseeing screening centres; a clear decision
structure and responsibility for healthcare management; a qual-
ity assurance system utilizing relevant data; and monitoring of
cancer occurrence in the target population.
The highest level of programme organization of population-
based screening requires that all persons eligible for screening
be identified and personally invited to attend a screening ex-
amination in each round of screening3and followed for the en-
tire screening pathway.
Since its establishment in 1990, the Italian group for mam-
mography screening (GISMa) has represented a cornerstone in
monitoring and performance evaluation of organized breast
screening programmes in Italy. Together with the National
centre for screening monitoring (ONS), created in 2002 by the
Italian Ministry of Health with the aim to monitor and promote
screening programmes nationwide, GISMa supports the annual
collection of data on national breast screening activities. Ag-
gregated data on implementation and performance are gathered
through a standardized form to calculate process and impact in-
dicators which have been agreed on a national level.4Results are
also compared with European standards.1
Despite some initial difficulties, annual surveys have improved
over the years, thanks to the collaborative efforts of all screen-
ing professionals, who work together to reduce and overcom-
ing heterogeneity in screening implementation, organization,
and management among Italian areas, trying to ensure higher
levels of standardization and data completeness.
The main aim of this work is to assess the time trend for selected
process and impact indicators – participation rate, recall rate,
overall detection rate and positive predictive value – in the pe-
riod 2006-2011 (2006-2012 for participation only).
The same parameters are also analyzed and cross-checked by
programme activity volumes.
This paper is an update of a previous report, published in the
2012 edition of the annual ONS Report.5
METHODS
In Italy there is no national breast cancer screening programme,
but rather a number of regionally-coordinated local initia-
tives. All 20 regions work under the umbrella of ONS, which
is responsible, with the GISMa group, for data collection and
monitoring. Data are collected annually by means of a struc-
tured questionnaire, in a computerized form, which allows in-
dicators to be calculated with automatic formulas. The ques-
tionnaire refers to the previous year’s activity and is stratified
by age group. It is sent out yearly by the ONS to the referent
for data collection in every region. The regional referent then
delivers the questionnaire to the referents of every programme
in the region.
The filled-in questionnaires are returned from the local pro-
grammes to the Regional Centre and, subsequently, if approved
by regional referents, to the National Centre. Logical and epi-
demiological checks are performed either at the regional or at
the national level. In particular, if data are logically impossible
or epidemiologically improbable (in comparison to historical
trends, to the performances of other programmes in the area,
etc.), a specific check on that information is carried out.
Questionnaires from 168 organized programmes (running for
the entire 2006-2012 period or only a part of it) were collected.
After a further check for completeness and consistency, 15
programmes with <100 tests per year and those providing in-
complete/inconsistent information were excluded. A total of
153 questionnaires were analyzed: 68 for the North (44.4%),
49 for the Centre (32.0%), and 36 for the South (23.5%).
accettabile GISMa del 50%, anche se si osservano livelli più alti di partecipazione al Nord e al Centro Italia rispetto al Sud/Isole.
L‘analisi temporale degli indicatori considerati (tasso totale di identificazione dei tumori, tasso di richiami in secondo livello e va-
lore predittivo positivo) mostra una buona qualità. Il tasso di richiami si mantiene adeguato nel tempo soprattutto nei passaggi
successivi (anche se sta avvicinandosi sempre di più alla soglia minima raccomandata) mentre, per i primi esami, non raggiunge
lo standard accettabile (<7%).
Buoni andamenti si osservano anche per il tasso totale di identificazione dei tumori e dal valore predittivo positivo. Entrambi mo-
strano una riduzione progressiva nel tempo ai primi esami (passando dal 6.2‰ nel 2006 al 4.5‰ nel 2011 e dall’8.0% nel 2006
al 5.2% nel 2011, rispettivamente) e un andamento buono e stabile agli esami successivi.
L‘analisi per volumi di attività indica che programmi con volumi più ampi (>10.000 test/anno) presentano indicatori migliori ri-
spetto a programmi in cui l'attività è più bassa.
Nonostante alcuni limiti dell’analisi, i risultati raggiunti sono rassicuranti e ricompensano gli sforzi intrapresi da tutti gli operatori
dello screening in questi anni. Resta comunque importante continuare il monitoraggio degli indicatori dello screening mammografico
e valutare nuove strategie per un continuo miglioramento delle prestazioni dei programmi organizzati di screening in Italia.
Epidemiol Prev 2015; 39(3) Suppl 1: 30-39)
Parole chiave: screening mammografico, trend temporali, volumi di attività, indicatori di processo, Italia
31
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Table 1 illustrates the number of tests, recalled women, and
screen-detected malignant cancers by the three Italian macro-
areas and time period. Analysis was performed for the follow-
ing indicators:
■
Participation rate, PR (%):
■
overall crude PR: the number of women who have a
screening test as a proportion of all women who are invited to
attend for screening;
■
adjusted PR: the number of women who have a screening
test as a proportion of all women who are invited to attend for
screening, excluding from the denominator women with a re-
cent (<12 months) mammogram outside the programme;
■
Recall rate, RR (%): the number of women recalled for fur-
ther assessments as a proportion of all women who had a screen-
ing examination;
■
Detection rate, DR (‰): the number of all malignant can-
cers detected every 1,000 screened women;
■
Positive predictive value, PPV (%): the ratio of lesions that
are truly positive to those that test positive.
These parameters were examined and cross-checked by time
trends for Italy and for the standard target population (50-69)
as a whole, by 5 year age-classes (50-54; 55-59; 60-64; 65-69)
and by geographical macro-areas (North, Centre, South-Is-
lands). For RR, DR, and PPV only, data were also disaggre-
gated by screening step: initial screening, referring to women
undergoing screening for the first time, and subsequent screen-
ing, referring to women who previously underwent screening
tests (for programmes implemented during the last two years
this category is not yet available).
These last indicators were also associated with the volume of
activity of the programmes, calculated as the number of tests
(both at initial and subsequent rounds) performed by the pro-
grammes yearly. Four classes of volume were considered:
<5,000; 5,000-9,999; 10,000-14,999, ≥15,000.
Breast cancer screening: time trends and activity volumes
32
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
2006 2007 2008 2009 2010 2011
North number of performed tests initial screening 174,640 175,280 176,375 161,885 164,838 156,173
subsequent screening 546,044 608,385 624,087 649,449 712,159 765,994
number of women recalled initial screening 13,719 13,628 13,662 12,598 14,209 13,954
for further assessments subsequent screening 21,648 24,423 25,558 25,799 29,263 31,524
number of screen-detected initial screening 1,262 1,072 967 801 879 809
malignant cancers subsequent screening 2,601 2,772 2,900 3,025 3,236 3,542
Centre number of performed tests initial screening 68,903 50,575 61,151 53,425 52,043 78,972
subsequent screening 189,298 191,649 228,545 210,381 227,910 232,433
number of women recalled initial screening 4,796 3,831 4,944 4,962 4,862 6,420
for further assessments subsequent screening 10,502 9,977 11,109 12,610 11,686 12,648
number of screen-detected initial screening 295 330 262 240 201 250
malignant cancers subsequent screening 820 937 878 877 950 1,003
South/ number of performed tests initial screening 32,982 53,105 74,144 86,669 23,271 25,171
Islands subsequent screening 46,326 76,323 44,304 28,789 128,056 128,943
number of women recalled initial screening 2,638 4,392 5,170 6,265 1,720 1,970
for further assessments subsequent screening 1,602 1,946 3,433 2,286 6,544 6,581
number of screen-detected initial screening 145 292 214 276 105 113
malignant cancers subsequent screening 71 74 108 105 402 417
Table 1. Number of performed tests,recalled women and screen-detected malignant cancers by Italian macro-areas.Years 2006-2011.
Tabella 1. Numero di test eseguiti, di donne richiamate per approfondimenti e di tumori maligni rivelati allo screening per macroaree.Anni 2006-2011.
RESULTS
Time trends analysis
Participation rate (PR)
For cancer screening programmes to bring about reductions in
mortality, a substantial proportion of the population must par-
ticipate. Programmes with low uptake can be ineffective and can
promote inequalities in health service. For these reasons, PR is
a key parameter to assess both the impact of the screening pro-
gramme and its acceptability among the target population.
However, evaluation and interpretation of results may be affected
by contextual aspects (e.g., opportunistic screening activities,
level of breast cancer awareness, socio-demographic characteris-
tics of the target population) and other organizational factors
(e.g., availability and accessibility of the services for diagnosis and
treatment, invitation system and communication strategies used
by the programme to increase informed participation). European
guidelines consider 50% an acceptable level of PR and indicate
70% as a desirable standard. In the considered period, the over-
all Italian crude PR always exceeded the minimum benchmark
(figure 1) although it never reached the optimal one.
Nevertheless, attendance rates by geographical macro-areas
confirmed, in 2012, a higher participation in northern and
central Italy compared to the South-Islands, where rates were
still inadequate and did not reach the recommended mini-
mum. Figure 2 shows the adjusted participation rate by 5-year
age classes during the same 2006-2012 period. For the whole
period, women of the intermediate classes had higher atten-
dance rates compared to younger and older women and by far
the highest participation was recorded for women who belong
to the 60-64 age group.
Recall rate (RR), detection rate (DR), positive predictive
value (PPV)
Although randomized controlled trials have shown that screen-
Breast cancer screening: time trends and activity volumes
33
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
2006
60.5 61.5
34.3
2007 2008 2009 2010 2011 2012
%
60
70
80
50
40
30
20
10
0
Figure 1. Overall crude parti-
cipation rate inItaly and byma-
cro-areas. Years 2006-2012.
Figura 1. Partecipazione com-
plessiva grezza in Italia e per
macroaree. Anni 2006-2012.
56.0
38.3
60.8
56.8
34.0
61.0 60.1 61.4 62.9
54.0
53.5
55.3
41.6 38.9
53.7
39.2
56.2
Italy
North
Centre
South/Islands
43.5
2006
59.6 60.6
2007 2008 2009 2010 2011 2012
%
60
70
80
50
40
30
20
10
0
Figure 2. Adjusted participa-
tion rate by 5-year age classes.
Years 2006-2012.
Figura 2. Partecipazione cor-
retta, per fasce d’età quin-
quennali.Anni 2006-2012.
65.7 65.7
61.4
58.5
64.2 64.8
60.2
55.9
61.5
57.6 57.8
60.3
63.2 64.8
64.2
63.364.5
61.0 62.1 61.4
63.160.8
64.9 64.7
50-54
55-59
60-64
65-69
62.4
58.4
Italy
ing mammography reduces the mortality for breast cancer, the
efficacy of mammography depends on the performance of the
interpreting radiologist, technical quality of the mammo-
grams, and proper implementation of a screening programme.
The purpose of mammography is detection of cancer (high
sensitivity), but this goal is ideally accomplished with reason-
able recall and biopsy rates (high specificity).
Good RR, DR, and PPV levels indicate that the programmes
are working in the right direction of getting a positive impact
on breast cancer mortality.
Recall rate
Recall rate represents a good indicator of screening specificity
(first level). In Italy in the whole period the percentage of
screened women referred for further assessments at initial
screening did not reach either the desirable (<5%) nor the ac-
ceptable standard (<7%), and the rate rose slightly over the
years. On the contrary, a good performance for this indicator
was achieved at subsequent screening, where the standard is
<5% and <3% for the acceptable and desirable level, respec-
tively. In subsequent screening tests, RR maintained a constant
performance throughout the period (average value: 4.4%), al-
though moving toward the warning threshold (figure 3, p. 34).
At initial screening, RR trend analysis by North, Centre, and
South-Islands presents the same increasing trends within the
three areas, while comparison between them does not reveal
substantial differences, with the exception of central Italy,
which had higher RRs in certain years.
Figure 3. Time trends of
recall rate (%) for women
50-69 years. Years 2006-
2011.
Figura 3.Andamento tem-
porale dei richiami per ap-
profondimento, età 50-69
anni.Anni 2006-2011.
Breast cancer screening: time trends and activity volumes
34
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
subsequent screening
initial screening
0 2 4 6 8 1010 8 6 4 2
7.6 2006 4.3
7.8 2007 4.1
7.6 2008 4.5
7.9 2009 4.5
8.7 2010 4.4
8.6 2011 4.4
At subsequent screening, RR trends appeared to be very sta-
ble in the North, less stable in the Centre, and in the South-
Islands where a high variation among periods was present
(table 2).
Analysis by 5-year age classes shows a fairly stable indicator
within each age group over time, both at first and subsequent
screening. Younger women have higher RRs whether they un-
dergo mammography for the first time or not (table 3).
2006 2007 2008 2009 2010 2011
RECALL RATE (%)
initial screening
North 7.9 7.8 7.7 7.8 8.6 8.9
Centre 7.0 7.6 8.1 9.3 9.3 8.1
South-Islands 8.0 8.3 7.0 7.2 7.4 7.8
Italy 7.6 7.8 7.6 7.9 8.7 8.6
subsequent screening
North 4.0 4.0 4.1 4.0 4.4 4.1
Centre 5.5 5.2 4.9 6.0 5.1 5.4
South-Islands 3.5 2.5 7.7 7.9 5.1 5.1
Italy 4.3 4.1 4.5 4.6 4.4 4.5
DETECTION RATE FOR MALIGNANT CANCERS (‰)
initial screening
North 7.2 6.1 5.5 4.9 5.3 5.2
Centre 4.3 6.5 4.3 4.5 3.9 3.2
South-Islands 4.4 5.5 2.9 3.2 4.5 4.5
Italy 6.2 6.1 4.6 4.4 4.9 4.5
subsequent screening
North 4.8 4.6 4.6 4.7 4.5 4.6
Centre 4.3 4.9 3.8 4.2 4.2 4.3
South-Islands 1.5 1.0 2.4 3.6 3.1 3.2
Italy 4.5 4.3 4.3 4.5 4.3 4.4
POSITIVE PREDICTIVE VALUE (%)
initial screening
North 9.2 7.9 7.1 6.4 6.2 5.8
Centre 6.2 8.6 5.3 4.8 4.1 3.9
South-Islands 5.5 6.6 4.1 4.4 6.1 5.7
Italy 8.0 7.8 6.1 5.5 5.7 5.2
subsequent screening
North 12.0 11.3 11.3 11.7 11.1 11.2
Centre 7.8 9.4 7.9 7,0 8.1 7.9
South-Islands 4.4 3.8 3.1 4.6 6.1 6.3
Italy 10.3 10.4 9.7 9.8 9.7 9.8
Table 2. Recall rate, detection
rate and positive predictive
value by North, Centre and
South-Islands. Years 2006-
2011.
Tabella 2. Tasso di richiamo,
tasso di identificazione e va-
lore predittivo positivo, per
macroaree. Anni 2006-2011.
Breast cancer screening: time trends and activity volumes
35
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
2006 2007 2008 2009 2010 2011
RECALL RATE (%)
initial screening
50-54 8.7 8.7 8.6 8.7 9.1 9.1
55-59 6.8 7.2 7.0 7.1 7.9 8.4
60-64 6.7 6.8 6.8 6.4 7.7 7.6
65-69 7.0 7.0 6.0 7.2 8.0 6.9
Italy 50-69 7.6 7.8 7.6 7.9 8.7 8.6
subsequent screening
50-54 5.3 5.2 5.4 5.8 5.6 5.4
55-59 4.4 4.2 4.4 4.6 4.4 4.5
60-64 4.1 3.8 4.3 4.2 4.1 4.2
65-69 3.8 3.7 4.1 4.3 4.1 4.1
Italy 50-69 4.3 4.1 4.5 4.6 4.4 4.5
DETECTION RATE FOR MALIGNANT CANCERS (‰)
initial screening
50-54 4.4 4.6 3.9 3.6 4.3 4.0
55-59 5.6 6.1 4.1 3.9 4.7 4.8
60-64 7.5 7.3 6.0 5.2 6.8 5.9
65-69 10.0 9.3 6.3 7.3 8.2 5.9
Italy 50-69 6.2 6.1 4.6 4.4 4.9 4.5
subsequent screening
N50-54 2.9 2.7 2.7 3.0 2.8 3.0
55-59 3.8 3.6 3.4 3.6 3.3 3.6
60-64 5.0 4.9 4.8 4.7 4.8 4.8
65-69 5.7 5.6 5.8 6.1 5.7 5.8
Italy 50-69 4.5 4.3 4.3 4.5 4.3 4.4
POSITIVE PREDICTIVE VALUE (%)
initial screening
50-54 5.0 5.2 4.5 4.2 4.7 4.4
55-59 8.3 8.5 5.8 5.5 5.9 5.7
60-64 11.3 10.6 8.8 8.0 8.9 7.8
65-69 14.3 13.2 10.5 10.2 10.3 8.6
Italy 50-69 8.0 7.8 6.1 5.5 5.7 5.2
subsequent screening
50-54 5.6 5.2 5.1 5.2 5.0 5.5
55-59 8.6 8.5 7.7 7.8 7.6 8.1
60-64 12.1 12.9 11.0 11.2 11.7 11.3
65-69 14.9 14.8 13.9 14.3 14.0 14.0
Italy 50-69 10.3 10.4 9.7 9.8 9.7 9.8
Table 3. Recall rate, detec-
tion rate and positive predic-
tive value by 5-year age-
classes. Years 2006-2011.
Tabella 3. Tasso di richiamo,
tasso di identificazione e va-
lore predittivo positivo, per fa-
sce d’età quinquennali. Anni
2006-2011.
Overall detection rate
It is one of the main indicators of the diagnostic sensitivity of the
programme. It should be referred to the expected cancer inci-
dence rate in the screening population in order to take into ac-
count the baseline risk for breast cancer. Detection of invasive
breast cancers is disaggregated into first and subsequent screen-
ing rounds because a woman is more likely to have a breast can-
cer detected the first time she visits the breast screening service
than in subsequent visits.This is because a woman’s first visit de-
tects prevalent cancers that may have been present for some time
rather than incident cancers that have grown between screens.
Concerning initial screening, despite a small increase in 2010
compared to 2009, the DR shows a progressive reduction over
time (from 6.2‰ in 2006 to 4.5‰ in 2011). This might be
associated with the percentage of women referred to in-depth
diagnosis at initial screening, which is higher than expected.
The trend is quite good and stable at subsequent screening (av-
erage 4.4‰) (figure 4, p. 36).
Higher detection rates were found in northern Italy at initial
screening in 2006 and 2007 (7.2‰ and 6.1‰, respectively),
with a continuous reduction till 2011, while for central and
southern Italy DRs were lower but more stable (table 2). At
subsequent screening, DR values were lower in the South/Is-
lands in 2006-2007 (1.5‰ and 1.0‰, respectively), with a
constant increase in the following years till 2011, when the
value doubled (3.2‰ in 2011 vs 1.5‰ in 2006).
Analysis by 5-year age classes shows higher detection rates for
65-69 year-old women (both at initial and subsequent screen-
ing) and lower DRs in women aged 50-59 years. Within each
age group, DR had no substantial change over time (table 3).
Positive predictive value
Recall rate and detection rate are brought together by the pos-
itive predictive value (defined as the number of cancers detected
as a percentage of all women recalled for further assessments).
PPV is used as a central indicator of the quality of screening
mammography programmes. A better performance of screen-
ing programmes is achieved when low rates of women re-
called for further assessments are associated with high rates of
screen-detected cancers and positive predictive value. In a pro-
gramme with a low PPV and high RR, compared with one
with the same cancer DR but high PPV and low RR, the work-
load on the screening staff and the anxiety experienced by
women will be considerably greater.6
In the period under study, Italian programmes presented good,
stable PPV at subsequent screening, while a progressive re-
duction in PPV at initial screening (from 8.0% in 2006 to
5.2% in 2011) was observed (figure 5).
In the analysis by macro-areas, PPV rates at first screening de-
creased over time in all areas, with the exception of the South-
Islands where there was a slight increase in the last period. PPV
in the latter area was generally lower compared to northern and
central Italy. The trend for PPV at subsequent screening was
quite stable in northern and central Italy compared to south-
ern Italy, where the trend was more unstable and the values
were significantly lower (table 2).
Analysis by age classes shows higher PPV rates for women aged
60-69 both at initial and subsequent screening compared to the
other groups (table 3). All these parameters were stable over time.
Activity volumes analysis
Current European guidelines recommend that radiologists who
report screening mammograms should read at least 5,000 cases
per year. Data gathered through the questionnaire were also an-
Breast cancer screening: time trends and activity volumes
36
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Figure 4. Time trends of
overall detection rate (‰)
for women 50-69 years.
Years 2006-2011.
Figura 4.Andamento tem-
porale del tasso di identifi-
cazione (‰), età 50-69
anni.Anni 2006-2011.
subsequent screening
initial screening
0 2 4 6 88 6 4 2
6.2 2006 4.5
6.1 2007 4.3
4.6 2008 4.3
4.4 2009 4.5
4.9 2010 4.3
-4.5 2011 4.4
Figure 5. Time trends of
positive predictive value (%)
for women 50-69 years.
Years 2006-2011.
Figura 5.Andamento tem-
porale del valore predittivo
positivo (%), età 50-69
anni.Anni 2006-2011.
subsequent screening
initial screening
002040608 10 1212 10 8 6 4 2
8.0 2006 10.3
7.8 2007 10.4
6.1 2008 9.7
5.5 2009 9.8
5.7 2010 9.7
5.2 2011 9.8
alyzed to compare the trend of RR, DR, and PPV according to
the annual activity volume of each programme. Thus, four ac-
tivity volume classes were defined, with a number of tests rang-
ing from <5,000/year to >15,000/year. This preliminary analy-
sis gives rise to some considerations about the impact of activity
volume on performance indicators (figures 6-8).
In programmes with greater activity (test/year ≥10,000) the RR
at both initial and subsequent screening was lower and, only
at repeat screening, within acceptable standards (4.3%, 4.0%).
This was also true for DR and PPV, for which programmes
with high volumes of activity show better performance, espe-
cially when compared with those who read fewer than 5,000
mammograms per year; the latter had a critical level for all an-
alyzed indicators, both at initial and subsequent screening.
Breast cancer screening: time trends and activity volumes
37
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
<5,000 5,000-9,999 10,000-14,999 ≥15,000
2
4
6
8
10
12
14
16
0
Figure 6. Recall rate for ac-
tivity volumes of screening
programmes.
Figura 6. Tasso di richiami
per volumi di attività dei pro-
grammi di screening.
9.4
7.6 7.7
4.3
initial screening
subsequent screening
8.5
5.0
7.8
4.0
<5,000 5,000-9,999 10,000-14,999 ≥15,000
2
4
6
8
10
12
14
16
0
Figure 7. Detection rate for
activity volumes of screening
programmes.
Figura 7. Tasso di identifica-
zione per volumi di attività dei
programmi di screening.
3.8 3.2
5.9
4.7
initial screening
subsequent screening
5.4
4.5
5.4
4.5
<5,000 5,000-9,999 10,000-14,999 ≥15,000
2
4
6
8
10
12
14
16
0
Figure 8. Positive predictive
value for activity volumes of
screening programmes.
Figura 8. Valore predittivo
positivo per volumi di attività
dei programmi di screening.
3.9 4.1
7.4
10.8
initial screening
subsequent screening
6.1
8.8
6.7
11.1
%
%
‰
CONCLUSIONS
GISMa surveys have progressively changed and have become
increasingly complete and systematic. Thanks to the work of
several operators, data collection makes it possible to evaluate
the quality of programmes, produce local and national statis-
tics, and compare different screening areas through standard-
ized indicators. These investigations and comparisons are im-
portant in helping screening staff to properly manage their
activity and improve programme effectiveness and quality.
However, GISMa surveys still have some limitations: data col-
lected are aggregated, and not all programmes, particularly
those covering large areas and with several territorial screening
units, are able to provide a complete data set every year.
In general, analysis of the four parameters discussed above (PR,
DR, RR, and PPV), though with due caution, shows a good
average quality of screening performance, which was main-
tained over time. Conversely, a number of failures in screen-
ing offer or functioning, rather than in the diagnostic process,
need to be highlighted.
The discrepancy between northern and southern Italy per-
sisted. The absence of an organized screening activity, as well
the chronic lack of dedicated professionals, invested resources,
and clear-cut, well-planned political actions for prevention in
southern Italy affect the overall quality of the programmes.
More in-depth investigations are needed to evaluate this dis-
crepancy in order to suggest and discuss corrective strategies.
Participation rate is a key indicator for measuring and com-
paring the quality of screening, essential for stakeholders to
evaluate the effectiveness of their choices. Low levels of atten-
dance can make the organizational and economic efforts that
go into screening ineffective.
In Italy, despite a good, constant time trend in activity, which
reaches and exceeds the acceptable standard, a great variabil-
ity still persists among central-northern and southern/Islands
programmes and within individual regions.
For a better understanding of this trend, the portion of women
undergoing spontaneous screening (quite relevant in some
settings in southern Italy) should be assessed.
The presence of a massive opportunistic screening activity can
explain both the difficulty for the programmes to invite all the
target population and the wide heterogeneity in participation
rates between and within Italian regions.
Furthermore, besides the presence of an opportunistic screen-
ing activity, participation rate can be influenced by many
other factors, such as individual and socio-cultural conditions,
and organizational aspects of the screening invitation design.
A centralized organization, as present in many northern Ital-
ian regions, can stimulate useful synergies among the various
screening phases, resulting in a wider and more successful in-
volvement of the target population. Resources and efforts
should move in this direction, together with a strong moni-
toring and regulation of the opportunistic activity that can in-
terfere with the efforts made by organized screening. In some
Italian contexts, many efforts have been made to channel op-
portunistic screening activities within the organized system
(e.g., in Piedmont a recent regional law banned the prescrip-
tion of preventive mammograms outside the organized pro-
gramme); for these efforts to be successful, the involvement of
health care professionals, family doctors in particular, is crucial.
The assessment of diagnostic indicators confirms the trend ob-
served in previous years.5Among these, RR is one of the more
carefully monitored indicators of a programme’s specificity.
Having too many women referred for additional examinations
(FNA, core or surgical biopsy) is a recognized problem both for
operational reasons and financial costs. In addition, increased lev-
els of anxiety and other adverse psychological consequences in
women who are recalled are well-documented.7,8
In our surveys RRs exceeded or were very close to the recom-
mended standards and call for further reflection. These values,
referred to programmes that have already been running for sev-
eral years, cannot be ascribed to the learning curve effect, typ-
ical of newly implemented programmes, even though the re-
cent, gradual replacement of analogue equipment with digital
devices could partly be responsible for this. High RRs, espe-
cially at initial screening, can also be due to an increasing num-
ber of screened women aged 50-54 years.
To better assess this trend, it would be useful to evaluate the
RR by screening units and by radiologists. Multidisciplinary
sessions on screen-detected lesions, collective revision of atyp-
ical outcomes and reinforcement of the training procedures can
represent some practical approaches to improve the perform-
ance of the programmes.
As concerns overall DR and PPV, despite the presence of small
annual fluctuations, Italian mammography screening pro-
grammes show good quality activity in general and over time.
No large variations, other than the expected ones, were ob-
served for age group analysis.
The results by geographical areas prompt distinct considera-
tions. A delay in the implementation of organized screening
programmes and the absence of structured coordination sys-
tems persisted in southern Italy. This has a strong impact both
on data completeness and on the intermediate outcomes that
are struggling to reach the recommended quality standard.
Southern Italian regions continue to present critical outcomes
which would require additional analysis involving health poli-
cies and health system organization.
Our results highlighting that activity volume can affect cancer
detection accuracy are not very surprising and are consistent with
those observed in other European programmes.9The volume of
procedures or patients has been repeatedly demonstrated to be
a strong determinant of quality in medical procedures.10
Indeed, the data from the Swedish population-based screening
studies, in which mammography is performed by experts in
high-volume centres, provide the foundation from which ev-
idence-based recommendations for mammography screening
are derived.11 It is essential to discourage the activation of
screening programmes with inadequate volumes of activity and
to facilitate screening centralization as much as possible.
Our results underline a direct correlation between higher vol-
ume activity and good performances, especially for DR and
Breast cancer screening: time trends and activity volumes
38
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
PPV. Programmes with higher volumes of activity are located
mainly in central and northern Italy, where the incidence rates
for breast cancer are higher. Since DR and PPV are greatly in-
fluenced by breast cancer incidence, this should be taken into
consideration when analyzing these outcomes.
Although this analysis has many limitations, as it considers pro-
grammes and not operators, it encourages to implement new
investigation strategies which combine sensitivity and speci-
ficity indicators with programme organizational characteristics.
Overall, the results here described, despite the specified weak-
nesses, continue to be reassuring and reward the great effort un-
dertaken by screening professionals over the years. It is there-
fore important to maintain the same level of co-operation and
participation within screening experiences and support and re-
inforce indicator monitoring. In addition, further opportuni-
ties for discussing observed difficulties must be offered to the
Italian screening community, in order to suggest, test, and eval-
uate strategies for continuous improvement.
Conflicts of interests: none declared
Breast cancer screening: time trends and activity volumes
39
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Perry N, Broeders M, deWolf C et al. European guidelines for
quality assurance in breast cancer screening and diagnosis. Fourth
edition. Office for official publications of the EC, Luxembourg
2006.
2. Vainio H, Bianchini F. Breast cancer screening. IARC press, Lyon
2002.
3. Karsa L, Anttila A, Ronco G et al. Cancer screening in the Euro-
pean Union: Report on the implementation of the Council Rec-
ommendations on cancer screening. European Communities,
Luxembourg 2008.
4. Giordano L, Giorgi D, Frigerio A e il gruppo GISMa. Indicatori e
standard per la valutazione di processo dei programmi di
screening del cancro della mammella. Epidemiol Prev 2006;2
(Suppl1):1-48.
5. Giordano L, Giorgi D, Ventura L et al. Time trends of process and
impact indicators in Italian breast screening programmes (2000-
2010). Epidemiol Prev 2012;6 (Suppl 1):28-38.
6. Yankaskas BC, Cleveland RJ, Schell MJ, Kozar R. Association of re-
call rates with sensitivity and positive predictive values of screen-
ing mammography. AJR 2001;177:543-49.
7. Brett J, Bankhead C, Henderson B et al. The psychological impact
of mammographic screening: A systematic review. Psychooncol-
ogy 2005;14(11):917-38.
8. Austoker J. Women who are recalled for further investigation for
breast screening: Psychological consequences 3 years after recall
and factors affecting re-attendance. J Public Health Med 2001;
23(4):292-300.
9. Blank RG, Bennet RL, Walli MG, Moss SM. Does individual pro-
gramme size affect screening performance? Results from the UKNHS
Breast Screening Programme. J Med Screening 2002;9(1):11-14.
10. Esserman L, Cowley H, Eberle C et al. Improving the accuracy of
mammography: Volume and outcome relationship. J Natl Cancer
Inst 2002;94:369-75.
11. Nystrom L, Rutqvist LE, Wall S et al. Breast cancer screening with
mammography: Overview of Swedish randomised trials. Lancet
1993;342:973-78.
WWW.EPIPREV.IT
40
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Audit system on Quality of breast cancer
diagnosis and Treatment (QT): results
of quality indicators on screen-detected
lesions in Italy, 2011-2012
Il “progetto SQTM” sulla qualità della diagnosi
e della terapia entro i programmi di screening
in Italia: risultati 2011-2012
Antonio Ponti,1Maria Piera Mano,1Mariano Tomatis,1Diego Baiocchi,2Alessandra Barca,2Rosa Berti,3
Denise Casella,1Enrico D'Ambrosio,4Erika Delos,5Giovanni Donati,3Fabio Falcini,6Brunella Frammartino,7
Alfonso Frigerio,8Fabiola Giudici,9Paola Mantellini,10 Carlo Naldoni,11 Carlo Olla Atzeni,4
Lorenzo Orzalesi,12 Giovanni Pagano,13 Francesca Pietribiasi,14 Sabina Pitarella,1Alessandra Ravaioli,6
Anna Silvestri,7Mario Taffurelli,15 Enrica Tidone,7Fabrizio Zanconati,16 Nereo Segnan1
Abstract
This annual survey, conducted by the Italian group for mammography screening (GISMa), collects in-
dividual data on diagnosis and treatment of about 50% of screen-detected, operated lesions in Italy.
The 2011-2012 results show good overall quality and an improving trend over time. A number of crit-
ical issues have been identified, including waiting times (which have had a worsening trend over the
years) and compliance with the recommendation of not performing frozen section examination on
small lesions. Pre-operative diagnosis improved constantly over time, but there is still a large variation
between Regions and programmes. For almost 90% of screen-detected invasive cancers a sentinel
lymph node (SLN) biopsy was performed on the axilla, avoiding a large number of potentially harm-
ful dissections. On the other hand, potential overuse of SLN dissection for ductal carcinoma in situ,
although apparently starting to decline, deserves further investigation.
The detailed results have been distributed, among other ways by means of a web-based data-ware-
house, to regional and local screening programmes, in order to allow multidisciplinary discussion and
identification of the appropriate solutions to any issues documented by the data. The problem of wait-
ing times should be assigned priority. Specialist Breast Units with adequate case volume and enough
resources would provide the best setting for making monitoring effective in producing quality im-
provements with shorter waiting times.
Epidemiol Prev 2015; 39(3) Suppl 1: 40-47)
Keywords: breast cancer screening quality treatment survey, Italy
1CPO Piemonte,
AOU Città della salute
e della scienza,
Torino
2ASP Lazio, Roma
3Servizio di chirurgia
toracica, Aosta
4Anatomia patologica
Ospedale Vito Fazzi, Lecce
5Chirurgia plastica
Ospedale Vito Fazzi, Lecce
6IRCCS, Istituto scientifico
romagnolo per lo studio
e la cura dei tumori (IRST),
Forlì
7ASL di Milano,
Prevenzione oncologica,
MPC
8SSCVD Senologia
di screening,
AOU Città della salute
e della scienza, Torino
9Dipartimento di scienze
mediche, chirurgiche
e della salute, Università
degli studi di Trieste
10Istituto per lo studio
e la prevenzione
oncologica (ISPO), Firenze
11Assessorato
alle politiche per la salute,
Regione Emilia-Romagna,
Bologna
12Breast Unit Chirurgia,
AOU Careggi, Firenze
13AUSL Roma H,
Albano Laziale
14Anatomia patologica,
ASL TO5 Moncalieri
15Dipartimento
di chirurgia generale
e dei trapianti d’organo,
Chirurgia generale,
Università di Bologna
16AOU Ospedali riuniti
di Trieste, Dipartimento
di scienze mediche,
chirurgiche e della salute
Corresponding author
Antonio Ponti
antonio.ponti@cpo.it
Riassunto
Questa survey annuale, condotta dal Gruppo italiano per lo screening mammografico (GISMa), rac-
coglie dati individuali su diagnosi e terapia di circa il 50% dei casi screen-detected operati in Italia.
I risultati 2011-2012 mostrano nel complesso una buona qualità e un trend in miglioramento nel
tempo. Sono stati identificati alcuni aspetti critici, tra cui i tempi di attesa (che continuano a peggio-
rare anno dopo anno) e il rispetto della raccomandazione di non eseguire l’esame estemporaneo al
Quality of breast cancer diagnosis and treatment, 2011-2012
41
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
Mammography screening rests upon a delicate balance of hu-
man benefits and costs which is highly sensitive to the quality,
not only of the screening itself, but of the entire process of care
for screen-detected lesions. Therefore, screening programmes
should perform audits of further assessments, histopathology,
diagnosis, and treatment, as well as the screening test itself.1,2
The mammography screening movement in Europe has been
on the front line in introducing quality assurance and moni-
toring in all stages of breast cancer management and care. The
European breast cancer screening network created an individ-
ual records database and audit system called QT (audit system
on Quality of breast cancer Treatment) which can be down-
loaded at www.qtweb.it. At the same site, extensive docu-
mentation is available. QT can be used in six languages (Eng-
lish, French, German, Italian, Spanish, and Hungarian) and has
been adopted by Breast units in several European countries.
Within the Italian group for mammography screening
(GISMa), a voluntary quality assurance programme for screen-
detected breast cancer care has been ongoing since 1997,3
and results of this activity have been published yearly in the re-
ports of the National centre for screening monitoring since
their first edition in 2003. The aim of this report is to publish
results of the monitoring of diagnosis and treatment indicators
in screen-detected lesions operated with open surgery in Italy
during 2011-2012.
METHODS
Individual data on diagnosis and treatment of screen-detected
operated lesions (benign or malignant) are recorded on QT ei-
ther by clinical staff in charge of the patients or by local screen-
ing organization and evaluation units. Regional programmes
report anonymous data yearly to the national co-ordination of-
fice, which performs data quality control and analysis.
Sources of outcome measures are Italian4,5 and European2,6-
8guidelines. This report includes indicators defined recently
by a Senonetwork-GISMa consensus group.9Regions were ex-
cluded from the analysis of a given indicator if missing values
for that indicator exceeded 30%.
Even though most programmes in Italy have designated sur-
gical units where the majority of the cases are referred, the study
protocol required that participating programmes record all
screen-detected cases, regardless of where treatment had taken
place. Piemonte, Valle d’Aosta, and Toscana use as index date
the date of the screening test that originated surgical referral,
while the remaining regions use date of surgery. To avoid se-
lection bias, the study protocol requires that participating pro-
grammes record all screen-detected operated lesions. Known
interval cases, operated in the index year, could also be in-
cluded, but this was not required.
The results reported here were presented, in their preliminary
version, at the National centre for screening monitoring’s an-
nual meeting in January 2014 in Bologna. Preliminary results
were checked locally and updated. In several of the regions,
data were discussed at specific multidisciplinary meetings prior
to publication. Data have been made available to regional and
screening coordinators on a web-based data-warehouse which
allows for analysis and benchmarking.
In 2011-2012, data were reported for a portion only of the fol-
lowing regions: Lombardia (Milano), Friuli-Venezia Giulia
(Trieste), Puglia (Lecce) and Toscana (Firenze). For the re-
maining four regions, data were reported region-wide. For
the first time, results in this report are shown for ages 45-74,
as some regions have extended the screening target population
beyond the traditional 50-69 age group.
All indicators are proportions; 95% confidence intervals are
given. Data analysis was performed with the tools included in
SQTM and statistical programme R.
RESULTS
During 2000-2012, about 40,000 lesions in thirteen Italian re-
gions were documented in QT. In 2011-2012, thirty-seven
screening programmes belonging to GISMa participated in the
QT project and individual data on 8,809 cases (including
7,284 malignant lesions) in eight regions were recorded in
women between 45 and 74 years of age (table 1, p. 42).
Ductal carcinoma in situ (DCIS) accounted for 16.0% of all
malignant lesions. Of invasive tumours, 35.1% had patho-
logical size ≤10 mm. Operated benign or intraepithelial lesions
(atypical hyperplasia, lobular “carcinoma” in situ grade 1 or 2,
congelatore nelle lesioni piccole. L’indicatore sulla diagnosi preoperatoria è migliorato progressivamente negli anni ma esiste an-
cora un’elevata variazione tra Regioni e tra programmi. In quasi il 90% dei casi di cancro invasivo identificati allo screening è
stato eseguito linfonodo sentinella (LNS) per la stadiazione, evitando un gran numero di dissezioni ascellari potenzialmente dan-
nose. D’altra parte, il possibile eccessivo utilizzo del LNS nei carcinomi duttali in situ, che peraltro negli ultimi anni accenna a ri-
dursi, merita indagini ulteriori.
I risultati dettagliati di questa survey sono stati distribuiti, anche attraverso una data-warehouse accessibile sul web, ai respon-
sabili dei programmi di screening regionali e locali, allo scopo di permettere la discussione multidisciplinare, la verifica dei dati
e l’identificazione delle soluzioni appropriate ai problemi che venissero così documentati. Al problema dei tempi di attesa do-
vrebbe essere assegnato carattere di priorità e urgenza. Unità diagnostico-terapeutiche di senologia con adeguati volumi di at-
tività e sufficienti risorse fornirebbero il contesto adeguato per far sì che il monitoraggio sia efficace nel produrre miglioramenti
nella qualità e tempi di attesa accettabili.
Epidemiol Prev 2015; 39(3) Suppl 1: 40-47)
Parole chiave: screening per il cancro della mammella, qualità, diagnosi, terapia, Italia
Quality of breast cancer diagnosis and treatment, 2011-2012
42
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
atypia with columnar cells, atypical papillary lesions) repre-
sented 13% of cases with known diagnosis. However, benign
and intraepithelial lesions were systematically recorded only by
5 out of 8 regions: Piemonte, Valle d’Aosta, Emilia-Romagna,
Lazio, and Puglia. Within these regions, benign or intraep-
ithelial lesions accounted for 15% of cases (benign/malignant
ratio= 0.18, a value very similar to the one found in the
GISMa aggregated data survey). The proportion of benign and
intraepithelial lesions, as well as of DCIS, was greater in
younger women (table 2).
The proportion of N+ invasive cases was 27.4% (missing:
9.1%). Grade of invasive carcinoma was distributed as fol-
lows: 20.5% I, 54.6% II, and 24.9% III (missing: 9.5%). Nu-
clear grade of DCIS was 25.4% I, 40.2% II, and 34.4% III
(missing: 10.5%).
Results of outcome measures are shown in tables 3 and 5.
Eighty-two per cent of cancers had pre-operative cytological or
micro-histological diagnosis (table 3). This figure is higher
compared to previous years and is over the new9acceptable tar-
get of 80%. However, considerable variation exists between re-
gions (range 45%-91%) and especially between programmes.
Cases for which pre-operative diagnosis was not available are
distributed by reason in table 4. Failure in performing any non-
operative diagnosis was responsible for 14% of these cases
(16% in 2010). A non-operative diagnosis involving “suspi-
cion” of malignancy – C4 or B4, according to the classification
proposed by the EC Working group on breast screening pathol-
ogy7– rather than a higher degree of certainty was responsi-
ble for 50% of the cases (48% in 2010). The proportion of in-
adequate cytology and absolute sensitivity7of C5 were above
the target (table 3).
Waiting times were still far from the target and had even wors-
ened compared to previous years (tables 5, p. 44 and 7, p. 46).
Forty-three per cent of cancers received surgery within one
month of referral (range between regions: 34%-79%), and
30% within two months of the screening date (22%-62%)
(table 5). Just slightly more than 65% of cases received surgery
within three months after screening (59%-92%).
Number 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
of programmes
Piemonte
and Valle d’Aosta 8910 10 10 10 10 10 10 10 10 10 10
Lombardia 1- - - 111- - 1111
Veneto 2112 12 12 12 10 91----
Friuli-Venezia Giulia ------------1
Emilia-Romagna 6899810 11 11 11 11 11 11 11
Toscana 111119911 11 1111
Umbria --1----------
Lazio 25377668810 11 11 12
Campania 1------------
Puglia -----------11
Sardegna -------1111--
Sicilia 212-1--------
Total 23 25 38 39 40 48 47 50 42 34 35 35 37
Number of cases
Piemonte
and Valle d’Aosta 589 709 812 852 1,170 1,175 1,212 1,098 1,216 1,229 1,196 1,563 1,538
Lombardia 69 ---51 138 139 --439 374 418 434
Veneto 158 76 270 426 369 432 392 191 176 ----
Friuli-Venezia Giulia ------------57
Emilia-Romagna 394 796 663 742 856 920 992 984 1,107 1,129 1,103 1,536 2,016
Toscana 144 138 151 195 213 522 526 710 551 192 88 75 71
Umbria --33 ----------
Lazio 137 142 128 245 339 239 286 375 325 567 467 502 443
Campania 9------------
Puglia -----------61 95
Sardegna -------74 72 17 62 --
Sicilia 135 23 36 - 10 --------
Total 1,635 1,890 2,093 2,460 3,008 3,426 3,547 3,432 3,447 3,573 3,290 4,155 4,654
Table 1. Italian survey on diagnosis and treatment of screen-detected breast lesions, 2000-2012,age 49-70 (up to 2010) age 45-74 (from 2011). Number of screening
programmes and cases, by region.
Tabella 1. Survey sulla diagnosi e la terapia delle lesioni mammarie screen-detected, 2000-2012, età 49-70 (fino al 2010), età 45-74 (dal 2011). Numero di programmi
e di casi, per Regione.
Quality of breast cancer diagnosis and treatment, 2011-2012
43
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Guidelines recommend avoiding intra-operative frozen section
examination (even on margins) in lesions under or equal to 10
mm because of limited accuracy and the risk of deteriorating
the specimen and impairing subsequent examination.1,4-7 The
result of this indicator (table 5) was still below the target, but
had improved compared to the previous period, as in 2007
frozen section examination was performed in about one fourth,
in 2008-2009 in about one fifth, and in 2010 and 2011-2012
in one eighth of cases only (the range between regions is wide:
9%-80%). Recent Italian guidelines9recommend the per-
formance of two-view specimen X-rays on all lesions showing
micro-calcifications only and set the numerical target at 90%.
The indicator (table 5) gives a result of 66.0%. The number
of missing data however is high (21%).
Breast conservation, both for invasive cancer (up to 3 cm)9and
DCIS (up to 2 cm), was at high levels, 85% the former and
90% the latter. The proportion of axillary dissections with an
adequate number of lymph nodes excised (92%) exceeded the
target (table 5). The indicator on performing no more than
one operation on the breast for clearing margins met the 90%
target both for invasive cancer and DCIS. Margins were left
wider than 1 mm in 93% of cases (table 5).
This survey investigated the gradual introduction over the
years of the sentinel lymph node (SLN) biopsy, which makes
staging possible with considerably fewer complications than
axillary clearance.4,8 An increasing proportion of invasive
cancers and DCIS were studied with SLN biopsy over time
until 2007-2008, then the use of SLN biopsy in invasive can-
cers reached a plateau around 87% while in DCIS it seemed
to start decreasing from a maximum of 62% in 2010 to 53%
in 2012 (figure 1, p. 44). The proportion of node-negative in-
vasive cases staged by SLN biopsy only (table 5 and table 7)
was 91% in 2011-2012, with an increasing trend over the
years and moderate variability by region (range 73%-100%).
In 92% of cases no more than 3 sentinel lymph nodes were ex-
cised, as prescribed by the target (table 5).
In 2011-2012, 3.3% of DCIS (range between regions: 0%-
7%) received clearance of the axilla (table 5), a procedure not
recommended in these cases. The result of this indicator has
improved over the years (table 7).
Overtreatment may also result from unnecessary open surgery
in the breast on benign lesions.This issue is illustrated in table
6(p. 45) where operated benign or intraepithelial lesions are
distributed by histopathology type. Benign lesions at no in-
creased risk (all except intraepithelial lesions, papilloma, scle-
rosing adenosis, radial scar, and phylloid tumours) were 524
in 2011-2012 (49% of all operated benign or intraepithelial
Histopathological diagnosis Age 45-49 Age 50-59 Age 60-69 Age 70-75 Missing Total
N%N%N%N%N%N%
benign 231 18.0 293 11.6 199 6.2 34 3.6 21 2.4 778 8.8
intraepithelial 118 9.2 115 4.6 80 2.5 14 1.5 3 0.3 330 3.7
lobular carcinoma in situ (LIN 3) 2 0.2 1 0.0 4 0.1 0 0.0 1 0.1 8 0.1
ductal carcinoma in situ 208 16.3 351 14.0 375 11.7 123 13.0 91 10.5 1,148 13.0
micro-invasive 15 1.2 40 1.6 43 1.3 14 1.5 2 0.2 114 1.3
invasive (1A/1B) 40 3.1 136 5.4 178 5.6 31 3.3 49 5.6 434 4.9
invasive (other) 172 13.4 461 18.3 760 23.8 264 27.8 145 16.7 1,802 20.5
invasive (unknown size) 443 34.6 949 37.7 1,414 44.3 439 46.3 292 33.6 3,537 40.2
malignant not specified 10 0.8 48 1.9 67 2.1 13 1.4 103 11.8 241 2.7
unknown 41 3.2 122 4.8 75 2.3 16 1.7 163 18.7 417 4.7
Total 1,280 100 2,516 100 3,195 100 948 100 870 100 8,809 100
Table 2. It. Italian survey on diagnosis and treatment of screen-detected breast lesions, 2011-2012. Distribution by final histopathology diagnosis and age.
Tabella 2. Survey sulla diagnosi e la terapia delle lesioni mammarie screen-detected, 2011-2012. Distribuzione per diagnosi istopatologica definitiva ed età.
Outcome measure Eligible Missing Result 95%CI Minimum % Target
cases %% required %
pre-operative diagnosis in cancers (C5,B5) 6,878 2.6 82.2 81.3 - 83.1 ≥80 ≥90
non-inadequate cytology if final diagnosis is cancer 4,381 0.6 91.9 91.1 - 92.7 ≥90
absolute sensitivity C5 4,381 0.6 67.6 66.2 - 69.0 ≥60
Table 3. Summary on diagnostic indicators,2011-2012, age 45-74. Results are calculated on eligible cases minus cases with missing information.
Tabella 3. Indicatori diagnostici, 2011-2012, età 45-74. I casi con informazione mancante sono esclusi dal denominatore.
N%
pre-operative diagnosis not performed 171 14.3
unsatisfactory 136 11.4
false negative (C2 or B2) 43 3.6
dubious (C3 o B3) 252 21.1
suspicious (C4 o B4) 592 49.6
Total 1,194 100.0
Table 4. Distribution of malignant cases without pre-operative diagnosis C5 or B5
by reason, 2011-2012, age 45-74.
Tabella 4. Distribuzione delle lesioni maligne senza diagnosi preoperatoria C5 o
B5, per motivo della mancata diagnosi preoperatoria,2011-2012, età 45-74.
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
%
10
20
30
40
50
60
70
80
90
100
0
Figure 1. Italian survey on di-
agnosis and treatment of
screen-detected breast can-
cers. Trend (%) in the use of
SLN technique (age 49-74).
Years 2001-2012.
Figura 1. Survey sulla dia-
gnosi e la terapia dei carci-
noma della mammella screen-
detected. Trend (%) nell’uso
della tecnica del linfonodo
sentinella (età 49-70). Anni
2001-2012.
invasive cancer
DCIS
Quality of breast cancer diagnosis and treatment, 2011-2012
44
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Table 5. Summary on surgical indicators, 2011-2012, age 45-74. Results are calculated on eligible cases minus cases with missing information. Due to missing values
exceeding 30%, some regions were excluded from the calculation of specific indicators.
Tabella 5. Indicatori chirurgici, 2011-2012, età 45-74. I casi con informazione mancante sono esclusi dal denominatore. Sono state escluse dal calcolo di specifici in-
dicatori le Regioni con una proporzione di valori mancanti >30%.
Outcome measure Eligible Missing Result CI95% Minimum Target Excluded
cases %% % required %
waiting time for surgery from referral ≤30 days 7,263 16.7 43.5 42.3-44.8 ≥75 ≥90 Lombardia, Puglia
waiting time for surgery from first diagnostic 7,263 8.3 28.8 27.8-30.0 ≥75 ≥90 Lombardia, Puglia
test ≤42 days
waiting time for surgery from screening test 7,123 10.2 29.9 28.8-31.0 ≥75 ≥90 Lombardia, Puglia,
≤60 days Toscana
waiting time for surgery from screening test 7,123 10.2 65.4 64.2-66.5 Lombardia, Puglia,
≤90 days Toscana
frozen section not performed in cancers 1,423 12.0 87.5 85.6-89.3 ≥95 Lazio, Lombardia,
≤10 mm Toscana
specimen X-ray in cases with 768 21.2 66.3 62.3-70.0 ≥90 ≥98 Puglia
microcalcifications only
only one operation after pre-operative 5,728 0.7 92.9 92.2-93.6 ≥80 ≥90
diagnosis (invasive)
only one operation after pre-operative 1,112 0.4 89.9 87.9-91.6 ≥80 ≥90
diagnosis (in situ)
conservative surgery in invasive cancers 5,367 10.5 84.7 83.6-85.7 ≥70 ≥90
≤30 mm
conservative surgery in DCIS (ductal carcinoma 511 1.2 90.1 87.1-92.5 ≥80 ≥90
in situ) ≤20 mm
margins >1 mm after last surgery 4,547 18.5 92.8 91.9-93.6 Lazio, Lombardia
number of lymph nodes >9 in axillary dissection 1,057 2.3 92.3 90.4-93.8 ≥80 ≥90
(sampling excluded)
axillary staging by SLN only in pN0 3,407 0 91.1 90.1-92.0 ≥80 ≥90
no axillary dissection (sampling included) 1,106 6.1 96.7 95.4-97.7 ≥90 ≥95
in DCIS
no more than 3 LNs at SLN biopsy 5,726 29.5 92.4 91.5-93.2 ≥80 ≥90 Lombardia, Puglia
Quality of breast cancer diagnosis and treatment, 2011-2012
45
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
lesions, excluding double lesions and lesions with missing his-
tological type: a result similar to previous years).
Table 7 shows time trends from 2000 to 2012 for selected per-
formance parameters. The frequency of pre-operative diagno-
sis and avoidance of frozen section examination in small lesions
showed improvement over time. Waiting times had a consis-
tent and important negative trend over the years.
DISCUSSION
In 2011-2012, most outcome measures were near or met the
target set by GISMa.5,9 Major exceptions, similarly to 2010,
were waiting times for surgery, compliance with the recom-
mendation on avoiding frozen section examination on small le-
sions and performing specimen X-rays.
The proportion of cancers with pre-operative diagnosis has
clearly increased over the years, due to increasing use of micro-
histology techniques, and reached the acceptable target for the
first time in 2005. However, the result only slightly increased
compared to 2007, despite the fact that a wide margin for im-
provement still exists in order to reach the European desirable
target of 90%.7This is also supported by the finding of a con-
siderable variation between programmes: about 25% did not
reach the acceptable target, while more than 20% did. Pathol-
ogists and radiologists should be involved with surgeons in an-
alyzing the reasons for underperformance in programmes scor-
ing in the lower part of the range. It may be worthy of notice
that fine needle aspiration cytology (FNA) was still used for
pre-operative diagnosis in the majority of cases: out of 7,449
lesions receiving needle biopsies, 3,560 (48%) received FNA
only, 2,620 (35%) core or vacuum assisted biopsy only, and
1,269 (17%) both.
Waiting time from screening to surgery embraces much of the
entire process of care (time from screening to first assess-
ment, time from first assessment to result, time from result
of assessment to first surgery). Results have been worsening
over the years, and in 2011-2012 the decreasing trend con-
tinued, with as few as 30% of patients being operated within
60 days of the screening examination. Regional authorities
should inspect the reasons for this considerable delay, espe-
cially in regions in the lower part of the range. Even though
two or three months of treatment delay are not expected to
affect clinical outcomes,10 they can cause anxiety and impair
quality of life, in addition to contradicting the idea itself of
early detection. Furthermore, many cases experience a delay
greater than three months.
Avoiding the use of frozen section examination entails a dif-
ficult change in attitude by the surgeon, when it is not due to
lack of pre-operative diagnosis. This procedure, even when
aimed at the evaluation of margins in impalpable lesions,
should be substituted by two-view specimen X-ray.4,9
Use of axillary dissection in DCIS was in compliance with
the target (less than 5%) but could further decrease, since this
procedure is useless in DCIS and is a potential cause of
complications. Pre-operative multidisciplinary discussion is
the way to minimize this problem, as only through discus-
sion with the pathologist and radiologist can the surgeon
learn about the non-invasiveness of the lesion.8This should
also help in decreasing the use in benign lesions, LIN, and
low- and intermediate-grade DCIS, of SLN dissection, which
is not free of complications. Importantly, for the first time,
this survey shows a decline in the use of SLN biopsy in
DCIS.
The proportion of missing values is still relatively large for
waiting time, frozen section examination, and performance of
specimen X-ray.
Although this survey includes a large share of screen-detected
N%
benign normal tissue 15 1.4
fibroadenoma 161 15.1
cysts 17 1.6
columnar cell change without atypia 8 0.7
fibrocystic breast diseae 102 9.5
benign phylloid tumour 20 1.9
sclerosing adenosis 80 7.5
radial scar 21 2.0
papilloma/papillomatosis 110 10.3
other 149 13.9
unknown 72 6.7
total benign 755 70.6
intraepithelial atypical lobular hyperplasia (LIN1) 16 1.5
lobular carcinoma in situ (LIN2) 65 6.1
atypical columnar cell change (DIN1a) 66 6.2
atypical ductal hyperplasia (DIN1b) 165 15.4
atypical papillary lesion 2 0.2
total intraepithelial 314 29.4
Total 1,069 100.0
Table 6. Distributionby histo-
logical type of benign and in-
traepithelial lesions operated by
open surgery (excluding syn-
chronous lesions), age 45-74.
Years 2011-2012
Tabella 6. Distribuzione per
tipo istologico delle lesioni be-
nigne e intraepiteliali operate
(lesioni sincrone escluse), età
45-74. Anni 2011-2012.
Quality of breast cancer diagnosis and treatment, 2011-2012
46
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
malignant cases in Italy (about 50% of cases documented in
the GISMa aggregated data survey), a selection towards in-
clusion of cases from better-organized Regions cannot be ex-
cluded. Benign operations, furthermore, are under-recorded
in some of the Regions. A larger participation in the survey
by Italian regions and programmes would be appropriate, per-
haps coupled with simplified data collection methods. On the
other hand, it is important to maintain the connection be-
tween screening and clinical Breast units11,12 that has been
established by this project over the years: a strong point of this
project is the production of timely and detailed information
of interest to both clinicians and public health professionals.
Acknowledgments
This survey was conducted by the multidisciplinary group on
therapy of the Italian group for mammography screening , with
co-ordination by CPO Piemonte. The project and develop-
ment of QT has been sponsored by the «Europe Against Can-
Indicator Eligible 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Min. Target
% req.
pre-operative diagnosis 33,397 52.4 58.1 61.4 66.5 69.9 73.2 73.7 75.8 78.2 76.9 80.3 81.5 84.3 ≥80 ≥90
in cancers (C5,B5)
waiting time for surgery 24,362 63.1 54.8 59.0 59.0 56.4 60.6 58.2 53.8 52.2 45.3 43.6 44.7 42.5 ≥75 ≥90
from referral ≤30 days
waiting time for surgery 29,560 69.2 49.6 47.4 46.6 41.3 42.7 42.3 36.8 32.9 35.3 31.3 30.2 27.9 ≥75 ≥90
from first diagnostic
test ≤42 days
waiting time for surgery 27,918 60.4 54.2 58.5 55.4 55.2 52.3 48.7 44.2 39.6 41.2 38.0 32.9 26.9 ≥75 ≥90
from screening test
≤60 days
waiting time for surgery 27,918 87.0 79.6 82.7 80.1 80.4 79.2 78.9 75.7 70.0 73.6 71.1 68.9 61.9
from screening test
≤90 days
frozen section not 6,200 44.4 51.8 59.6 68.3 79.5 73.0 69.3 75.8 81.0 86.1 87.2 90.8 89.4 ≥95 ≥95
performed in cancers
≤10 mm
specimen X-ray in cases 1,960 77.7 58.2 61.2 34.2 45.1 45.3 57.1 32.9 44.2 64.8 68.8 64.2 68.4 ≥90 ≥98
with microcalcifications
only
only one operation after 23,523 84.9 85.4 87.1 87.8 87.9 88.7 90.0 90.4 91.3 91.8 92.8 92.4 92.4 ≥80 ≥90
pre-operative diagnosis
(invasive)
only one operation after 4,443 74.8 81.6 82.9 86.0 86.0 86.6 86.1 87.3 86.4 88.5 90.5 90.3 89.0 ≥80 ≥90
pre-operative diagnosis
(non-invasive)
conservative surgery 20,680 85.2 84.3 83.1 86.6 86.9 88.4 87.9 88.0 88.9 88.6 86.6 87.1 84.7 ≥70 ≥90
in invasive cancers
≤30 mm
conservative surgery 2,956 89.8 89.4 89.0 88.5 93.5 93.0 89.1 92.3 91.0 95.5 93.9 92.8 88.2 ≥80 ≥90
in DCIS (ductal
carcinoma in situ)
≤20 mm
margins >1 mm after 20,579 85.5 85.1 83.2 87.3 89.0 90.1 89.4 89.2 89.4 93.6 90.9 93.5 93.4
last surgery
number of lymph nodes 7,048 92.9 95.0 95.1 92.1 90.4 93.3 92.4 92.6 91.0 90.2 91.5 93.8 90.8 ≥80 ≥90
>9 in axillary dissection
(sampling excluded)
axillary staging by SLN 14,741 0 14.7 47.9 60.2 69.1 75.6 82.9 86.3 89.4 91.7 90.1 90.3 92.2 ≥80 ≥90
only in pN0
no axillary dissection 4,103 79.7 85.9 93.2 89.2 96.0 94.5 93.6 93.8 97.4 97.3 97.8 95.0 98.3 ≥90 ≥95
in DCIS
no more than 3 LNs 20,276 - 94.0 95.5 93.2 94 94.5 92.8 92.9 92.3 93.6 94.0 92.7 94.2 ≥80 ≥90
at SLN biopsy
Table 7. Time trends for selected indicators (%), 2000-2012, age 49-70. Only regions having contributed data for the whole period (Piemonte, Valle d’Aosta, Emilia-
Romagna, Toscana, Lazio) were included. Due to missing values exceeding 30%, Lazio was excluded from the indicators for waiting time for surgery from referral, spec-
imen X-ray, and no more than 3 LNs at SLN biopsy.
Tabella 7. Andamento temporale (%) per alcuni indicatori, 2000-2012, età 49-70. Sono incluse solo le Regioni che hanno contribuito per l’intero periodo (Piemonte,
Valle d’Aosta,Emilia-Romagna,Toscana e Lazio).Avendo una proporzione di valori mancanti >30%, il Lazio è escluso dal calcolo degli indicatori sui tempi di attesa dalla
prescrizione, l’esecuzione della Rx sul pezzo e il numero di linfonodi sentinella escissi.
Quality of breast cancer diagnosis and treatment, 2011-2012
47
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. National Co-ordination group for surgeons working in breast can-
cer screening. Quality assurance guidelines for surgeons in breast
cancer screening. NHSBSP, Publication n. 20, 1996.
2. Perry N, Blichert-Toft M, Cataliotti L et al. Quality assurance in the
diagnosis of breast disease. Eur J Cancer 2001;37:159-72.
3. Distante V, Mano MP, Ponti A. Monitoring surgical treatment of
screen-detected breast lesions in Italy. Eur J Cancer 2004;40:1006-12.
4. Forza operativa nazionale sul carcinoma mammario. I tumori della
mammella. Lineeguida sulla diagnosi, il trattamento e la riabilitazione.
Firenze, 2003. Updated in: Attualità in Senologia 2005;46:33-106.
5. Mano MP, Distante V, Ponti A et al. Monitoraggio e promozione
della qualità del trattamento del carcinoma mammario nelle Unità
di senologia e nei programmi di screening in Italia. Attualità in
Senologia 2001;10 (Suppl 1).
6. Rutgers EJT, Bartelink H, Blamey R et al. Quality control in loco-re-
gional treatment for breast cancer. Eur J Cancer 2001;37:447-53.
7. Perry N, Broeders M, de Wolf C et al. European guidelines for
quality assurance in breast cancer screening and diagnosis. 4th edi-
tion. European commission, Europe against cancer programme,
Luxembourg 2006.
8. Rosselli del Turco MR, Ponti A, Bick U et al. Quality indicators in
breast cancer care. Eur J Cancer 2010;46:2344-56.
8. Mano MP, Ponti A, Angiolini C et al. Indicatori di qualità per la cura
del carcinoma mammario nelle Breast Unit in Italia: una proposta
congiunta GISMa-Senonetwork. [www.senonetwork.org].
10. Richards MA, Westcombe AM, Love SB et al. Influence of delay on
survival in patients with breast cancer: a systematic review. Lancet
1999;353:1119-26.
11. Blamey R, Blichert-Toft M, Cataliotti L et al. Breast units: future stan-
dards and minimum requirements. Eur J Cancer 2000, 36: 2288-93.
12. Wilson RA, Marotti L, Bianchi S et al. The requirements of a specialist
Breast Centre. Eur J Cancer 2013;49:3579-87.
cer» and EUNICE (European network for information on can-
cer) programmes of the European Commission, the Ministry of
Health, and Lega italiana per la lotta contro il cancro through
ONS, Regione Piemonte, and Fondazione San Paolo, Turin.
We are grateful to the many clinical specialists and persons in-
volved in screening evaluation and organization who con-
tributed to data collection, and to the regional screening co-or-
dination centres.
Conflicts of interests: none declared
Information provided by Italian breast
cancer screening programmes:
a comparison between 2001 and 2014
Informazioni fornite dai programmi di screening
mammografico in Italia: un confronto tra il 2001
e il 2014
Roberta Castagno,1Debora Canuti ,2Marco Petrella,3Lauro Bucchi,4Chiara Fedato,5Francesca Garena,1
Livia Giordano1
Abstract
Debate on efficacy, benefits, and risks of breast cancer screening continues to rage, and scientific con-
troversy surrounding overdiagnosis, false positives/false negatives, raises questions about communi-
cation to women attending screening programmes.
The study compares information provided by invitation letters and leaflets of Italian breast screening
programmes in 2001 (N=47) and 2014 (N=80). At both times, nearly all programmes provided ade-
quate practical information and details about screening objectives and test procedures. Information
regarding epidemiology/figures was scarce or absent in 2001, while in 2014 a number of programmes
began to inform women about screening risks (false negative and positive results and overdiagnosis,
65%, 16%, and 21% respectively) although actual figures were rarely supplied.
Despite this small improvement, Italian programmes are still far from giving balanced information.
Further efforts should be addressed to providing accurate and transparent information, enabling
women to make an informed choice.
(Epidemiol Prev 2015; 39(3) Suppl 1: 48-51)
Keywords: breast cancer screening, invitation letter, leaflets, overdiagnosis, Italy
1Unità di epidemiologia
dei tumori, CPO Piemonte,
AOU Città della salute
e della scienza, Torino
2Azienda sanitaria locale
della Romagna,
Emilia-Romagna
3ASL 2 Umbria - last
affiliation
4Registro tumori
della Romagna, Istituto
scientifico romagnolo
per lo studio e la cura
dei tumori (IRST) IRCCS,
Meldola, Forlì
5Settore promozione
e sviluppo igiene e sanità
pubblica, Regione Veneto,
Venezia
Corresponding author
Roberta Castagno
roberta.castagno@cpo.it
Riassunto
Il dibattito sull’efficacia, i benefici ed i rischi dello screening mammografico, in termini di sovradiagnosi
e sovratrattamento, falsi positivi/negativi, hanno portato a riflettere su quale tipo di comunicazione oc-
corre dare alle donne. Lo studio confronta le informazioni fornite dalle lettere di invito e gli opuscoli
dei programmi di screening mammografico italiani nel 2001 (N=47) e nel 2014 (N=80). Quasi tutti i
programmi, sia nel 2001 che attualmente, forniscono adeguate informazioni logistico-organizzative
e dettagli sugli obiettivi dello screening e la procedura del test. Le informazioni epidemiologiche/nu-
meriche, nel 2001, sono per lo più assenti o solo raramente presenti, mentre nel 2014 alcuni pro-
grammi cominciano a dare informazioni anche sui rischi dello screening (falsi negativi, falsi positivi e
sovradiagnosi, rispettivamente 65%, 16% e 21% ), anche se solo raramente quantificano tali concetti.
Nonostante qualche miglioramento, i programmi italiani non forniscono ancora informazioni com-
plete e bilanciate. Saranno quindi necessari ulteriori sforzi per migliorare la capacità dei programmi nel
produrre e trasmettere un’efficace comunicazione sullo screening mammografico al fine di permet-
tere alle donne di fare una scelta informata.
(Epidemiol Prev 2015; 39(3) Suppl 1: 48-51)
Parole chiave: screening mammografico, lettere di invito, opuscoli, sovradiagnosi, Italia
WWW.EPIPREV.IT
48
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Communication in breast cancer screening programmes
INTRODUCTION
The former approach to breast cancer screening information
emphasized screening benefits for the population, following the
imperative of achieving adequate uptake to have an impact on
mortality.1
Over the last few years, there has been growing European con-
cern about risks and benefits of mammography screening2-3 and
how to communicate this to women.4Whether breast screen-
ing causes more harm than good has been widely debated. The
main issues are how great the benefits of screening are in terms
of reduced breast cancer mortality2,5 and how significant the
harms are, especially in terms of overdiagnosis (defined as can-
cers detected at screening that would not have otherwise become
clinically apparent in a woman's lifetime2) and false positive and
false negative outcomes.7-9 Therefore, the entire scientific com-
munity now supports the need for balanced information that
explains both the harms and benefits for women attending
screening. Invitation letters and written information material are
the most common means of communication used by organized
screening programmes.
In spring 2014, the Italian group for mammography screening
(GISMa) promoted a survey to investigate what information
Italian organized breast cancer screening programmes provide
to women. The same investigation had been carried out in
2001. The aim of the current study is to compare the two sur-
veys to verify how mammography screening information has
evolved over time.
METHODS
In 2001 and 2014, invitation letters and leaflets in use by Ital-
ian organized mammography screening programmes were col-
lected and evaluated through a score sheet designed for this
purpose. The score sheet assesses the presence of logistic and
organizational information, screening objectives, mammogra-
phy and screening information, and epidemiological/quanti-
tative data, including the presence of epidemiological figures
and estimates. All issues are detailed in table 1 (p. 50).
All materials were assessed by two readers with the support of
a supervisor.There was no evaluation concerning layout qual-
ity and wording of these tools in this phase of the study.
RESULTS
Nearly 90% of active programmes in Italy responded both
years (53/60 programmes in 2001 and 110/124 in 2014).
Among these, 47 and 80 information sets (invitation letter
plus leaflet) were included in the 2001 and 2014 analysis, re-
spectively.
The main results of the two surveys are presented in table 1 and
summarized below.
Logistic and organizational information
Compared to 2001, in 2014 a greater number of programmes
notified women about how and when to obtain their mam-
mography results (88.8% vs 61.6% and 33.8% vs 17.0%, re-
spectively).
In 2001, no programme conveyed messages of informed con-
sent and only 6.4% informed on data confidentiality. In
2014, 25% of programmes mentioned informed consent and,
after the Data Protection Code came into effect in 2003,
many more of the information tools in use referred to data
confidentiality (45.1%). In addition, in 2014 nearly 75% of
programmes provided explanations about quality control ac-
tivities and the involvement of properly trained professionals
(compared to only 17.0% in 2001).
This type of practical information was present and carefully
described in both surveys and was essentially conveyed by the
invitation letter.
Screening objectives
The percentage of tools describing «what a screening pro-
gramme is» more than doubled over time (44.6% in 2001 vs
92.5% in 2014).
The entirety of programmes fully described the target popu-
lation and benefits of mammography screening (in terms of
the importance of early detection to reduce breast cancer
mortality and increase the chances of recovery), both in 2001
and 2014.
Mammography and screening information
In both surveys almost all programmes described «what a
mammography is» (93.6% in 2001 vs 98.9% in 2014) and the
interval between the two tests (95.7% in 2001 vs 100% in
2014).
The percentage of tools specifying the double reading of the
test was nearly twice in 2014 compared to 2001 (56.3% vs
27.7%).
In 2001, poor information about side effects (pain and dis-
comfort caused by the test) and radiation-related risks were
provided (34% and 6%, respectively) compared with today’s
material (86.4% on both topics).
In 2014, 97.5% of programmes informed women about the
possibility to be recalled for further assessments (68% in 2001),
55.1% described what further assessments consist of (no pro-
grammes in 2001) and 25% also stated the rate (recall rate).
Detailed information related to test procedures was provided
both in 2001 and 2014 almost exclusively by the information
leaflet.
Epidemiological information/quantitative
data
In 2001, epidemiological information and numerical data
were very rare or missing.The data mentioned by programmes
were breast cancer incidence (14.9%), lifetime risk of devel-
oping breast cancer (8.5%), and relative risk reduction mor-
tality (23.4%). No information was given about overdiagno-
sis, false negative and false positive results.
In 2014, a greater number of tools illustrated information
about breast cancer incidence (16.3%), lifetime risk of devel-
oping breast cancer (20%), and relative risk reduction mor-
tality (25.1%). Furthermore, some programmes also began to
49
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
Communication in breast cancer screening programmes
50
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
inform women about overdiagnosis (21.3%), false negative re-
sults (65.1%), and false positive results (16.3%). Nevertheless,
numerical data were seldom provided. All this information,
when present, was conveyed by the leaflets.
DISCUSSION
Comparison between the two surveys shows that completeness
of information has increased over time. Nevertheless, there is
still a great degree of variation in the information provided by
mammography screening programmes, especially relating to
epidemiological and numerical information.
Italian programmes have consistently provided adequate lo-
gistic and organizational information, which is conveyed
mainly by the invitation letters.
Leaflets attached to invitation letters give more detailed in-
formation about screening programme organization, test pro-
cedures and benefits and harms of mammography screening.
The latter, such as overdiagnosis, false negative and false pos-
Invitation letter only (%) Leaflet only (%) Both (%) Total (%)
2001 2014 2001 2014 2001 2014 2001 2014
(N=47) (N=80) (N=47) (N=80) (N=47) (N=80) (N=47) (N=80)
Logistic and organizational information
How to fix and/or how to change 51.1 41.3 2.1 1.3 40.4 52.5 93.6 95.1
the appointment
Documents women should bring 38.3 61.3 6.4 0.0 40.4 38.8 85.1 100.0
Free test or not 17.0 15.0 0.0 0.0 83.0 83.8 100.0 98.8
How to get the results 10.6 18.8 25.5 25.0 25.5 45.0 61.6 88.8
When to get the results 2.1 10.0 14.9 23.8 0.0 0.0 17.0 33.8
Informed consent 0.0 12.5 0.0 10.0 0.0 2.5 0.0 25.0
Data confidentiality 2.1 38.8 4.3 3.8 0.0 2.5 6.4 45.1
Quality control/operator training 0.0 1.3 17.0 63.8 0.0 8.8 17.0 73.9
Screening objectives
What a screening programme is 2.1 20.0 40.4 7.5 2.1 65.0 44.6 92.5
Mammography benefits 0.0 3.8 66.0 63.8 34.0 32.5 100.0 100.0
Who the test is for 0.0 6.3 36.2 36.3 63.8 57.5 100.0 100.0
Mammography and screening information
What a mammography is 0.0 3.8 63.8 73.8 29.8 21.3 93.6 98.9
Screening interval 4.2 6.3 51.1 47.5 40.4 46.3 95.7 100.0
How it is performed 0.0 0.0 29.8 47.5 0.0 3.8 29.8 51.3
How long it takes 0.0 0.0 59.6 48.8 26.7 3.8 86.3 52.6
Who reads the test 0.0 1.3 27.7 50.0 0.0 5.0 27.7 56.3
Side effects 2.1 1.3 29.8 78.8 2.1 6.3 34.0 86.4
Radiation risk 0.0 1.3 6.4 83.8 0.0 1.3 6.4 86.4
Breast awareness 0.0 0.0 0.0 77.5 0.0 1.3 0.0 78.8
Further assessments (mentioned) 23.4 2.5 36.2 45.0 8.5 50.0 68.1 97.5
Further assessments (described) 0.0 1.3 0.0 53.8 0.0 0.0 0.0 55.1
Epidemiological and quantitative data
Breast cancer incidence 0.0 2.5 14.9 13.8 0.0 0.0 14.9 16.3
Lifetime risk of developing breast cancer 0.0 0.0 8.5 20.0 0.0 0.0 8.5 20.0
Lifetime risk of dying from breast cancer 0.0 0.0 0.0 6.3 0.0 0.0 0.0 6.3
Survival from breast cancer 0.0 0.0 2.1 2.5 0.0 0.0 2.1 2.5
Relative risk reduction mortality 0.0 0.0 23.4 21.3 0.0 3.8 23.4 25.1
Absolute risk reduction mortality 0.0 0.0 0.0 11.3 0.0 0.0 0.0 11.3
Proportion of screened women 0.0 0.0 0.0 25.0 0.0 0.0 0.0 25.0
who would be recalled
Proportion of breast cancers detected 0.0 1.3 2.1 18.8 0.0 0.0 2.1 20.1
by mammography (sensitivity)
Proportion of women without breast cancer who 0.0 1.3 0.0 0.0 0.0 0.0 0.0 1.3
would have a positive mammogram (specificity)
Proportion of women with positive mammogram 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
who would have a breast cancer (PPV)
False negative results 0.0 0.0 0.0 61.3 0.0 3.8 0.0 65.1
False positive results 0.0 0.0 0.0 16.3 0.0 0.0 0.0 16.3
Overdiagnosis 0.0 0.0 0.0 21.3 0.0 0.0 0.0 21.3
Table 1. Information provided in invitation letters and leaflets in 2001 and 2014.
Tabella 1. Informazioni fornite nelle lettere di invito e nelle brochure distribuite nel 2001 e nel 2014.
Communication in breast cancer screening programmes
51
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Raffle AE. Information about screening – is it to achieve high up-
take or to ensure informed choice? Health Expectations 2001;4(2):
92-98.
2. Independent UK Panel on Breast Cancer Screening. The benefits
and harms of breast cancer screening: an independent review.
Lancet 2012;380:1778–86.
3. Gotzsche PC, Nielsen M. Screening for breast cancer with mam-
mography. Cochrane Database Syst Rev 2011;1:CD001877.
4. Giordano L, Cogo C, Patnick J, Paci E. Communicating the balance
sheet in breast cancer screening. J Med Screen 2012;19(1):67-71.
5. Broeders M, Moss S, Nystrom L et al. The impact of mammo-
graphic screening on breast cancer mortality in Europe: a review
of observational studies. J Med Screen 2012;19(Suppl. 1):14-25.
6. Puliti D, Duffy SW, Miccinesi G et al. Overdiagnosis in mammo-
graphic screening for breast cancer in Europe: a literature review.
J Med Screen 2012;19(Suppl 1):42-56.
7. Jørgensen KJ, Gøtzsche PC. Overdiagnosis in publicly organised
mammography screening programmes: systematic review of inci-
dence trends. BMJ 2009;9:339:b2587.
8. Duffy SW, Chen THH, Smith RA et al. Real and artificial contro-
versies in breast cancer screening. Breast Cancer Management
2013;2: 519-28.
9. Hofvind S, Ponti A, Patnick J et al. False-positive results in mam-
mographic screening for breast cancer in Europe: a literature re-
view and survey of service screening programmes. J Med Screen
2012;19(Suppl 1):57-66.
itive results, were mentioned more often in 2014, although
rarely quantified.
Despite this small improvement, Italian programmes are still
far from providing balanced information. Adequate commu-
nication (including figures and estimates) about all negative ef-
fects of screening is still a challenge that requires the efforts and
resources of the entire screening community.
This analysis may be taken as a starting point for defining the
most appropriate tools and circumstances to facilitate an in-
formed choice. It could also help to evaluate strategies to im-
prove the quality of information.
In a screening context, information can be conveyed by vari-
ous means, even though written materials (invitation letter plus
leaflet) remain the main source of communication, especially
in organized screening programmes. A crucial issue that needs
to be discussed within the GISMa group is that of how to pro-
mote consistency of breast cancer screening information among
Italian programmes. In particular, discussion should focus on
the need for recommendations concerning the contents of in-
vitation letters and leaflets, to standardize invitation tools na-
tionwide.
Moreover, the quality of layout and wording of the material
should also be studied in depth, to assess information accuracy,
especially in terms of clarity of language and syntax.
Conflicts of interests: none declared
WWW.EPIPREV.IT
52
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
Problems, solutions, and perspectives in the evaluation
of interval cancers in Italian mammography screening
programmes: a position paper from the Italian group
for mammography screening (GISMa)
Problemi, soluzioni e prospettive nella valutazione dei cancri
d’intervallo nei programmi italiani di screening mammografico:
un position paper del Gruppo italiano screening
mammografico (GISMa)
Lauro Bucchi,1Alfonso Frigerio,2Manuel Zorzi,3Chiara Fedato,4Giovanni Angiolucci,5Daniela Bernardi,6Cinzia Campari,7Emanuele Crocetti,8
Stefano Ferretti,9Daniela Giorgi,10 Francesca Marchisio,11 Doralba Morrone,12 Carlo Naldoni,13 Marco Petrella,14 Antonio Ponti,15
Alessandra Ravaioli,1Gianni Saguatti,16 Dolores Santini,11 Priscilla Sassoli de Bianchi,13 Monica Serafini,17 Viviana Vergini,15 Livia Giordano15
1Registro tumori della Romagna, Istituto scientifico romagnolo per lo studio e la cura dei tumori (IRST) IRCCS, Meldola, Forlì
2Centro di riferimento regionale per lo screening mammografico, Torino
3Registro tumori del Veneto, Regione Veneto, Padova
4Coordinamento regionale screening, Regione Veneto, Venezia
5Senologia diagnostica, Azienda unità sanitaria locale, Arezzo
6Unità operativa di senologia clinica e screening mammografico, Azienda provinciale servizi sanitari, Trento
7Centro screening, Azienda unità sanitaria locale, Reggio Emilia
8Unità di epidemiologia clinica e descrittiva, Istituto per lo studio e la prevenzione oncologica, Istituto tumori toscano, Firenze
9Registro tumori di Ferrara, Ferrara
10Unità di epidemiologia, Azienda sanitaria locale, Istituto tumori toscano, Lucca
11Centro screening mammografico, Azienda unità sanitaria locale, Modena
12Unità operativa di senologia, Istituto per lo studio e la prevenzione oncologica, Istituto tumori toscano, Firenze
13Assessorato alle politiche per la salute, Regione Emilia-Romagna, Bologna
14Unità operativa di epidemiologia, Azienda unità sanitaria locale 2 dell’Umbria, Perugia
15Unità di epidemiologia dei tumori, Centro di riferimento per l'epidemiologia e la prevenzione oncologica in Piemonte, AOU Città della salute
e della scienza, Torino
16Unità operativa di senologia, Azienda unità sanitaria locale, Bologna
17Centro di prevenzione oncologica, Azienda unità sanitaria locale, Ravenna
Corresponding author: Lauro Bucchi, lauro.bucchi@irst.emr.it
anno 39 (3) maggio-giugno 2015
e
&
p
This article is dedicated to the memory of Stefano Ciatto
53
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
GISMa position paper on interval cancers anno 39 (3) maggio-giugno 2015
e
&
p
Abstract
In this position paper, a self-convened team of experts from the Italian Group for Mammography Screening (Gruppo italiano
screening mammografico, GISMa) pointed out the problems that increasingly hamper the feasibility and validity of the estimate
of the proportional incidence of interval breast cancer (IBC) in Italy, suggested potential solutions and an agenda for research,
and proposed that the question of the sensitivity of mammography be viewed in a larger perspective, with a greater attention
to radiological review activities and breast radiology quality assurance programmes.
The main problems are as follows: the coverage of cancer registration is incomplete; the robustness of using the pre-screening
incidence rates as underlying rates decreases with time since the start of screening; the intermediate mammograms performed
for early detection purposes may cause an overrepresentation of IBCs; the classification of many borderline screening histories
is prone to subjectivity; and, finally, the composition of cohorts of women with negative screening results is uncertain, because
several mammography reports are neither clearly negative nor clearly positive, and because of the limitations and instability of
the electronic mammography records.
Several possibilities can be considered to cope with these issues: standard methods for using the hospital discharge records in the
identification of IBCs should be established; for the calculation of regional estimates of the underlying incidence, a suitable math-
ematical model should be identified; the definition of IBC according to the 2008 GISMa guidelines needs to be updated, especially
with respect to in situ cancers and to invasive cancers with borderline screening histories; a closer adherence to standard screen-
ing protocols, with a simplified patient management, would make it easier to objectively identify IBCs; alternative methods for es-
timating the sensitivity of mammography should be taken into consideration; and, finally, analysis could be restricted to the absolute
incidence rate of IBC, which would make comparison of the risk between neighbouring populations possible.
Epidemiologists must extend their attention to the prevention of the risk of IBC and the implementation of breast radiology qual-
ity assurance practices. Epidemiologists and radiologists can share common objectives: it is necessary to promote the idea that
the availability of a registry-based series of IBCs is not a prerequisite for their radiological review; radiological review of breast
cancers greater than 20mm in size detected at second and subsequent screens, that are potential substitutes for IBCs, needs ra-
diological and epidemiological validation studies; the advent of digital mammography brings about the possibility to create li-
braries of mammograms accessible online, which enables the conduct of large studies of the diagnostic variability of radiologists;
and, finally, epidemiologists and radiologists have the responsibility to monitor the effects that a loss of cumulative professional
experience in screening centres, due to the imminent retirement of a substantial proportion of healthcare workforce, could
cause on their performance.
(Epidemiol Prev 2015; 39(3) Suppl 1: 52-57)
Keywords: screening, mammography, quality assurance, breast cancer, interval cancer
Riassunto
In questo position paper, un team spontaneo di esperti associati al Gruppo italiano screening mammografico: (1) puntualizza i
limiti metodologici e i fattori distorsivi che compromettono la valutazione dell’incidenza dei cancri d’intervallo nei programmi di
screening in Italia, (2) suggerisce le possibili soluzioni e un’agenda per la ricerca, e (3) propone che il problema dei cancri d’in-
tervallo sia inserito in una prospettiva più ampia, con una maggiore attenzione per le attività di revisione radiologica e per i pro-
grammi di quality assurance in radiologia senologica.
(Epidemiol Prev 2015; 39(3) Suppl 1: 52-57)
Parole chiave: screening, mammografia, cancro della mammella, cancro d’intervallo
54
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
GISMa position paper on interval cancers anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
The sensitivity of mammography is a major factor for the ef-
fectiveness of a breast screening programme. The reference
method to evaluate the sensitivity of mammography is based
on the estimate of the proportional incidence of interval breast
cancer (IBC).
IBCs are cancers diagnosed after a negative mammography re-
sult and before next invitation to screening, or within two years
if the woman has reached the age for screening cessation. The
proportional incidence of IBC is the incidence observed dur-
ing the screening interval as compared to the incidence that
would be expected in the absence of screening, or underlying
incidence. This proportion gives an approximation of the rate
of mammography failures in abolishing the incidence of breast
cancer during the screening interval. In other words, the pro-
portional incidence of IBC is equal to 1 – the sensitivity of
mammography.
According to the 2008 guidelines from the Italian group for
mammography screening (Gruppo italiano screening mam-
mografico, GISMa),1the scientific society that gathers all pro-
fessionals involved in any aspect of mammography screening
in the country, the performance indicators of every screening
programme must include the absolute and proportional rates
of IBC, as well as the rate of IBCs interpreted to be visible on
retrospective radiological review. In the epidemiological guide-
lines chapter of the European guidelines for quality assurance in
breast cancer screening and diagnosis,2the estimate of the pro-
portional incidence of IBC is among the impact indicators, al-
though it is stated that it suffers from «several limitations».
This position paper originated from an initiative of members
of GISMa’s Coordinating Committee, who drafted a working
document and asked for amendments and proposals from epi-
demiologists and radiologists members of the society. The pa-
per aims at:
■
pointing out the problems that increasingly hamper the fea-
sibility and validity of the estimate of the proportional incidence
of IBC in Italy;
■
suggesting potential solutions and an agenda for research;
■
proposing that the question of IBC be viewed in a larger per-
spective, with a greater attention to radiological review activ-
ities and breast radiology quality assurance programmes.
The authors of this paper will submit a set of essential proposals
to the incoming Coordinating Committee of the GISMa.
PROBLEMS
The problems that affect the estimate of the proportional in-
cidence of IBC can be summarized as follows.
■
With respect to the identification of IBCs, the main limi-
tations are the incomplete coverage of cancer registration and
the delay – of a few years – by which the annual case series are
completed. The only available alternative is to create efficient
special breast cancer registries, whether based on standard meth-
ods of cancer registration or hospital discharge records. This
can also be done by the screening centres themselves. GISMa
guidelines accepted the use of hospital discharge records, al-
though they stated that developing standard methods was an
urgent need.1To this end, they proposed the formation of a
workgroup.
■
The robustness of using the pre-screening incidence rates as
underlying rates decreases with time since the start of the screen-
ing programme. It is unsafe both to assume that those rates,
if not modified by screening, would have been stable over time,
and to linearly extrapolate them to the present time.This lim-
itation is mentioned in the epidemiological guidelines chap-
ter of the European guidelines.2The 2008 GISMa guidelines
suggested the calculation and use of regional incidence esti-
mates.1These too were defined as an urgent need.
■
Intermediate mammograms performed at clinical radiolo-
gy facilities for early detection purposes may cause an over-
representation of IBCs. They lead to the detection of asymp-
tomatic cancers that cancer registries, if lacking information
on their actual clinical status, inevitably classify as IBCs. The
same may happen following intermediate mammograms ac-
tively offered within the screening programmes (early rescreen),
if they are recorded as diagnostic examinations rather than true
screening examinations. It is an epidemiological paradox that
the practice of performing intermediate mammograms, while
increasing the sensitivity of mammography for early breast can-
cer, causes apparently the opposite effect.
■
GISMa guidelines took into consideration the question of
whether the definition of interval cancer may include the can-
cers diagnosed during the third interval year or later, or after
a negative or an inconclusive assessment, or after a woman’s re-
fusal to undergo assessment, or after discontinuation of par-
ticipation in the programme, or after a previous diagnosis of
breast cancer.1The definition of IBC was expanded to include
some of these screening histories, but their interpretation in
a real-world screening setting remains prone to subjectivity.
■
Another source of variability is the eligibility of in situ breast
cancers, which is interconnected with the problem of their reg-
istration. GISMa guidelines suggested excluding in situ breast
cancers from the estimate of the proportional incidence of IBC,
given that they are incompletely registered and given their be-
nign and generally non-progressive behaviour. Nevertheless, the
guidelines recommended that interval in situ breast cancers
known to the screening centres be subject to radiological re-
view.1
■
Along with the diffusion of mammography screening into
widespread use, the procedure has become increasingly het-
55
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
GISMa position paper on interval cancers anno 39 (3) maggio-giugno 2015
e
&
p
erogeneous and complex. This change is connected to the emerg-
ing idea of an individually tailored screening.3One of the most
notable consequences of this is that the classic dichotomous clas-
sification of mammography results has been abandoned in cer-
tain screening programmes and in certain circumstances.
More and more often there are borderline mammography re-
ports that are neither clearly negative nor clearly positive. In the
estimate of the proportional incidence of IBC, this introduces
a degree of uncertainty both in the composition of cohorts of
women with negative screening results and in the detection mode
of incident breast cancers.
■
The composition of cohorts of women with negative screen-
ing results is also uncertain because of the limitations and in-
stability of the electronic mammography records. Screening cen-
tres are equipped with a variety of computer systems and soft-
wares. Many of these are designed solely for the delivery of the
service, not for the evaluation of results.
SOLUTIONS
Several possibilities can be considered to resolve these issues, at
least to a certain extent.
■
The GISMa guideline recommending that a workgroup be
appointed to establish standard methods for using the hospi-
tal discharge records in the identification of IBCs1should be
implemented. The workgroup can be comprised of those epi-
demiologists who are currently using the hospital discharge
records as a basis for registration.
■
As far as the underlying incidence rates are concerned, the
GISMa guideline recommending the calculation and use of re-
gional estimates1remains valid. It can be suggested to GISMa’s
Coordinating committee to formally ask the National centre
for screening monitoring (Osservatorio nazionale screening,
ONS) to examine the mathematical models that are being used
to estimate breast cancer incidence, and to select the most suit-
ed one.
■
Certain issues of the 2008 GISMa guidelines1need to be up-
dated. In particular, it would be advisable to re-examine the el-
igibility of in situ cancers and of invasive cancers diagnosed dur-
ing the third interval year or later, or after a negative or an in-
conclusive assessment, or after a woman’s refusal to undergo as-
sessment, or after discontinuation of participation in the pro-
gramme, or after a previous diagnosis of breast cancer. More-
over, the chapter on the definition of IBC should include a def-
inition of what a negative mammography result is, taking the
problem of borderline screening histories into consideration.
Epidemiologists with previous experience in the classification
of IBC detection modes should compare their methods with
each other and with the radiologists’ point of view.
■
Theoretically, an option to objectively classify IBC detection
mode is to draw the attention of screening units to the op-
portunity of adhering more closely to standard screening pro-
tocols. A simplified patient management would make it easi-
er to identify IBCs and – no less important – limit the diffu-
sion of unplanned forms of individually tailored screening.This
could be coupled with an effort to standardize the nomencla-
ture used in mammography reports as well as their format, at
least on a regional scale.
■
An innovative approach to the evaluation of the sensitivity
of mammography, which is commonly referred to as the un-
biased set method,4is not to use estimates of the underlying
incidence nor pre-screening incidence rates.The method requires
the availability of a general or a special cancer registry and of
information on the detection mode of registered breast cancers.
However, it uses only screen-detected cancers (except those de-
tected in the prevalence screen) and IBCs.The method was ex-
plicitly proposed for screening programmes of long duration,
which is the case for most programmes in Italy. It could be sug-
gested to GISMa as well as ONS to consider adopting the un-
biased set method as a reference method.
■
A minimalist approach to the evaluation of the incidence of
IBC, which has already been advised by European guidelines,2
would be to restrict analysis to the absolute incidence rate. On
the one hand, this would mean neglecting the estimate of mam-
mography sensitivity. On the other hand, however, it would al-
low the risk of IBC to be compared between neighbouring pop-
ulations (for example, those living in different health care dis-
tricts of an administrative region) who can be assumed to have
the same underlying breast cancer incidence. This would also
provide radiologists with a practical self-evaluation tool.
■
Until workable and effective solutions are found, the limi-
tations in estimating the proportional incidence of IBC need
to be well understood across the health system.The present pa-
per aims at preventing the use of currently available estimates
for legal and administrative purposes.
■
The same caution should be used in public communication
concerning the harms of mammography screening, which is rec-
ommended by European guidelines.5In the presentation of
screening programmes (public advertising campaigns and in-
vitation letters), information on false-negative mammography
results is insufficient. However, the information material
should simply state that false-negative results are possible, and
should describe the radiology facility characteristics that may
influence the accuracy of diagnosis (for example, the range of
annual screening mammogram reading volume of local radi-
ologists). Numerical estimates of the sensitivity of mammog-
raphy, which are poorly reliable and difficult to communicate,
must be avoided.
■
Lastly, we suggest a change in the scientific paradigm that has
so far underlain IBC evaluation. GISMa guidelines recommend
not only to estimate the proportional incidence of IBC, but also
to retrospectively review the mammograms.1More attention
and resources should be devoted to the reviewing process. The
value of radiological review, both for quality assurance and con-
tinuing education purposes, is repeatedly emphasized by Eu-
ropean guidelines.2,6
■
Approaching the problem of IBCs from the perspective of
breast radiology quality assurance would give practical imple-
mentation to a 2008 document from the Ministry of Health
(Direzione generale della prevenzione sanitaria del Ministero
della salute) in which it was stated that the registration of IBCs
should be accompanied by actions aimed at increasing the lev-
els of quality of the screening process.7The document suggested
that the review process be done in a semi-informed manner,
which has a greater educational impact in that it focuses on cri-
teria for women’s recall and not on medico-legal evaluations.
For medico-legal purposes, the reference method is a blinded
review of mammograms performed by a group of non-informed
expert consultants from a national certified register.
PERSPECTIVES
Epidemiologists must extend their attention to the prevention
of the risk of IBC, and consider that the implementation of
breast radiology quality assurance practices can be a common
point of interest with radiologists. The proportional incidence
of IBC, which is generally calculated at the screening pro-
gramme level and not at the single radiologist level, does not
provide clues to improve the sensitivity of mammography,
because it has no specific feedback on the diagnostic per-
formance. Conversely, radiological review of IBCs has a direct
educational impact.8The estimate of the proportional inci-
dence of IBC is a good example of how the descriptive epi-
demiology of cancer provides valuable information about the
size of problems, but often without the capacity to make a real
contribution to cope with them.9Unfortunately, in the long
run, a descriptive work that fails to promote appropriate actions
loses its rationale.
Following this line of reasoning, it must be noted that the an-
nual GISMa surveys of results of mammography screening in
Italy have shown for years a situation where the recall at sec-
ond and subsequent screens is above the acceptable standard
of 5% for one-third of local programmes, and where the prac-
tice of performing intermediate mammograms is widespread.10
This would require regular training programmes that are cur-
rently insufficient.
The following are some suggestions on how epidemiologists
and radiologists can interact positively and fruitfully.
■
The absence of radiological review activities in those screen-
ing centres that are served by a general or a special cancer reg-
istry is an unacceptable situation, in addition to being an orig-
inal type of underuse of cancer registration.9Where this occurs,
epidemiologists and radiologists should work together to find
a solution.
■
At the same time, both epidemiologists and radiologists should
promote the idea that the availability of a complete, registry-
based series of IBCs is a prerequisite only for estimating their
proportional incidence, not for their radiological review. Be-
sides, this should be done as soon as an IBC is detected or be-
comes known to the screening centre.
■
GISMa guidelines1and a study from the screening unit of
Trento11 have supported the radiological review of breast can-
cers greater than 20 mm in size detected at second and subse-
quent screens. GISMA’s Workgroup on diagnosis (Gruppo di
lavoro area diagnosi) has proposed, in particular, that these can-
cers be used as substitutes for IBCs in radiological review ac-
tivities at those screening centres where reviewing IBCs is prob-
lematic.12 The radiological review of screen-detected breast can-
cers greater than 20 mm in size is potentially feasible on a na-
tional scale and would make it possible to set up true nation-
al standards for all screening programmes in the country. This
approach, however, requires radiological and epidemiological
validation studies.
■
Screening units that have already estimated the proportion-
al incidence of IBC should be encouraged to determine
whether, in their data, there is a relationship between the sen-
sitivity of mammography and the prevalence of breast cancers
greater than 20 mm in size detected at second and subsequent
screens.13
56
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
GISMa position paper on interval cancers anno 39 (3) maggio-giugno 2015
e
&
p
57
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125 NATIONAL CENTRE FOR SCREENING MONITORING 11TH REPORT
GISMa position paper on interval cancers anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Ciatto S, Naldoni C, Ponti A et al. Interval cancers as indicators of per-
formance in screening programmes. Epidemiol Prev 2008;2:93-98.
2. Broeders M, Nyström L, Ascunce N et al. Epidemiological guidelines
for quality assurance in breast cancer screening. In: Perry N, Broed-
ers M, de Wolf C, Törnberg S, Holland R, von Karsa L (eds). Euro-
pean guidelines for quality assurance in breast cancer screening
and diagnosis. 4th edition. Office for Official Publications of the Eu-
ropean Communities, Luxembourg, 2006:15-56.
3. Paci E, Mantellini P, Giorgi Rossi P et al. Tailored Breast Screening
Trial (TBST). Epidemiol Prev 2013;4-5:317-27.
4. Zorzi M, Guzzinati S, Puliti D, Paci E. A simple method to estimate
the episode and programme sensitivity of breast cancer screening
programmes. J Med Screen 2010;17(3):132-38.
5. Giordano L, Webster P, Segnan N, Austoker J. Guidance on breast
screening communication. In: Perry N, Broeders M, de Wolf C,
Törnberg S, Holland R, von Karsa L (eds). European guidelines for
quality assurance in breast cancer screening and diagnosis. 4th edi-
tion. Office for Official Publications of the European Communities,
Luxembourg, 2006:379-94.
6. Rosselli del Turco M, Hendriks J, Perry N et al. Radiological guide-
lines for quality assurance in breast cancer screening. In: Perry N,
Broeders M, de Wolf C, Törnberg S, Holland R, von Karsa L (eds).
European guidelines for quality assurance in breast cancer screen-
ing and diagnosis. 4th edition. Office for Official Publications of the
European Communities, Luxembourg, 2006: 181-95.
7. Ministero della salute, Direzione generale della prevenzione sani-
taria. Programma di screening mammografico: significato e ges-
tione dei cancri d’intervallo. Seconda Edizione. Roma, 2008.
[http://www.senologiasirm.org/index.php?option=com_content&
view=article&id=12&Itemid=15].
8. Houssami N, Irwig L, Ciatto S. Radiological surveillance of interval
breast cancers in screening programmes. Lancet Oncol 2006;
7(3):259-65.
9. Armstrong BK. The role of the cancer registry in cancer control.
Cancer Causes Control 1992;3(6):569-79.
10. Giorgi D, Giordano L, Ventura L et al. Lo screening mammografico
in Italia: survey 2010. Epidemiol Prev 2012;6(Suppl.1):8-27.
11. Ciatto S, Bernardi D, Pellegrini M et al. Proportional incidence and
radiological review of large (T2+) breast cancers as surrogate in-
dicators of screening programme performance. Eur Radiol 2012;
22(6):1250-54.
12. Verbale Gruppo di lavoro area diagnosi, Torino 2013, Convegno
Nazionale GISMa.
[http://www.gisma.it/index.php?view=article&catid=38%3A
gruppi-di-lavoro&id=263%3Averbale-gruppo-di-lavoro-area-
diagnosi&option=com_content&Itemid=81].
13. Zorzi M, Fedato C, Baracco S. Confronto tra pT2+ agli esami suc-
cessivi e cancri intervallo per la stima della sensibilità. Convegno
Nazionale GISMa 2013: Sessione poster.
[http://www.gisma.it/index.php?option=com_content&view=arti-
cle &id=258:convegno-nazionale-gisma-2013-sessione-poster-&
catid=39:atti-dei-convegni&Itemid=129].
14. Scott HJ, Gale AG. Breast screening: PERFORMS identifies key
mammographic training needs. Br J Radiol 2006;79(2):S127-33.
■
The advent of digital mammography has brought about the
possibility to create, in conjunction with central radiological re-
view activities, libraries of mammograms accessible online. Al-
though there remains the problem of obtaining the informed
consent of patients, radiologist access to reviewed mammograms
would represent an important opportunity for research and train-
ing. Online libraries could be completed with images representing
a larger spectrum of mammographic abnormalities. Epidemi-
ologists could contribute to these developments by designing
studies of the radiologist variability in interpretation of mam-
mography findings.
■
Between mid-2013 and early 2014, GISMa’s Coordinating
committee carried out a national questionnaire survey of ra-
diologist’s experience-related characteristics (for example, an-
nual screening mammogram reading volume, and the percentage
of working time devoted to breast radiology).The survey is par-
ticularly topical given that the budget constraints that the Ital-
ian National Health Service is facing may lead to increasing flex-
ibility of mammogram-reading teams, as has been reported from
the United Kingdom.14 The data from the survey will have to
be evaluated both from a radiological and an epidemiological
point of view, because they could support the hypothesis that
radiologist’s experience-related characteristics are associated with
current results of local screening programmes.10
■
In the 1980s, the implementation of the new National Health
Service, coupled with an aggressive policy of deficit spending,
originated a dramatic increase in the number of medical and
paramedical staff in the public sector. The imminent retirement
of this workforce makes it urgent to assess the professional and
training needs of screening centres. The basic role of high-lev-
el specific training would suggest that part of the resources cur-
rently devoted to mammography screening programmes be al-
located to the creation and maintenance of a limited number
of multidisciplinary national training centres, following the ex-
perience of other European countries. Epidemiologists are re-
sponsible, in collaboration with radiologists, for monitoring the
effects that a loss of cumulative professional experience in screen-
ing centres could cause on their performance.
Conflicts of interests: none declared
Cervical cancer
screening
60 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Theoretical or potential or nominal extension: percentage of women involved in a screening programme out of the
total female population in the 25-64 age range resident in the area covered by an organized screening programme.
Actual extension or Extension of invitations: percentage of women involved in a screening programme out of the
total female population in the 25-64 age range who actually received an invitation to screening during the analyzed
period.
Compliance with invitation or Attendance: percentage of women attending screening out of invited women.
Referral rate: percentage of women referred to colposcopy (for any reason) out of the total number of screened women.
Recommendation to repeat cytology: percentage of women recommended to repeat cytology out of the total number
of screened women.
Compliance to recommendation to repeat cytology: percentage of women who actually repeated cytology among
those who were recommended to do so.
Compliance to colposcopy for ASCUS+: percentage of women who underwent colposcopy out of women referred to
colposcopy because of ASCUS or more severe cytology.
Compliance to colposcopy for HSIL+: percentage of women who underwent colposcopy out of women referred to
colposcopy because of HSIL or more severe cytology.
Positive predictive value (PPV) of referral to colposcopy because of ASCUS+ cytology for histologically
confirmed CIN2+: proportion of women with histologically confirmed CIN2+ out of women referred to colposcopy
because of ASCUS or more severe cytology.
Detection rate (DR) CIN2+ unadjusted: number of women who had a CIN2+ detected out of 1,000 screened women.
Detection rate (DR) CIN2+ stand. Ita.: number of women who had a CIN2+ detected out of 1,000 screened women,
adjusted for age in 5-year groups on the Italian population.
Piemonte: Nereo Segnan, Centro prevenzione oncologia – CPO
Piemonte, Torino
Provincia Autonoma di Bolzano (Alto Adige): Antonio Fanolla,
Assessorato alla sanità, Osservatorio epidemiologico, Provincia
autonoma di Bolzano, Bolzano
Provincia Autonoma di Trento (Trentino): Sivano Piffer, Giovanni
De Pretis, Osservatorio epidemiologico, Azienda provinciale
per i servizi sanitari, Trento
Puglia: Vincenzo Pomo, Cinzia Annatea Germinario, Agenzia
regionale sanità, Regione Puglia, Bari
Sardegna: Pierina Thanchis, Assessorato dell’igiene e sanità
e dell’assistenza sociale, Regione Sardegna, Cagliari
Toscana: Paola Mantellini, Istituto per lo studio e la prevenzione
oncologica, Regione Toscana, Firenze
Umbria: Mariadonata Giaimo, Direzione regionale salute, coesione
sociale e società della conoscenza, Regione Umbria, Perugia
Valle D’Aosta: Gabriella Furfaro, Servizio dipendenze patologiche,
salute mentale e promozione della salute, Aosta
Veneto: Chiara Fedato, Registro tumori del Veneto, Padova
Abruzzo: Tamara Agostini, Direzione politiche della salute, Regione
Abruzzo, Pescara
Basilicata: Vincenzo Barile, Angelo Sigillito, Sergio Schettini, AO
San Carlo, Potenza
Calabria: Liliana Rizzo, Dipartimento Tutela della salute e politiche
sanitarie, Regione Calabria, Catanzaro
Campania: Renato Pizzuti, Osservatorio epidemiologico regionale,
Assessorato alla sanità, Regione Campania, Napoli
Emilia-Romagna: Carlo Naldoni, Assessorato alle politiche
per la salute, Regione Emilia-Romagna, Bologna
Friuli-Venezia Giulia: Nora Coppola, Direzione centrale salute,
integrazione socio sanitaria, politiche sociali e famiglia, Regione
Friuli-Venezia Giulia, Trieste
Lazio: Alessandra Barca, Lazio sanità, Agenzia di sanità pubblica,
Roma
Liguria:Luigina Bonelli, Gabriella Paoli, Istituto nazionale
per la ricerca sul cancro, Genova
Lombardia: Direzione generale salute, Regione Lombardia, Milano
Marche: Lucia Di Furia, Servizio salute, Regione Marche, Ancona
Molise: Ospedale Cardarelli, Regione Molise, Campobasso
Italian cervical cancer screening survey group:
Gruppo di lavoro italiano per la survey sullo screening cervicale:
Glossary
Cervical cancer screening: 2011-2012 acivity
61 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
WWW.EPIPREV.IT
Extension of organized cervical cancer
screening programmes in Italy and their
process indicators, 2011-2012 activity
Estensione dei programmi organizzati di screening
cervicale in Italia e loro indicatori di processo
Guglielmo Ronco,1Pamela Giubilato,1Francesca Carozzi,2Giovanni Maina,3Paolo-Giorgi-Rossi,4,5
Marco Zappa2and the Italian cervical cancer screening survey group
Abstract
Italian national guidelines recommend regional implementation of organized screening programmes for
cervical cancer. As we have been doing since 1998, we collected aggregated tables of data from Italian
organized cervical screening programmes in order to centrally compute process indicators. Data on
women invited during 2011 and 2012 and screened up to April of the subsequent year were considered.
In 2012, the target population of Italian organized screening programmes included 14,497,207
women, corresponding to 87.3% of Italian women aged 25-64 years.
Compliance to invitation was 41.2% in 2011 and 40.8% in 2012, with a strong decreasing North-South
trend. However, it should be considered that many women are screened outside any organized pro-
grammes. In 2012, of the women screened, 3.5% were referred for repeat cytology and 71.1% of
them complied; 2.4% of screened women were referred to colposcopy.
Compliance with colposcopy referral was 85.3% among women referred because of ASC-US or more
severe cytology and 90.4% among those referred because of HSIL or more severe cytology. The pos-
itive predictive value (PPV) of referral because of ASC-US or more severe cytology for CIN2 or more
severe histology was 16.9%. The unadjusted detection rate of CIN2 or more severe histology was 3.4
per 1,000 screened women (3.6 standardized on the Italian population, truncated 25-64). CIN2 or more
severe histology was detected in 64.6% of colposcopies classified as grade 2 or higher. Of all colpo-
scopies during which a CIN2 or more severe histology was obtained, 33.6% were classified as grade
2 or higher. Follow-up only was recommended to 81.7% of women with CIN1.
Excision by radio-frequency device was the most common treatment for women with CIN2 (52.8%)
and CIN3 (57.0%). However 0.4% of all CIN2 and 2.3% of all CIN3 had hysterectomy.
(Epidemiol Prev 2015; 39(3) Suppl 1: 61-76)
Keywords: cervical cancer, Pap test, colposcopy, mass screening, Italy
1AOU Città della salute
e della scienza,
CPO Piemonte, Torino
2Istituto per lo studio
e la prevenzione
oncologica (ISPO), Firenze
3AOU Città della salute
e della scienza,
Ospedale S. Anna, Torino
4Servizio interaziendale
di epidemiologia, AUSL
Reggio Emilia
5IRCCS-Arcispedale
S. Maria Nuova,
Reggio Emilia
Corresponding author:
Guglielmo Ronco
guglielmo.ronco@cpo.it
Riassunto
Le linee guida nazionali italiane raccomandano alle Regioni di attivare programmi organizzati di
screening per il cervicocarcinoma. Come negli anni precedenti, a partire dal 1998, dai programmi or-
ganizzati italiani di screening cervicale si sono raccolte tabelle aggregate di dati per calcolare central-
mente indicatori di processo. Si sono considerati i dati delle donne invitate nel corso del 2009 e scree-
nate fino ad aprile 2011.
Nel 2012 i programmi organizzati italiani includevano nella loro popolazione obiettivo 14.497.207
donne, corrispondenti all’87,3% delle donne italiane di età 25-64 anni. La compliance all’invito è stata
Cervical cancer screening: 2011-2012 acivity
62 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
The Italian health system is managed by Italy’s 20 regions. Since
1996, Italian national guidelines have recommended to regions
to implement organized screening programmes for cervical can-
cer.1-3 Recommendations, largely based on European guide-
lines,4,5 include personal invitations to women aged 25 to 64
years for a Pap test every three years, a monitoring system, and
quality assurance for each phase of the programme.
Surveys designed to assess the level of implementation of or-
ganized programmes in Italy and to collect process indicators
have been conducted by GISCi (Italian group for cervical
screening) since 1997. Their results have been evaluated and
published by the ONS (Osservatorio nazionale screening, Na-
tional centre for screening monitoring), on behalf of the Ital-
ian Ministry of Health, since 2002.6-15 Diagnostic work-up
and particularly treatment have also been monitored in order
to reduce under- and over-treatment.
A number of programmes moved to HPV-based screening tests
as pilot projects or as routine activity after the recommendation
of the national Ministry of Health.16 Detailed data on HPV-
based screening are presented elsewhere.17
In the present report, data on coverage and compliance and data
on second-level activities included all women, independently of
the primary screening test. Conversely, process indicators for
first-level tests include only women screened with Pap smears.
METHODS
Surveys of organized cervical screening programmes active in
Italy in 2011 and 2012 were conducted by the ONS on behalf
of the Italian Ministry of Health in 2012 and 2013. A pro-
gramme was considered active each year if at least 1,000
women were invited during that year. For each year, women in-
vited during that year and screened within the first 4 months
of the subsequent year were considered.
Given the different approaches to integration of invitations and
spontaneous activity, some programmes reported data only on
women screened after invitation and others on all screened
women, independently of invitation. In the latter case, data on
spontaneous activity included women screened during the rel-
evant year.
We collected tables of aggregated data, in general nested,so that
each table was the denominator of the next. They were used to
centrally compute process indicators (most of those recom-
mended by Italian2,3 and European5guidelines) and to study
their distribution. Data were centrally checked for complete-
ness and consistency. Each region appointed a person to pro-
vide data and finally verify them. We interacted, sometimes re-
peatedly, with providers, to obtain clarifications and
integrations, if needed.
For each indicator we computed the national overall mean, i.e.,
the value obtained by pooling all the population for which data
were available. In addition, we analyzed the distribution of in-
dicators between regions and between local programmes within
each region.
“Programme” is defined as each entity for which we obtained ag-
gregated data. In general, according to national guidelines,1-3 this
corresponds to an organizational unit that manages and co-or-
dinates the different steps of screening, from invitation to diag-
nostic assessment and treatment. These units are generally well
defined, but sometimes they underwent re-organization (typi-
cally, aggregation of smaller programmes). Furthermore, their
size is highly variable. For example, in some regions there is a sin-
gle programme (e.g., Basilicata and Friuli) while others have
many local programmes with regional co-ordination and eval-
uation (e.g., Piemonte, Veneto, Emilia-Romagna, Toscana).
We report (table 3, p. 66) the mean national value, of some in-
dicators and their 10th and 90th percentile. The values of the
last three surveys are reported. The year denotes the period of
screening activity considered (therefore the year before the con-
duction of the survey). In addition, we present graphs with the
distribution between regions in 2011 and 2012. Figures 2 (p.
67) and 5(p. 68) report the mean for 2011 and 2012.
Data on second-level activities (about correlation between col-
poscopic findings and histology and about the management of
women with screen-detected CIN or invasive cancer) are pre-
sented at an overall national level as tables including data
from all programmes that provided them in 2010 and 2011.
Colposcopic findings were classified according to the Interna-
tional classification (IFCPC).The Rome 1990 classification18
was adopted in the first experimental surveys and kept in use
41,2% nel 2011 e 40,8% nel 2012, con un deciso trend a diminuire da Nord a Sud. Bisogna comunque ricordare che molte donne
vengono screenate al di fuori dei programmi organizzati.
E’ stato raccomandato di ripetere la citologia al 3,5% delle donne e il 71,1% di esse l’ha fatto. Il 2,4% delle donne screenate è
stato inviato in colposcopia. La compliance alla colposcopia è stata 85,3% tra le donne inviate per citologia ASC-US o più grave e
90,4% tra quelle inviate per citologia HSIL o più grave. Il valore predittivo positivo (VPP) dell’invio in colposcopia per citologia ASC-
US o più grave per istologia CIN2 o più grave è stato 16,9%. La detection rate (DR) grezza di istologia CIN2 o più grave è stata 3,4
ogni 1.000 donne screenate (3,6 quella standardizzata sulla popolazione italiana, troncata 25-64). Nel 64,6% delle colposcopie clas-
sificate come di grado 2 o più elevato l’esame istologico ha dato un responso CIN2 o più grave. Tra tutte le colposcopie con isto-
logia CIN2 o più grave, il 33,6% è stato classificato come di grado 2 o più elevato. All’81,7% delle donne con esito CIN1 si è con-
sigliato il follow-up. L’escissione con radiofrequenza è stato il trattamento più comune per donne con istologia CIN2 (52,8%) e CIN3
(57,0%). Lo 0,4% delle donne con istologia CIN2 e il 2,3% di quelle CIN3 ha avuto un’isterectomia.
(Epidemiol Prev 2015; 39(3) Suppl 1: 61-76)
Parole chiave: cancro cervicale, Pap test, coloscopia, screening di massa, Italia
for comparability. In this section each colposcopy was consid-
ered as a statistical unit. In case of multiple biopsies during a
same colposcopy, the most severe histology was considered. In
the section on management of women with screen-detected
CIN/cancer each woman was a unit. For this purpose we con-
sidered the worst histology before treatment and the first treat-
ment. A “see and treat” approach – i.e., treatment in the ab-
sence of a histological diagnosis – is very limited in Italian
organized programmes.
RESULTS
Extension of organized programmes
and invitation of the target population
Concerning this section of the survey, we obtained question-
naires from 116 and 119 programmes for 2011 and 2012, re-
spectively. The target population of active organized pro-
grammes in these and previous surveys is reported in table 1
(p. 64). Target populations are also expressed as the percent-
age of women aged 25 to 64 years resident in a given area. It
must be kept in mind that denominators are based on census
for 2012 and estimated for previous years.
Active programmes in Italy had a target population of
14,301,979 women in 2011 and 14,497,209 women in 2012,
representing 84.1% and 87.3%, respectively, of the Italian fe-
male population aged 25-64, compared to 80.1% in 2010. In
2012, active programmes included in their target population
the entire female population aged 25 to 64 years in 15/21 re-
gions, over 95% in 3 regions and close to 80% in 2 regions
(Sardegna and Liguria). Incomplete nominal extension is
mainly caused by the choice of not implementing a population-
based screening in the region of Lombardia, where only local
initiatives are active (table 2, p. 65).
The values above consider the entire target population re-
gardless of the proportion actually invited. It is obviously rel-
evant that active programmes invite women at a rate sufficient
to reach the entire target population within the standard
screening interval (3 years for cytology-based screening). Table
1reports the ratio between the number of women invited dur-
ing each year and the number that should have been invited
in case of full implementation (i.e., 1/3 of the resident popu-
lation aged 25-64 years). In 2012, actual extension was 70.4%
at national level. Because variations between years can result
from local criteria of organization, the percentage of women
in the target population invited in the last 3 years is reported
in table 2.The completeness of invitation is also computed ex-
cluding from the denominator women not invited because of
recent testing or for other specified reasons (adjusted %). Pro-
grammes adopt different criteria for exclusion and some pro-
grammes do not exclude any women at all. There is a clear
North-South gradient in completeness of invitation. As the in-
terval between HPV-based screens is now 5 years, actual cov-
erage is now underestimated, but this effect is minimal for
2011 and 2012 activity.
During 2011 and 2012, 41.2% and 40.8%, respectively, of in-
vited women were screened, compared to 39.8% in the previ-
ous year (table 1). A clear decreasing trend in compliance
with invitation from northern to central and especially south-
ern Italy (49.1%, 40.2%, and 29.5%, respectively, in 2012) was
Cervical cancer screening: 2011-2012 acivity
63 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
%
60
70
50
40
30
20
10
0
Figure 1. Uptake percentage by region. 2011-2012 activity. / Figura 1. Adesione percentuale all’invito, per Regione. Attività 2011-2012.
Each bar represents one region, 2011-2012 activity.
Lines represent the range between 10th and 90th percentile of programme distrubution within each region
Basilicata 11
Basilicata 12
Liguria 11
Liguria 12
Campania 11
NORTH CENTRE SOUTH and ISLANDS
Campania 12
Sicilia 11
Sicilia 12
Lazio 11
Lazio 12
Puglia 11
Puglia 12
Molise 11
Molise 12
Trento 11
Trento 12
Abruzzo 11
Abruzzo 12
Calabria 11
Calabria 12
Marche 11
Marche 12
Bolzano 11
Bolzano 12
Lombardia 11
Lombardia 12
Sardegna 11
Sardegna 12
Umbria 11
Umbria 12
Piemonte 11
Piemonte 12
Toscana 11
Toscana 12
Veneto 11
Veneto 12
Emilia-Romagna 11
Emilia-Romagna 12
Aosta 11
Aosta 12
Friuli 11
Friuli 12
2006 2007 2008 2009 2010 2011 2012
Number of women 25-64 yrs. included 11,362,580 11,872,810 13,094,025 13,133,604 13,538,080 14,301,979 14,497,209
in the target population
of organized programmes
Population 25-64 yrs 16,463,948* 16,543,059* 16,693,052* 16,812,052* 16,900,554* 17,006,946* 16,600,566**
Nominal extensiona69.01 71.77 78.44 78.12 80.10 84.10 87.33
Actual extensionb52.91 54.80 59.85 63.30 63.64 62.19 70.37
(2,873,202/ (3,021,734/ (3,330,289/ (3,547,457/ (3,584,955/ (3,525,522/ (3,893,773/
5,487,982) 5,514,353) 5,564,350) 5,604,016) 5,633,511) 5,668,982) 5,533,522)
Compliance to invitation (%)c38.49 39.83 39.69 39.27 39.84 41.17 40.76
(1,116,006/ (1,217,000/ (1,332,376/ (1,393,243/ (1,374,745/ (1,451,056/ (1,600,796/
2,899,817) 3,055,353) 3,356,931) 3,547,457) 3,450,755) 3,524,863) 3,927,403)
Percentage of population screenedd20.34% 22.07% 23.94% 24.86% 24.40% 25.60% 28.93%
(1,116,006/ (1,217,000/ (1,332,376/ (1,393,243/ (1,374,745/ (1,451,056/ (1,600,796/
5,487,983) 5,514,353) 5,564,351) 5,604,017) 5,633,518) 5,668,982) 5,533,522)
Northern Italy
Number of women 25-64 yrs. included 4,911,641 4,942,788 5,210,405 5,133,658 5,155,376 5,513,736 5,590,488
in the target population
of organized programmes
Population 25-64 yrs 7,545,425 7,555,407 7,615,828 7,674,160 7,712,312 7,771,110 7,564,052
Nominal extensiona65.09 65.42 68.42 66.90 66.85 70.95 73.91
Actual extensionb52.91 55.38 55.38 59.75 60.32 62.51 69.25
(1,330,768/ (1,394,613/ (1,525,113/ (1528455/ (1,550,770/ (1,619,150/ (1,745,942/
2,515,141) 2,518,469) 2,538,609) 2558053) 2,570,768) 2,590,370) 2,521,348)
Compliance to invitation (%)c45.62 46.93 47.67 49.15 49.39 49.87 49.12
(612,069/ (664,344/ (734,577/ (751,283/ (742,219/ (815,607/ (867,589/
1,341,812) 1,415,361) 1,541,010) 1,528,455) 1,502,820) 1,635,630) 1,766,270)
Percentage of population screenedd24.34% 26.38% 28.94% 29.37% 28.87% 31.49% 34.41%
(612,069/ (664,344/ (734,577/ (751,283/ (742,219/ (815,607/ (867,589/
2,515,142) 2,518,469) 2,538,609) 2,558,053) 2,570,771) 2,590,370) 2,521,351)
Central Italy
Number of women 25-64 yrs. included 3,029,340 3,008,931 3,252,167 3,113,448 3,277,736 3,308,299 3,246,268
in the target population
of organized programmes
Population 25-64 yrs 3,224,341 3,275,594 3,315,532 3,347,197 3,367,589 3,391,992 3,283,420
Nominal extensiona93.95 91.86 98.09 93.02 97.33 97.53 98.87
Actual extensionb75.05 74.54 80.51 80.26 80.62 79.42 81.48
(806,609/ (813,887/ (889,801/ (895,459/ (904,993/ (897,918/ (891,778/
1,074,780) 1,091,865) 1,105,177) 1,115,732) 1,122,528) 1,130,664) 1,094,473)
Compliance to invitation (%)c35.70 40.23 40.17 38.12 37.98 38.52 40.18
(290,632/ (330,925/ (357,846/ (341,325/ (327,029/ (346,654/ (358,958/
814,208) 822,548) 890,868) 895,459) 860,981) 899,824) 893,437)
Percentage of population screenedd27.04% 30.31% 32.38% 30.59% 29.13% 30.66% 32.80%
(290,632/ (330,925/ (357,846/ (341,325/ (327,029/ (346,654/ (358,958/
1,074,780) 1,091,865) 1,105,177) 1,115,732) 1,122,530) 1,130,664) 1,094,473)
Southern Italy and Islands
Number of women 25-64 yrs. included 3,421,599 3,921,091 4,631,453 4,886,498 5,104,968 5479944 5,660,453
in the target population
of organized programmes
Population 25-64 yrs 5,694,182 5,712,058 5,761,692 5,790,695 5,820,653 5,843,844 5,753,109
Nominal extensiona65.63 68.65 80.38 84.39 87.70 95.39 98.39
Actual extensionb38.77 42.71 47.66 58.21 58.20 51.77 65.50
(735,825/ (813,234/ (915,375/ (1,123,543 (1,129,192/ (1,008,454/ (1,256,053/
1,898,060) 1,904,019) 1,920,564) /1,930,231) 1,940,215) 1,947,948) 191,7701)
Compliance to invitation (%)c28.68 27.12 27.73 26.76 28.11 29.19 29.52
(213,305/ (221,731/ (239,953/ (300,635/ (305,497/ (288,795/ (374,249/
743,797) 817,444) 925,053) 1,123,543) 1,086,954) 989,409) 1,267,696)
Percentage of population screenedd11.24% 11.65% 12.49% 15.58% 15.75% 14.83% 19.52%
(213,305/ (221,731/ (239,953/ (300,635/ (305,497/ (288,795/ (374,249/
1,898,061) 1,904,019) 1,920,564) 1,930,232) 1,940,218) 1,947,948) 1,917,703)
apercentage of the resident 25-64 year-old population that is included in the target population of active organized programmes.
bnumerator: population invited in the relevant year; denominator: 1/3 of the resident population aged 25-64 (invited women include both those invited for cytology and those invited for
HPV testing as primary screening test).
cdenominator: number of women invited; numerator: number of women who underwent screening among them (by the first 4 months of the following year).
dnumerator: number of women who underwent screening among invited women (by the first 4 months of the following year); denominator: 1/3 of the resident 25-64 year-old population.
* estimated by the National institute of statistics (Istat).
**obtained by census.
Cervical cancer screening: 2011-2012 acivity
64 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Table 1. Target population of active organized screening programmes in Italy,population invited and compliance to invitation.
Tabella 1. Popolazione obiettivo dei programmi organizzati di screening cervicale in Italia, quota di donne invitate e donne che hanno effettivamente risposto.
Cervical cancer screening: 2011-2012 acivity
65 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Region Programmes active in 2012 Target Nominal Target Target Target Adjusted
polpulation extension population population population target
25-64 yrs invited invited* invited* population
invited**
(%) (%) (%) (%) (%) (%)
(2012) (2012) (2011) (2012) (2010+ (2010+
2011+2012) 2011+2012)
Valle d'Aosta Single regional programme 35,777 100 29.6 30.4 96.59 96.59
Piemonte Regional programme. Fully activea1,206,933 100 27.8 30.1 87.51 87.51
Città di Torino, Cuneo, Alessandria, Moncalieri,
Rivoli, Ivrea, Biella-Vercelli, Novara, Asti
Liguria Regional programme. 336,105 79.1 6.9 11.0 28.34$
Genova 3, Imperia, Savonese
Lombardia Regional programme. The following are active: 778,096 28.7 28.1 31.6 87.31 100
Brescia. Cremona, Lodi, Mantova, Pavia,
Vallecamonica
Self-governing Single regional programme 145,719 100 46.2 63.9 100 100
province of Trento
Self-governing Single regional programme 137,647 100 24.3 23.6 71.78$
province of Bolzano
Veneto Regional programme. Fully activea1,353,553 100 28.2 32.1 87.52 100
Adria, Alta Padovana, Alto Vicentino, Asolo,
Bassano Del Grappa, Belluno, Bussolengo, Chioggia,
Este, Feltre, Legnago, Dolo Mirano, Padova, Vicenza
Ovest Vicentino,Verona, Pieve Di Soligo, Rovigo,
Treviso, Veneto Orientale, Veneziana
Friuli-Venezia Giulia Single regional programme 343,353 100 28.4 30.3 83.70 100
Emilia-Romagna Regional programme. Fully activea1,255,986 100 35.6 34.2 100 100
Bologna, Cesena, Ferrara, Forlì, Imola, Modena,
Parma, Piacenza, Ravenna, Reggio Emilia, Rimini
Toscana Regional programme. Fully activea1,022,925 100 31.2 33.0 95.89 100
Arezzo, Empoli, Firenze, Grosseto, Livorno, Lucca,
Massa, Pisa, Pistoia, Prato, Siena,Viareggio
Umbria Single regional programme 265,114 100 24.2 37.7 83.82 99.48
Marche Regional programme. Fully activea422,224 100 31.7 31.7 93.98 100
Area vasta 1,Area vasta 2, Area vasta 3,
Area vasta 4,Area vasta 5
Lazio Regional programme. The following are active: 1,536,005 96.4 23.6 20.8 66.48 68.41
Latina, Rieti, Roma A, Roma B, Roma C, Roma D,
Roma E, Roma G, Roma H, Viterbo
Molise Single regional programme 85,637 100 13.1 14.0 53.60 53.67
Abruzzo Single regional programme. Fully activea368,882 100 15.9 29.5 61.05 100
Campania Regional programme. The following are active:
Avellino, Benevento, Caserta, Napoli 1, Napoli 2,
Napoli 3, Salerno 1,624,086 100 14.7 15.7 46.15 54.12
Basilicata Single regional programme 167,348 100 --100 100
Calabria Catanzaro, Cosenza, Lamezia Terme, Locri, Palmi,
Reggio Calabria, Vibo Valentia 530,517 97.7 19.6 14.8 57.44 60.25
Sicilia Regional programme. The following are active:
Agrigento, Catania, Caltanissetta, Enna, Messina,
Palermo, Ragusa, Siracusa, Trapani 1,375,898 99.3 19.2 30.0 78.25 78.27
Sardegna Regional programme. The following are active:
Cagliari, Carbonia, Nuoro, Olbia, Oristano, Sanluri 377,031 79.4 30.5 24.9 78.76 80.79
Puglia Single regional programme 1,131,054 100 14.0 18.3 50.43 34.5
afully active means that the entire regional female population aged 25-64 is included in the target population of active cervical screening programmes.
* only women aged 25-64 years are considered both in the numerator and denominator.
**numerator: women aged 25-64 years invited in the last 3 years. Denominator: target population aged 25-64 years minus women excluded before invitation because already invited or
due to other reason.
$active only for 2 years.
Table 2. Active organized cervical screening programmes and target population (age 25-64), by region.Years 2011-2012.
Tabella 2. Programmi organizzati di screening cervicale attivi e popolazione obiettivo (25-64 anni), per Regione.Anni 2011-2012.
present, as previously observed. In 2012, compliance was over
50% in Umbria, Valle d’Aosta, Friuli-Venezia Giulia, Emilia-
Romagna, and the province of Trento (figure 1).
Process indicators in organized
programmes
Data in this section include only women screened by cytology.
In 2011 and 2012, programmes that provided this type of data
were 107 and 104, while screened women were 1,508,959 and
1,467,808, respectively. Some programmes reported data only
on women screened after invitation. Decreases in number of
programmes and screened women are due to the increase in
HPV-based screening. Table 3 reports for each indicator the
number of programmes for which that indicator could be
computed.
In 2011 and 2012, 4.1% and 3.5% of screened women were
recommended to repeat cytology, compared to 4.7% in 2010
and values between 5% and 7% in 2005-2009. In 2012, in two
regions cytology repeat was recommended to more than 15%
of screened women and in three others to more than 6% (fig-
ure 2). In three of these regions, many repeats were due to
«other reasons», likely reactive changes, which represent a rel-
evant source of variability. Repeats for unsatisfactory smears
were very high in Molise. In some regions, a proportion of
women was recommended to repeat the smear after ASC-US,
AGC, and L-SIL cytology. However, these reasons represent a
substantial proportion of repeats only in Sardegna and Veneto.
Among women who had been recommended to repeat the
smear, 65% actually had a new one in 2011 and 71% in
2012, following a monotonously increasing trend (60% in
2007). In 2012 two regions were below 50% and seven were
above 80% (figure 3). These values do not take into account
that some women should have repeated cytology after a time
interval that had not ended when data were collected.
Cervical cancer screening: 2011-2012 acivity
66 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Year of activitya2010 2011 2012
N* Mean centile N* Mean centile N* Mean centile
(%) (%) (%) (%) (%) (%)
(num/den) 10th 90th (num/den) 10th 90th (num/den) 10th 90th
Population screenedb118 1,456,665 107 1,508,595 104 1,467,808
Recommendation to repeat 111 4.7 1.2 10.0 103 4.1 1.0 9.0 100 3.5 1.0 7.5
cytologyc(71,820 (59,934/ (51,674/
1,512,430 1,449,562) 1,467,808)
Compliance to recommendation 100 62.7 40.8 86.9 98 64.8 41.1 94.1 94 71.1 41.9 95.3
to repeat cytologyd(33,410 (34,591 (32,507/
53,288) 53,405) 45,691)
Referral ratee114 2.5 1.0 4.2 105 2.4 1.1 4.2 102 2.4 1.0 4.2
(36,647/ (36,525/ (36,432/
1,445,138) 1,492,349) 1,494,122)
Compliance to colposcopy 106 85.9 64.8 98,6 101 87.7 64.8 98.3 99 85,3 72.6 100
for ASC-US+f(29,725/ (30,115/ (25,510/
34,600) 34,346) 34,605)
Compliance to colposcopy 105 88.7 64,0 100 98 89.5 66,6 100 99 90.4 66.7 100
for HSIL+g(2,834/ (2,749/ (2,868/
3,194) 3,072) 3,172)
PPV of referral to colposcopy 102 16.0 6.4 28.3 95 15.3 5.2 29.0 92 16,9 5.8 31.1
because of ASC-US+ cytology (4,597/ (4,268 (4,724/
for histologically confirmed 28,723) 27,802 27,988)
CIN2+h
DR CIN2+ unadjustedi102 3.2 1.1 5.2 95 3.2 0.8 5.0 92 3.4 1.5 5.2
(4,597/ (4,268/ (4,741/
1,393,654) 1,323,390 1,393,544)
DR CIN2+ stand. Ita.j98 3.5 0.9 5.6 88 3.2 1.2 5.5 89 3.5 1.2 5.8
* number of programmes that provided information
ayear before the conduction of the survey; each survey includes women invited during the previous year and screening within the first 4 months of the current year (see text).
bin some programmes this includes only women screened after invitation, in others all screened women, independently of invitation (see text)
cdenominator: number of screened women; numerator: number of women recommended to repeat cytology.
ddenominator: total number of women recommended to repeat cytology; numerator: women who repeated within 15 April 2013.
edenominator: number of screened women; numerator: number of screened women referred to colposcopy (any reason).
fdenominator: number of women referred to colposcopy because of ASC-US or more severe citology; numerator: number of the latter who underwent colposcopy.
gdenominator: number of women referred to colposcopy because of H-SIL or more severe citology; numerator: number of the latter who underwent colposcopy.
hdenominator: number of women who underwent colposcopy because of ASC-US or more severe citology; numerator: number of the latter who had CIN2 or more severe detected (histo-
logically confirmed – most severe lesion within six months from cytology considered).
idenominator: number of screened women; numerator: number of the latter who had a CIN2+ detected (histologically confirmed – most severe lesion within six months from cytology con-
sidered). Cases per 1,000 screened women.
jsee (i); adjusted for age in 5-yeargroups on the Italian population (census 1991, truncated 25-64); the nationalmean was directly computed for the pool of all programmes with valid needed
data; percentiles were obtained after computing the standardized DR for each programme with valid required data.
Table 3. Value of some process indicators (national mean, 10th,and 90th percentile) in the last three surveys.
Tabella 3. Valore di alcuni indicatori di processo (media nazionale, 10° e 90° percentili) nelle ultime tre survey.
The referral rate to colposcopy was 2.4% both in 2011 and
2012 (table 3). Values between 2.3% and 2.5% had been reg-
istered in all years from 2005 to 2010.
The referral rate was above 4% in both 2011 and 2012 in Valle
d’Aosta and in 2011 in Molise and Basilicata (figure 4, p. 68).
There was a high variability within some regions. In 2012, out
Cervical cancer screening: 2011-2012 acivity
67 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Toscana Bolzano Emilia
Romagna
Marche Puglia Calabria Umbria Friuli Lombardia Veneto Lazio Campania Sicilia Sardegna
%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0
Figure 2. Percentage of screened women referred for repeat cytology, by region. 2011-2012 activity.
Figura 2. Percentuale della popolazione screenata che ha avuto indicazione a ripetere la citologia per qualsiasi causa, per Regione.Attività 2011-2012.
Piemonte Aosta Trento Molise Liguria
repeat unsatisfactory repeat ASC-US+AGC repeat L-SIL repeat ASC-H repeat other
%
60
70
50
40
30
20
90
100
80
10
0
Figure 3. Compliance with repeat cytology.Women who repeated cytology by 15 April 2012 and by 15April 2013 out of all those referred for repeat cytology. 2011-
2012 activity.
Figura 3. Compliance alla ripetizione della citologia. Donne che hanno ripetuto entro il 15 aprile 2012 ed entro il 15 aprile 2013 su tutte le donne che hanno avuto in-
dicazione a ripetere.Attività 2011-2012.
Each bar represents one region, 2011-2012 activity.
Lines represent the range between 10th and 90th percentile of programme distribution within each region.
Molise 11
Molise 12
Aosta 11
Aosta 12
Calabria 11
NORTH CENTRE SOUTH and ISLANDS
Calabria 12
Piemonte 11
Piemonte 12
Basilicata 11
Umbria 11
Umbria 12
Campania 11
Campania 12
Sardegna 11
Sardegna 12
Trento 11
Trento 12
Lazio 11
Lazio 12
Veneto 11
Veneto 12
Toscana 11
Toscana 12
Marche 11
Marche 12
Lombardia 11
Lombardia 12
Emilia-Romagna 11
Emilia-Romagna 12
Sicilia 11
Sicilia 12
Bolzano 11
Bolzano 12
Puglia 11
Puglia 12
Friuli 11
Friuli 12
Liguria 11
Liguria 12
of 102 programmes with relevant data, 68 (66.7%) referred to
colposcopy fewer than 3% of screened women, and 89 (87.3%)
fewer than 4%. However, in 8 programmes the referral rate was
>5% and in two of them >6%. With respect to the reason for
referral (figure 5), ASC-US cytology was still a major source of
variability and reached very high levels in Molise and Valle
d’Aosta. Clearly, the regions with the lowest referral rate invited
a very low number of women with ASC-US directly to col-
poscopy but did a previous repeat of cytology or a triage by
HPV testing. However, L-SIL cytology has become now the
most frequent reason in many regions and is a second major
cause of variability.
68 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
8. Cervical cancer screening: 2011-2012 acivity
%
6
7
5
4
3
2
1
0
Figure 4. Proportion of women referred to colposcopy for any reason, by region. 2011-2012 activity.
Figura 4. Proporzione di donne invitate in colposcopia per qualsiasi motivo, per Regione.Attività 2011-2012.
Each bar represents one region, 2011-2012 activity.
Lines represent the range between 10th and 90th percentile of programme distribution within each region.
Trento 11
Trento 12
Umbria 11
Umbria 12
Campania 11
NORTH CENTRE SOUTH and ISLANDS
Campania 12
Puglia 11
Puglia 12
Friuli 11
Friuli 12
Toscana 11
Toscana 12
Liguria 11
Liguria 12
Piemonte 11
Piemonte 12
Lazio 11
Lazio 12
Veneto 11
Veneto 12
Lombardia 11
Lombardia 12
Emilia-Romagna 11
Emilia-Romagna 12
Marche 11
Marche 12
Calabria 11
Calabria 12
Molise 11
Molise 12
Sardegna 11
Sardegna 12
Sicilia 11
Sicilia 12
Basilicata 11
Aosta 11
Aosta 12
Trento Umbria Campania Toscana Friuli Puglia Marche Lombardia Liguria Piemonte Lazio Veneto Emilia
Romagna
Calabria
%
1
2
3
4
5
6
7
0
Figure 5. Proportion of women referred to colposcopy,by region and reason. 2011-2012 activity.
Figura 5. Proporzione di donne invitate in colposcopia, per Regione e motivo. Attività 2011-2012.
Sardegna Sicilia Molise Aosta
Ca H-SIL L-SIL after ASC-US and HPV test new ASC-US after ASC-US referred after first ASC-US ASCH AGC other
At a national level, the positive predictive value (PPV) of
ASC-US or more severe cytology for CIN2 or more severe his-
tology was 15.3% in 2011 and 16.9% in 2012. Its value was
just above 16% from 2006, after a rising trend which started
in 2000 (when PPV was 11.4%).
Figure 6 shows the distribution of PPV in Italian regions dur-
ing 2011 and 2012. Its value was inversely correlated to the re-
ferral rate (data not shown) and was <10% in Valle d’Aosta and
Calabria both years and for one year in Basilicata, Puglia (only
one available), Sicilia, and Molise. In 2012 Sicilia was just
above 10%, but Molise registered a remarkable increase, reach-
ing 22%. Values stably >20% were observed in four regions
(Umbria, province of Trento, Friuli-Venezia Giulia, and
Toscana).
Three of them refer to colposcopy no or very few women at the
first diagnosis of ASC-US, as a result of the implementation of
triage systems for this cytological category. However, PPV was
not very high in some regions where no or few women with
ASC-US were directly referred to colposcopy but referral be-
cause of L-SIL is relevant. Indeed, looking at specific reasons
of referral (table 4) L-SIL cytology had a PPV for CIN2+
<10%. In addition, women referred to colposcopy because of
persistent ASC-US cytology had a lower PPV for CIN2+ than
that of women referred at the first ASC-US cytology.
Among women referred to colposcopy because of an ASC-US or
more severe cytology during 2011 and 2012, 87.7% and 85.3%
respectively actually had one colposcopy, compared to 85.9 in
2010 and 85.1% in the two previous years (figure 7, p. 70).
Among women referred to colposcopy because of a H-SIL or
more severe cytology, compliance was 89.5% in 2011 and
90.4% in 2012 (figure 8, p. 71).
Figure 9 (p. 71) shows the detection rate (DR) of histologically
confirmed CIN2 or more severe lesions. The standardized (on
the Italian population truncated 25-64 yrs) DR was 3.2 lesions
detected per 1,000 screened women in 2011 and 3.5 in 2012.
Previously, DR increased from 3.0 in 2004 to 3.5 in 2010.
Overall, there was a decreasing trend from North to South.
However, high DRs, despite being lower than in 2009 and
2010, were still observed in Sardegna, where a new programme
was recently started, and there was a strong increase in Sicilia
in 2012 following invitation extension. An increase from 3.2
in 2011 to 4.9 in 2013 was also observed in Marche (central
Italy), again related to an increased proportion of women at
their first cervical screen.
Second-level activity
Colposcopic findings and their correlation with histology
Data were reported from 81 programmes both in 2011 and
2012 (table 5, p. 72). Most of the 54,776 colposcopies in-
cluded in the analysis were classified as normal (38.9%), G1
(34.8%) or unsatisfactory (11.9%).
At least one biopsy was performed in 49.5% of all colpo-
scopies: 84.2% of those with abnormal findings, 33.4% of un-
satisfactory colposcopies, and 16.8% of normal colposcopies.
When considering only colposcopies with biopsy, CIN1 or
more severe histology was detected in 69.0% of those classi-
fied as grade 1 and CIN2+ in 65.0% of those classified as
grade 2 and 89.2% of those suggestive of cancer, but just in
Cervical cancer screening: 2011-2012 acivity
69 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
%
30
35
25
20
15
10
40
45
50
5
0
Figure 6. Positive predictive value, by region. 2011-2012 activity. / Figura 6. Valore predittivo positivo, per Regione. Attività 2011-2012.
Each bar represents one region, 2011-2012 activity.
Lines represent the range between 10th and 90th percentile of programme distribution within each region.
Basilicata 11
Aosta 11
Aosta 12
Puglia 12
Calabria 11
NORTH CENTRE SOUTH and ISLANDS
Calabria 12
Sicilia 11
Sicilia 12
Campania 11
Campania 12
Lazio 11
Lazio 12
Liguria 11
Liguria 12
Veneto 11
Veneto 12
Sardegna 11
Sardegna 12
Piemonte 11
Piemonte 12
Emilia-Romagna 11
Emilia-Romagna 12
Marche 11
Marche 12
Lombardia 11
Lombardia 12
Toscana 11
Toscana 12
Molise 11
Molise 12
Friuli 11
Friuli 12
Trento 11
Trento 12
Umbria 11
Umbria 12
50.8% of those with atypical vessels. When excluding from
computations the lesions diagnosed during unsatisfactory or
unclassified colposcopies, 95.1% of CIN3+ and 93.7% of
CIN2 were identified during colposcopies with abnormal
findings (58.9% and 33.6% of CIN3+ and CIN2, respectively,
during colposcopies classified as G2, atypical vessels, or sug-
gestive of cancer).
Management and treatment of women
with screen-detected biopsy-proven CIN
Data were reported by 86 programmes in 2011 and 93 in 2012
(table 6, p. 73). No information was available for 4.8% of
women and the type of treatment was unknown for a further
1.8%.
Of all women with CIN1, 81.7% were recalled for follow-
Cervical cancer screening: 2011-2012 acivity
70 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Criterion of referral Endpoint 2011 2012
Mean (%) P10 P90 Mean (%) P10 P90
(num/den) (%) (%) (num/den) (%) (%)
H-SIL cytology CIN2+ 70.2 46.0% 100.0 71.7 50.0 100.0
(1,844/2,626) (1,719/2,397)
H-SIL cytology CIN3+ 46.4 16.0 79.3 47.2 12.5 68.4
(1,218/2,626) (1,132/2,397)
L-SIL cytology CIN2+ 9.7 3.3 20.0 9.1 3.0 20.0
(1,227/12,622) (1,098/12,022)
L-SIL cytology CIN3+ 2.9 0.0 8.0 3.1 0.0 8.1
(362/12,622) (367/12,022)
ASC-US cytology followed CIN2 13.7 2.1 33.3 12.2 0.0 17.2
by TRIAGE HPV (247/1,808) (172/1,416)
ASC-US cytology followed by CIN3 4.9 0.0 12.3 6.0 0.0 10.8
TRIAGE HPV (89/1,808) (85/1,416)
Repeat ASC-US cytology CIN2 4.0 0.0 11.5 4.5 0.0 10.0
(13/324) (17/380)
Repeat ASC-US cytology CIN3 0.3 0.0 3.85 1.8 0.0 9.1
(1/324) (7/380)
First ASC-US cytology CIN2 5.6 0.0 15.6 4.8 0.0 13.64
(440/7,814) (427/8,845)
First ASC-US cytology CIN3 2.4 0.0 5.8 2.1 0.0 5.8
(188/7,814) (187/8,845)
The table includes data from the 95 programmes that provided data in 2011 and the 92 that provided data in 2012.
Table 4. Positive predictive value for CIN2 or more severe histology and for CIN3 or more severe histology according to different criteria of referral.
Tabella 4. Valore predittivo positivo per istologia CIN2 o più grave e per istologia CIN3 o più grave, secondo diversi criteri di invio.
52.83 49.38 49.02 49.50
79.25
74.07
69.60
73.27
66.67
2002 2003 2004 2005 2006 2007 2008 2009 2010
%
10
20
30
40
50
60
70
80
90
100
0
Figure 7. Compliance with colposcopy (referral because of ASC-US or more severe cytology). Percentage of programmes that reach “acceptable” and “desirable” val-
ues by year of activity.
Figura 7. Compliance alla colposcopia (invio per citologia ASC-US o più grave).Percentuale di programmi che raggiungono valori “accettabili”e “desiderabili”, per anni
di attività.
desirable acceptable % compliance colposcopy ASC-US+, national average
51.35 52.29 50.91 47.32 50.94
74.31
70.00
75.00 76.42
2011
55.45
80.20
2012
53.72
79.34
84.65 85.91
87.68
85.27
85.09
82.25
81.61
84.75
84.72
86.00
88.00
Cervical cancer screening: 2011-2012 acivity
71 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Figure 8. Compliance with colposcopy (referral because of H-SIL or more severe cytology). Percentage of programmes that reach “acceptable” and “desirable” values
by year of activity.
Figura 8. Compliance alla colposcopia (invio per citologia H-SIL o più grave). Percentuale di programmi che raggiungono valori “accettabili” e “desiderabili”,per anni
di attività.
52.94
49.37
53.54 51.04
74.51
60.76
63.60 64.58
61.32
2002 2003 2004 2005 2006 2007 2008 2009 2010
%
10
20
30
40
50
60
70
80
90
100
0
55.60
50.48
55.14
50.94 47.06
71.43
67.29
70.75
65.69
89.63 88.21
2011
44.90
66.33
89.50
2012
44.90
66.10
90.40
89.3189.52
87.10
90.00
88.30
91.10
93.50
desirable acceptable % compliance colposcopy H-SIL+, national average
%
6
7
5
4
3
2
1
0
Figure 9. Unadjusted detection rate (per 1,000 women) of histologically confirmed CIN2+, by region.2011-2012 activity.
Figura 9. Tasso di identificazione grezza di CIN2+ con conferma istologica,per Regione.Attività 2011-2012.
Each bar rapresents one Region, 2011-2012 activity.
Lines represent the range between 10th and 90th centile of programme distrubution within each Region
Puglia 12
Molise 11
Molise 12
Calabria 11
Calabria 12
NORTH CENTRE SOUTH and ISLANDS
Campania 11
Campania 12
Basilicata 11
Trento 11
Trento 12
Piemonte 11
Piemonte 12
Aosta 11
Aosta 12
Lazio 11
Lazio 12
Sicilia 11
Sicilia 12
Liguria 11
Liguria 12
Toscana 11
Toscana 12
Umbria 11
Umbria 12
Veneto 11
Veneto 12
Emilia-Romagna 11
Emilia-Romagna 12
Sardegna 11
Sardegna 12
Friuli 11
Friuli 12
Lombardia 11
Lombardia 12
Marche 11
Marche 12
up only, in agreement with the recommendation not to treat
these lesions except if persistent.5-7 This proportion increased
from previous years (78.8% in 2010 and 73.0% in 2009).
Most of women with CIN2 (52.8%) and CIN3 (57.0%) were
treated by stand-alone radio-frequency devices. Laser conisation
was applied in 7.2% of women with CIN2 and 9.1% of those
with CIN3. Destructive treatments were still used in association
with radio-frequency devices (laser in 5.4% and 3.2% of women
with CIN2 and CIN3 respectively), but very uncommonly
alone, especially for CIN3. Cold knife conisation was limited to
8.8% of women with CIN2 and 14.5% of those with CIN3.
Of the women with adenocarcinoma in situ, 35.8% had hys-
terectomy, 15.6% cold knife conisation, and 32.1% other
more conservative excisional treatment. As first treatment,
some 59% of women with invasive cancer had hysterectomy,
7% cold knife conisation and 10% LLETZ.These plausibly in-
clude diagnostic assessment procedures. We do not know
about subsequent treatments.
No recommendation of treatment was registered for 7.0% of
CIN2 and 2.2% of CIN3. On the other hand, hysterectomy
was reported in 0.1%, 0.4%, and 2.3% of women with CIN1,
CIN2, and CIN3, respectively. Italian guidelines recommend
no more than 2% hysterectomies on CIN2/3 and virtually
none on CIN1.1,2 Diathermocoagulation, which is not rec-
ommended by guidelines,5,19 was still applied for 4.3% of
CIN1 and 1.6% of CIN2.
No treatment was registered, despite referral, in 3-4% of
women with CIN2/3 or adenoCa. In most of these cases, re-
ferrals were made >3 months in advance, suggesting refusal.
Correlation between colposcopy-guided biopsy and
excised specimen histology
Excisional histology was CIN1 or lower in 13% of women with
a CIN2-3 colposcopy-guided biopsy, similar to what was ob-
served in 2010 (14%). Among women with CIN1 who had a
colposcopy-guided biopsy, 23% had CIN2 or more severe his-
tology on the excised specimen. Higher values had been observed
in previous years: 30% in 2009 and 32% in 2010 (table 7).
DISCUSSION
Organized cervical screening programmes have now reached al-
most complete nominal extension. Italian women not in-
cluded in organized programmes are substantially only those
from most of Lombardia, which chose not to implement in-
vitational programmes. However, the programmes active in
northern and central Italy now frequently reach complete or al-
most complete invitational coverage, while in some regions of
southern Italy the invitation rate is much lower than needed.
Some decrease in invitational coverage, compared to previous
years, was observed in regions of southern but also northern
and central Italy.
It is important to avoid that funding restrictions due to the eco-
nomic crisis result in an inversion of the growing trend ob-
served up to now: this would mean losing the results of a great
amount of efforts and resources allocated for many years. Re-
cent national results20,21 confirm the early local observation22
Cervical cancer screening: 2011-2012 acivity
72 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Table 5. Colposcopic findings and histology in the colposcopies performed by 81 Italian cervical screening programmes during 2011 and 2012.
Tabella 5. Grading colposcopico ed esito istologico delle colposcopie effettuate da 81 programmi italiani di screening negli anni 2011 e 2012.
Colposcopic findings Histology
no biopsy no CIN CIN1 CIN2 CIN3 adeno invasive invasive total total
performed carcinoma squamous adeno with
in situ carcinoma carcinoma biopsy
normal colposcopic 17,701 2,217 992 215 137 8473,580 21,281
findings-transformation
zone fully visible (N)
% of total 83.2% 10.4% 4.7% 1.0% 0.6% 0.0% 0.0% 0.0%
% of total with biopsy 61.9% 27.7% 6.0% 3.8% 0.2% 0.1% 0.2%
grade 1 (N) 3,379 4,874 7,634 2,038 1103 25 25 7 15,706 19,085
% of total 17.7% 25.5% 40.0% 10.7% 5.8% 0.1% 0.1% 0.0%
% of total with biopsy 31.0% 48.6% 13.0% 7.0% 0.2% 0.2% 0.0%
grade 2 (N) 233 460 1,092 1125 1626 37 79 15 4,434 4,667
% of total 5.0% 9.9% 23.4% 24.1% 34.8% 0.8% 1.7% 0.3%
% of total with biopsy 11.6% 23.6% 24.8% 36.8% 1.4% 1.5% 0.3%
atypical vessels (N) 194 51 9 13 27 2 16 4 122 316
% of total 61.4% 16.1% 2.8% 4.1% 8.5% 0.6% 5.1% 1.3%
% of total with biopsy 41.8% 7.4% 10.7% 22.1% 1.6% 13.1% 3.3%
colposcopic features 11 55414 10 39 16 93 104
suggestive of invasive
cancer (N)
% of total 10.6% 4.8% 4.8% 3.8% 13.5% 9.6% 37.5% 15.4%
% of total with biopsy 5.4% 5.4% 4.3% 15.1% 10.8% 41.9% 17.2%
other - unsatisfactory 4,343 1,262 613 144 141 6662,178 6,521
colposcopy (N)
% of total 66.6% 19.4% 9.4% 2.2% 2.2% 0.1% 0.1% 0.1%
% of total with biopsy 57.9% 28.1% 6.6% 6.5% 0.3% 0.3% 0.3%
Number of colposcopies 1,783 514 284 95 100 3671,009 2,792
where colposcopy result
is not available (N)
% of total 63.9% 18.4% 10.2% 3.4% 3.6% 0.1% 0.2% 0.3% 54,766
% of total with biopsy 50.9% 28.1% 9.4% 9.9% 0.3% 0.6% 0.7%
Cervical cancer screening: 2011-2012 acivity
73 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Table 6. Treatment or management of the intraepithelial lesions, performed by 86 Italian screening programmes in 2011 and 93 in 2012.
Tabella 6. Trattamento o gestione delle lesioni intraepiteliali effetuati da 86 programmi italiani di screening nel 2012.
Most severe histology before treatment
CIN1* % CIN2* % CIN3* % adeno % invasive % total
carcinoma carcinoma
in situ
First treatment
laser vaporisation 231 2.3 103 2.5 24 0.7 0 0.0 0 0.0 358
cryotherapy 1 0.0 0 0.0 0 0.0 0 0.0 0 0.0 1
radical diathermy 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 0
diathermocoagulation 436 4.3 65 1.6 6 0.2 0 0.0 0 0.0 507
excision by radio-frequency device 469 4.7 2,147 52.8 1,929 57.0 24 22.0 25 10.0 4,594
cold knife conisation 84 0.8 359 8.8 490 14.5 17 15.6 18 7.2 968
laser conisation 47 0.5 293 7.2 308 9.1 10 9.2 4 1.6 662
LLETZ + Laser 20 0.2 220 5.4 108 3.2 1 0.9 0 0.0 349
hysterectomy 13 0.1 15 0.4 79 2.3 39 35.8 147 59.0 293
Other treatments
conisation NOS 0 0.0 2 0.0 2 0.1 0 0.0 0 0.0 4
radio/chemotherapy 0 0.0 0 0.0 0 0.0 0 0.0 3 1.2 3
photo-thermocoagulation 3 0.0 1 0.0 1 0.0 1 0.9 0 0.0 6
trachelectomy 0 0.0 1 0.0 1 0.0 0 0.0 0 0.0 2
polipectomy 2 0.0 0 0.0 0 0.0 0 0.0 0 0.0 2
type of treatment unknown 66 0.7 122 3.0 112 3.3 4 3.7 19 7.6 323
not treated - no treatment 8,233 81.7 285 7.0 73 2.2 3 2.8 0 0.0 8,594
recommended
not treated - treatment 36 0.4 29 0.7 24 0.7 0 0.0 2 0.8 91
recommended <3 months before
not treated - treatment 74 0.7 92 2.3 85 2.5 4 3.7 5 2.0 260
recommended ≥3 months before
unknown if treated 365 3.6 331 8.1 140 4.1 6 5.5 26 10.4 868
Total 10,080 100.0 4,065 100.0 3,382 100.0 109 100.0 249 100.0 17,885
Histology on excised specimen
worse histology before treatment negative CIN1 CIN2/3 adeno Ca invasive total not total
(<CIN) in situ cervical available available
(%) (%) (%) (%) cancer (%) (%)
CIN1 188 (20.4) 521 (56.6) 205 (22.3) 5 (0.5) 2 (0.2) 921 51 (5.2) 972
CIN2/3 154 (2.9) 537 (10.2) 4,428 (83.7) 43 (0.8) 126 (2.4) 5,288 184 (3.4) 5,472
Adeno Ca in situ 4 (5.7) 2 (2.9) 8 (11.4) 42 (60.0) 14 (20.0) 70 1 (1.4) 71
Invasive cervical cancer 7 (3.7) 0 (0.0) 40 (21.2) 7 (3.7) 135 (71.4) 189 9 (4.5) 198.
The numberof women is given,followedby percentages in brackets. "Notavailable" percentagesare computedon row totals. Theother percentagesare computed based onthe “totalavailable” data.
Table 7. Correlation between colposcopy-guided biopsy and excised specimen histology.
Tabella 7. Correlazione tra biopsia guidata dalla colposcopia e istologia del campione prelevato.
that organized programmes can increase the overall proportion
of women screened within the needed interval, thus showing
their utility. A recent nationwide analysis of the screening his-
tories of women who developed invasive cancer23 also showed
(in agreement with previous local analyses24,25) that the large
majority of those women did not comply with invitation.
Therefore, an effort to increase compliance and reduce its
negative North-South trend is needed.
When interpreting time trends of performance indicators it
must be taken into account that the population examined has
partly changed over time, mainly because of the increased ex-
tension of organized programmes. Furthermore, the detection
rate of high-grade CIN is expected to be higher in newly acti-
vated programmes than in screening programmes that are already
at subsequent screening rounds. In some areas, this phenome-
non was however compensated by an increase in immigrants
from high migration pressure countries, who have a higher
prevalence of high-grade CIN than Italian women.26-29
Performance indicators show little variation in the last years at
a national level. There is surely a long-term trend to reduce re-
call for cytology repeat (which is plausibly attributable to the
training activity in cytology interpretation, mainly performed by
GISCi) and increasing compliance to recall for repeat. Referral
to colposcopy was stable or slightly on the rise. However, PPV
was also substantially stable, after a previous increase from 2000.
Many indicators show increased homogeneity between re-
gions for the past few years. A number of outliers, however,
are still present: two regions recall over 10% of screened
women to repeat cytology and a group of regions has an ex-
tremely low PPV. A crucial factor in determining PPV is
clearly the management of ASC-US, and its heterogeneity ex-
plains part of the heterogeneity in PPV. However, variability
in criteria of interpretation of cytology still plays a relevant
role. Although implementation of triage systems for ASC-US
is needed in order to obtain high PPVs, it is not always suf-
ficient. In fact, triaging ASC-US by repeat cytology did not
reach high PPVs, possibly because criteria of interpretation
were too loose in any case. In addition, high frequency of
ASC-US was replaced by high frequency of LSIL reports,
without reaching high PPVs.
Programmes with low PPV are mainly (but not only) from
southern Italy, where organized programmes started their ac-
tivity more recently. The very low CIN2+ detection rate ob-
served in a few regions in southern Italy also requires attention.
In southern Italy recent data showed a prevalence of HPV in-
fection similar to that in central and northern Italy. This sug-
gests a similar baseline risk at least in younger cohorts.30
Therefore, the low detection rate could be the result of a se-
lective uptake of invitation of women who had already been in-
tensively screened, low sensitivity of cytology and/or histology,
or low compliance to colposcopy. Part of the low compliance
to colposcopy may depend upon incomplete registration (es-
pecially of colposcopies performed outside reference centres).
Moreover, part of the low detection rate may depend upon in-
Cervical cancer screening: 2011-2012 acivity
74 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
complete registration or missing links with histology results. In
any case, it is essential to strive for the implementation of good
fail-safe systems. Lack of diagnostic work-up can make the ef-
forts made for primary screening useless.
In conclusion, data suggest that most of the programmes that
have been active for many years reached a good quality, likely
thanks to the long-lasting monitoring and intensive activity of
quality assurance. On the other hand, the newly started pro-
grammes in southern Italy need strong support to improve qual-
ity, particularly as for the specificity of first-level cytology and
the completeness of follow-up and registration.There is a need
for intervention, as, in some areas, the current situation does not
guarantee effectiveness of screening and acceptable levels of un-
desired effects. A shift to HPV testing could solve problems con-
cerning quality of cytology interpretation, but it would not be
an effective solution to problems of loss to follow-up.
The application of appropriate treatments has largely im-
proved during the last few years and has now reached levels
that are acceptable – although still not optimal – in almost all
geographical areas. In addition, these data are still missing
from many programmes. High quality of diagnostic work-up
and treatment and strict adherence to positive women man-
agement algorithms are needed in view of a shift to HPV-
based screening.
Conflicts of interests: none declared
Gatti, E. Borciani (Piacenza); L. Lombardozzi, M. Zatelli (Parma); S. Prandi,
L. Paterlini, C. Campari (Reggio Emilia); R. Andresini, C. Goldoni, A. Ven-
turelli (Modena); P. Cristiani, M. Manfredi, A. Pasquini, P. Biavati (Bologna);
R. Nannini, L. Caprara (Imola); A. De Togni, M.C. Carpanelli, C. Palmonari
(Ferrara); M. Serafini, N. Morini, B. Vitali (Ravenna); M. Palazzi, M. Farneti,
C. Imolesi, M. Severi (Cesena); F. Desiderio, D. Canuti, G. Monticelli (Rimini)
Toscana:
A. Iossa, C. Di Pierro, C. Visioli (Firenze); P. Amico, F. Marini (Em-
poli); S. Coccioli, D. Giorgi (Lucca); M. Rapanà, P. Marini, L. Ieri (Pistoia);
P. Vivani, C. Nicolai, G. Tornabene (Massa Carrara); R. Turillazzi, E. Monti
(Siena); A. Scarfantoni (Viareggio); T. Bechelli, L. del Chicca (Pisa); C. Maf-
fei, A. Zani (Livorno); P. Piacentini, R. Rosati (Grosseto); C. Epifani, L. Ab-
delghani (Prato); F. Mirri (Arezzo)
Umbria:
M.D. Giaimo, S. Prandini, C. Bietta (ASL 2 Perugia); G. Vinti (ASL
1 Città di Castello); R. Corvetti (ASL 4 Terni); A. Di Marco (ASL 3 Foligno)
Marche:
L. Di Furia, A. Giustozzi (Regione Marche); M. Marcucci, A. San-
tone (Ascoli Piceno); M.G. Volpini, N. Budini Gattai (Camerino); R. Passa-
tempo, M. Malizia, M.G. Volpini (Civitanova); M. Agostini,
A. Vaccaro (Fano); G. Picchietti, S. Paci (Jesi); L. Marinelli, M.G. Volpini (Ma-
cerata); G. Giacomucci, M. Agostini (Pesaro); A. Barzetti, S. Bozzi Cima-
relli (Senigallia); M.T. Capasso, M. Agostini (Urbino); F. Minnucci, P. Benti-
voglio (Ancona); A. Santone, M.R. Taraborelli (San Benedetto del Tronto);
C. Ragaglia, M.T. Lanciotti (Fermo); E. Bruschelli (Fabriano)
Molise:
C. Giammaria, L. Di Lullo, A. Di Credico
Lazio:
A. Barca, D. Baiocchi (Regione Lazio); M. Amato, M. Martini (Roma
A); F. Odoardi, M. Pontani (Roma B); S. Marzani, P. Capparucci (Roma C);
G. Petricone, F. Puddu (Roma D); M.G. Acampora (Roma E); M.C. Tufi, P.
Barbarino, M. Arcara (Roma G); P. Rigato, M. Mammola (Roma H); S.
Brezzi, P. Raggi (Viterbo); G. Baldi (Rieti); A. Di Cesare, P. Bellardini (Latina);
L. Martufi (Frosinone)
Abruzzo:
C. Fortunato, E. Altobelli, A. Lattanzi (ASL Teramo); M. Minna,
A. Calabrese (ASL Pescara); A. Macerola (ASL Av-Sulmona-L’Aquila); D. Ca-
raceni, M. Muzii, F.M. Lattanzio (ASL Lanciano-Vasto-Chieti)
Valle D’Aosta:
G. Furfaro, T. Meloni, M. Cognein
Piemonte: N. Segnan, E. Mancini (Torino); G. Faragli (Alessandria); S. Polizzi
(Moncalieri); L. Orione (Cuneo); M.P. Alibrandi (Ivrea); T. Miroglio (Asti); M.
Sartori, M. Ciminale (Rivoli-Val di Susa); C. Magnani, P. Bestagini (Novara); F.
Germinetti (Vercelli-Biella)
Lombardia:
D. Cereda, L. Coppola, M. Gramegna, L. Zerbi (DG Salute Re-
gione Lombardia); M. Schivardi, M. Crisetig, F. Speziani (ASL Brescia); S. Gotti,
M. Dal Soldà, L. Boldori (ASL Cremona); A. Belloni, E. Rossetti, G. Marazza
(ASL Lodi); E. Anghinoni (ASL Mantova); G. Magenes, L. Camana (ASL Pa-
via); S. Domenighini, G. Pieracci (ASL Valle Camonica Sebino)
Liguria:
L. Bonelli, I. Valle (Genova 3); A. Franco (Savonese); M. Orlando
(Imperiese)
PA Trento:
S. Piffer, M.A. Gentilini, E. Polla, P. Dalla Palma
PA Bolzano:
A. Fanolla
Friuli-Venezia Giulia:
L. Zanier, A. Franzo, E Clagnan
Veneto:
M. Zorzi, C. Fedato, C. Cogo, F. Soppelsa, R. Mel, S. Di Camillo
(ULSS 1 Belluno); L. Cazzola, G. Orsingher (ULSS 2 Feltre); C. Sannino, M.A.
Zanella, M. Perli (ULSS 3 Bassano); F. Banovich, S. Saccon (ULSS 4 Thiene-
Alto Vicentino); M. Lestani, N. Scomazzon (ULSS 5 Ovest Vicentino); B. Co-
ria (ULSS 6 Vicenza); S. Cinquetti, T. Moretto (ULSS 7 Pieve di Soligo); G.
Lustro, G. Diacono (ULSS 8 Asolo); L. Laurino, L. Finotto (ULSS 9 Treviso);
A. Favaretto (ULSS 10 Veneto Orientale); F. Zago, M. Lorio (ULSS 12 Ve-
neziana); A. Montaguti, (ULSS 13 Mirano); M.L. Polo (ULSS 14 Chioggia);
A. Pupo, F. Ortu, S. Callegaro (ULSS 15 Camposampiero Cittadella); I. Si-
moncello, M. Matteucci (ULSS 16 Padova); A. Ferro, S. Bertazzo, (ULSS 17
Este-Monselice); L. Gallo, A. Stomeo, N. Volpe, R. Buoso (ULSS 18 Rovigo);
A. Del Sole, R. Spitti (ULSS 19 Adria); R. Colombari, P. Cattani, M. Bona-
mini, I. Brunelli (ULSS 20 Verona); P. Coin, O. Dal Pezzo, K. Greco (ULSS 21
Legnago); C. Capaldo, A. Ganassini (ULSS 22 Bussolengo)
Emilia-Romagna:
C. Naldoni, A.C. Finarelli, P. Sassoli de’ Bianchi (Asses-
sorato politiche per la salute); F. Falcini, R. Vattiato, L. Bucchi, S. Mancini,
A. Colamartini (Forlì e Assessorato politiche per la salute); P.G. Dataro, G.
Data for the years 2011-2012 was provided by:
Hanno fornito i dati per gli anni 2011-2012:
Cervical cancer screening: 2011-2012 acivity
75 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Puglia); V. Piazzolla (ARES Puglia); N. Albano (Resp. RIP NSISR); C. Ger-
minario, D. Martinelli, M.S. Gallone (OER Puglia); N. Morelli (ASL Bari);
A. Belsanti (ASL Bt); G. Spagnolo (ASL Brindisi); R. Prato (ASF Foggia);
M.R. Canitano (ASL Lecce); S. Sabato (ASL Taranto)
Basilicata:
R. Maglietta, A. Basanisi, A. De Bartolomeo
Sicilia:
G. Dardanoni, I Schimmenti (Ass. Reg. Sanità); A. Matina, F. La
Porta (ASP Agrigento); F. Sferrazza (ASP Caltanissetta); R. Scillieri, A. Sca-
lisi (ASP Catania); G. Macaluso, L. De Simone (ASP Enna); S. Paratore (ASP
Messina); M. Spedale (ASP Palermo); S. D’Amanti (ASP Ragusa); S. Mali-
gnaggi (ASP SR); R. Candura, A. Barraco (ASP Trapani)
Sardegna:
R. Masala e P. Tanchis (dal 2012) (Ass. Reg. Sanità); S. Tilocca
(Cagliari); F. Congiu (Sanluri); O. Frongia (Oristano); M.A. Atzori (Nuoro);
U. Stochino (Lanusei); M. Piga (Olbia); A. Onnis (Carbonia)
Campania:
R. Pizzuti, A. Pugliese, G. Albano (Avellino 1);
M. Cozza (Avellino 2); G. Ragozzino (Benevento); G. Capone (Caserta 1);
C. Vatiero, E. Frezza (Caserta 2); R. Papa (Napoli 1); M.T. Pini (Napoli 2);
M. Panico (Napoli 3); A. Esposito, C. Maione, R.P. Esposito (Napoli 4);
F.S. Manco (Napoli 5); U. Scala (Salerno 1); G. Auriemma (Salerno 2);
G. Martuscelli (Salerno 3)
Calabria:
L. Rizzo; A. Giorno (AS 4 Cosenza); T. Landro (ASP Vibo V.); M.
Viola (ASP Reggio C.); C. Spadafora (ASP Crotone); M.P. Montesi (ASP Ca-
tanzaro, ambito territoriale Lamezia T.); E. Bova (ASP Catanzaro, ambito
territoriale Catanzaro)
Puglia: V. Pomo (Dir. Area politiche per la promozione della salute,
delle persone e delle pari opportunità, Regione Puglia); G. Labate, C.
Ladalardo (Servizio PATP, Assessorato alle politiche della salute, Regione
References/Bibliografia
1 Commissione oncologica nazionale. Proposte operative in tema di
prevenzione secondaria del cervico-carcinoma uterino. In: Linee gui-
da elaborate dalla Commissione oncologica nazionale, in applica-
zione di quanto previsto dal Piano Sanitario Nazionale per il trien-
nio 1994-96, relativo all’azione programmata “Prevenzione e cura
delle malattie oncologiche”, concernenti l’organizzazione della pre-
venzione e dell’assistenza in oncologia. Supplemento ordinario alla
Gazzetta Ufficiale n.127, 1.6.1996.
2. Conferenza permanente per i rapporti tra lo Stato, le Regioni e le
Province autonome di Trento e Bolzano. “Provvedimento 8 marzo
2001. Accordo tra il Ministro della sanità e le Regioni e Province au-
tonome di Trento e Bolzano sulle linee-guida concernenti la pre-
venzione, la diagnostica e l’assistenza in oncologia”. Gazzetta Uf-
ficiale, 2.5.2001.
3. Ministero della salute, Direzione generale della prevenzione.
Screening oncologici. Raccomandazioni per la pianificazione e l’ese-
cuzione degli screening di popolazione per la prevenzione del can-
cro della mammella, del cancro della cervice uterina e del cancro
del colon retto. Roma, 2006.
4. Coleman D, Day N, Douglas G et al. European guidelines for quality
assurance in cervical cancer screening. Eur J Cancer 1993; 29A
(Suppl 4):S1-38.
5. Arbyn M, Anttila A, Jordan J et al. (eds). European guidelines for qual-
ity assurance on cervical cancer screening. 2nd edition. Office for of-
ficial publications of the European communities, Luxembourg 2008.
6. Ronco G, Pilutti S, Naldoni C et al. Stato dello screening cervicale
in Italia. In: Rosselli Del Turco M, Zappa M (eds). Osservatorio na-
zionale per la prevenzione dei tumori femminili. Primo rapporto.
Roma 2002:32-49.
7. Ronco G, Ricciardi V, Naldoni C et al. Livello di attivazione e indi-
catori di processo dei programmi organizzati di screening cervica-
le in Italia. In: Rosselli Del Turco M, Zappa M (eds). Osservatorio na-
zionale per la prevenzione dei tumori femminili. Secondo rappor-
to. Roma 2003:36-51.
8. Ronco G, Giubilato P, Naldoni C et al. Livello di attivazione e indi-
catori di processo dei programmi organizzati di screening in Italia.
In: Roselli Del Turco M, Zappa M (eds). Osservatorio per la preven-
zione dei tumori femminili. Terzo rapporto. Roma 2004:38-51.
9. Ronco G, Giubilato P, Naldoni C et al. Activity level and process
indicators of organised programmes for cervical cancer screening
in Italy. Epidemiol Prev 2006;2 (Suppl.1):27-40.
10. Ronco G, Giubilato P, Naldoni C et al. Extension of organised cer-
vical screening programmes in Italy and their process indicators. Epi-
demiol Prev 2007;2-3(Suppl 2):33-47.
11. Ronco G, Giubilato P, Naldoni C et al. Extension of organised cer-
vical cancer screening programmes in Italy and their process indi-
cators. Epidemiol Prev 2008;2(Suppl 1):37-54.
12. Ronco G, Giubilato P, Naldoni C et al. Extension of organised cervi-
cal cancer screening programmes in Italy and their process indica-
tors: 2007 activity. Epidemiol Prev 2009;3(Suppl 2):41-56.
13. Ronco G, Giubilato P, Naldoni C et al. Extension of organised
cervical cancer screening programmes in Italy and their process
indicators: 2008 activity. Epidemiol Prev 2010;5-6(Suppl 4):35-51.
14. Ronco G, Giubilato P, Naldoni C et al. Extension of organised cervi-
cal cancer screening programmes in Italy and their process indica-
tors: 2009 activity. Epidemiol Prev 2011;5-6(Suppl 5):39-54.
15. Ronco G, Giubilato P, Naldoni C et al. Extension of organised cervi-
cal cancer screening programmes in Italy and their process indica-
tors: 2010 activity. Epidemiol Prev 2012;6(Suppl 1):39-54.
16. Ministero della salute. Piano Nazionale della Prevenzione 2010-2012.
Azione centrale prioritaria concernente la definizione di documenti
tecnici di sintesi delle evidenze scientifiche a supporto della pro-
grammazione, monitoraggio e valutazione degli interventi di pre-
venzione oncologica nella popolazione a rischio. Documento di in-
dirizzo sull’utilizzo dell’HPV-DNA come test primario per lo screen-
ing del cancro del collo dell’utero.[www.osservatorionazionale-
screening.it/sites/default/files/allegati/Screening.pdf]
17. Ronco G, Giorgi-Rossi P, Giubilato P et al. A first survey of HPV-based
screening in routine cervical cancer screening in Italy. Epidemiol Prev
2015;3(Suppl 1):77-83.
18. Stafl A,Wilbanks GD. An International Terminology of Colposcopy.
Report of the Nomenclature Committee of the International
Federation of Cervical Pathology and Colposcopy. Obstet Gynecol
1991;77:313-14.
19. Wright TC, Massad LS, Dunton CJ et al. 2006 consensus guide-
lines for the management of women with cervical intraepithelial
neoplasia or adenocarcinoma in situ. Am J Obstet Gynecol
2007;197:340-45.
20. Istat.Indagine multiscopo «Condizioni di salute e ricorso ai servizi san-
itari». Istat, Roma 2006. [http://www3.istat.it/salastampa/comunicati/
non_calendario/20061204_00/testointegrale.pdf – accessed: Apr 15,
2014].
21. PASSI (Progressi delle Aziende Sanitarie per la Salute in Italia). Rap-
porto nazionale PASSI 2012: screening cervicale. [http://www.epicentro.
iss.it/passi/rapporto2012/ScreeningCervicale.asp – accessed: Apr 15,
2014].
22. Ronco G, Segnan N, Giordano L et al. Interaction of spontaneous
and organised screening for cervical cancer in Turin, Italy. Europ J
Cancer 1997;8:1262-67.
23. Zucchetto A, Ronco G, Giorgi Rossi P et al. Screening patterns with-
in organized programs and survival of Italian women with invasive
cervical cancer. Prev Med 2013 Jun 1. doi:pii: S0091-7435(13)00181-
3. 10.1016/j.ypmed.2013.05.018
24. Ronco G, Pilutti S, Patriarca S et al. Impact of the introduction of or-
ganised screening for cervical cancer in Turin, Italy: cancer incidence
by screening history 1992-98. Br J Cancer 2005; 8;93(3):376-78.
25. Zucchetto A, Franceschi S, Clagnan E et al. Screening history of
women with invasive cervical cancer in north-east Italy. Eur J Ob-
stet Gynecol Reprod Biol 2010;152:200-204.
26. Di Felice E, Caroli S, Paterlini L et al. Cervical cancer epidemiology in
Cervical cancer screening: 2011-2012 acivity
76 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
foreign women in Northern Italy: role of HPV prevalence in country
of origin. Eur J Cancer Prev [In press].
27. Vallesi G, Bietta C, Marri M, Petrella M. [Immigration from countries
with a strong migratory pressure and participation in cervical cancer
screening program in the Local Health Unit 2, Umbria Region. Impact
on the probability of high-grade lesions and cervical cancer]. [Article
in Italian] Epidemiol Prev 2012;2:95-99.
28. Visioli CB, Crocetti E, Zappa M et al. Participation and risk of high
grade cytological lesions among immigrants and Italian-born
women in an organized cervical cancer screening program in cen-
tral Italy. J Immigr Minor Health 2014 Jun 11.
29. Campari C, Fedato C, Iossa A et al and the GISCi Migrant Working
Group. Cervical cancer screening in immigrant women in Italy: a sur-
vey on participation, cytology, and histology results. Eur J Cancer Prev
[submitted].
30. Giorgi Rossi P, Bisanzi S, Paganini I et al. Prevalence of HPV high and
low risk types in cervical samples from the Italian general population:
a population based study. BMC Infect Dis 2010;10:214.
A first survey of HPV-based screening
in routine cervical cancer screening in Italy
Prima survey sull’utilizzo routinario del test HPV
nello screening cervicale in Italia
Guglielmo Ronco,1Paolo Giorgi-Rossi,2,3 Pamela Giubilato,1Annarosa Del Mistro,4Marco Zappa,5
Francesca Carozzi5and the HPV screening survey working group
Abstract
Pilot HPV-based cervical screening programmes have recently started in Italy, partly on the strength of a
large randomized trial. The Ministry of Health recommended that regions shift toward HPV-based scree-
ning in early 2013 and provided guidelines for its application (stand-alone HPV testing by validated me-
thods, cytological triage of HPV positives, beginning at age 30-35, 5-year intervals). A first survey on the
2012 activity was conducted in 2013.
In 2012, 19 Italian organized cervical screening programmes from 10 regional programmes invited
311,856 women (8.0% of all women invited for cervical screening in 2012 in Italy) for HPV-based scree-
ning; 41.5% complied, with a decreasing North-South trend. Among screened women, 7.9% (range
4.3%-13.9%) were HPV positive, decreasing to 6.6% (range 4.0%-12.4%) when considering women
aged 35-64 years. Among HPV positive women, 34.8% (with high variability between programmes: range
11.1%-59.3%) were judged to have ASC-US or more severe cytology (5.3% ASC-US, 26.6% L-SIL, 5.2%
H-SIL). Out of all screened women, those referred to colposcopy based on HPV and cytology results were
2.9% (range 0.6%-4.8%), whereas they were 2.0% when considering only women aged 35-64 years.
(Epidemiol Prev 2015; 39(3) Suppl 1: 77-83)
Keywords: cervical cancer, mass screening, HPV test, Italy
1AOU Città della salute
e della scienza,
CPO Piemonte, Torino
2Servizio interaziendale
di epidemiologia, AUSL
Reggio Emilia
3IRCCS-Arcispedale
S. Maria Nuova,
Reggio Emilia
4Istituto oncologico
veneto IRCSS
5Istituto per lo studio
e la prevenzione
oncologica (ISPO), Firenze
Corresponding author:
Guglielmo Ronco
guglielmo.ronco@cpo.it
WWW.EPIPREV.IT
77 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Riassunto
Recentemente, in parte sull’onda dei risultati di un ampio trial randomizzato, in Italia sono stati attivati pro-
grammi pilota di screening cervicale basati sul test HPV. All’inizio del 2013 il Ministro della salute ha rac-
comandato alle Regioni di passare a screening basati sul test HPV e ha fornito linee guida per la sua appli-
cazione (utilizzo del solo test HPV applicando metodi validati, triage citologico dei casi positivi al test, inizio
all’età di 30-35 anni, intervalli di 5 anni). Una prima survey sull’attivià del 2012 è stata condotta nel 2013.
Nel 2012, 19 programmi organizzati di screening cervicale afferenti a 10 regioni hanno invitato allo scree-
ning basato sul test HPV 311.856 donne (8,0% di tutte le donne invitate allo screening cervicale nel 2012
in Italia) . Di queste, il 41,5% ha aderito con un trend decrescente da Nord a Sud. Tra le donne sottopo-
ste a screening, il 7,9% (range 4,3%-13,9%) era HPV positivo, percentuale che diminuisce al 6,6% (range:
4,0%-12,4%) se si considerano solo le donne di età fra 35 e 64 anni. Tra le donne positive all’HPV, il test
citologico ha dato esito ASC-US o più grave (5,3% ASC-US; 26,6% L-SIL; 5,2% H-SIL) nel 34,8% dei casi
(con un’alta variabilità fra programmi, range: 11,1%-59,3%).
Di tutte le donne sottoposte a screening, quelle inviate in colposcopia sulla base dei risultati del test HPV e
degli esiti ctologici sono state il 2,9% (range: 0,6%-4,8%), percentuale che si abbassa al 2,0% se si con-
siderano solo le donne di età fra 35 e 64 anni.
(Epidemiol Prev 2015; 39(3) Suppl 1: 77-83)
Parole chiave: cancro cervicale, scereening di massa, test HPV, Italia
HPV-based cervical cancer screening: first survey
78 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
Testing for the DNA of oncogenic HPV types as a primary
screening test for cervical cancer precursors has been in-
tensively studied over the past few years.
Randomized controlled trials (RCTs) comparing
HPV-based to cytology-based screening have been
conducted in Sweden (Swedescreen1), the Netherlands
(POBASCAM2), England (ARTISTIC3), Italy (NTCC4),
India,5Finland,6and Canada (CCCast7and FOCAL8).
The first four studies1-4 published data on two screening
rounds showing increased detection of high-grade CIN at
the first round and decreased detection in the second when
comparing the HPV and cytology groups. This proves that
HPV-based screening allows earlier detection of persistent
high-grade CIN than cytology. In addition, the Indian
study showed a reduced incidence of cervical cancer mor-
tality and advanced cancers after a once-in-a-lifetime screen
by HPV.5These findings were confirmed by a pooled
analysis of the RCTs that published results on two screen-
ing rounds with respect to invasive cancer incidence, which
provided direct evidence of increased protection with HPV-
based screening.9
Over the past few years, partly on the strength of the
NTCC experience, a number of pilot HPV-based screen-
ing projects have started up within Italian organized cer-
vical screening programmes. They have mainly aimed at
evaluating the feasibility of HPV-based screening in rou-
tine activity.
In the meanwhile, an Italian Health Technology Assess-
ment report was published in 2012.10 It concluded that
HPV-based screening was more effective than cytology-
based cervical screening and entailed little or no increase in
negative effects if appropriate protocols were applied. This
included using stand-alone HPV as a primary screening
test, with only clinically validated DNA-HPV tests,11 start-
ing HPV-based screening at 30 to 35 years of age, adopt-
ing 5-year intervals and applying a «cytological triage»
protocol. The latter entailed testing HPV-positive women
for cytology (using material taken during the HPV sam-
pling visit) and referring directly to colposcopy only
women with ASC-US or more severe cytology, while re-
testing HPV-positive women with normal cytology after
one year for stand-alone HPV and referring them to col-
poscopy only if HPV was still positive (figure 1).
Among other recommendations, the HTA report recom-
mended strict monitoring of HPV-based screening. In early
2013, the Italian Ministry of Health adopted these recom-
mendations as a guide for screening planning by regional
health authorities.12
In 2013, a first Italian national survey of HPV-based
screening was conducted as part of the yearly survey of cer-
vical cancer screening by organized programmes.
Figure 1. Protocol recommended
by the Italian HTA report.
Figura 1. Protocollo raccoman-
dato dal report italiano di Healt
Technology Assessment.
negative
referred to new
screening round referred to coloscopy
positive
samples for HPV and cytology taken
HPV test (HC2)
cytology stained and interpreted
INFORMED OF HPV POSITIVITY
cytology WNL
(Within Normal Limits)
invited after 1 year for new HPV test
cytology ASC-US+
or unsatisfactory
HPV test negative HPV test still positive
METHODS
Surveys designed to assess the level of implementation of
organized cervical screening programmes in Italy and col-
lect process indicators are conducted every year by the
ONS (Osservatorio nazionale screening, National centre
for screening monitoring) on behalf of the Italian Min-
istry of Health. Data are collected through a question-
naire as aggregated tables of data. Details are provided
elsewhere.13
A survey section dedicated to HPV-based screening was
added to the general survey in 2013, related to the 2012 ac-
tivity. It was designed assuming that the protocol suggested
by the HTA and Ministry guidelines was applied. As the
protocol entails 1-year repeats for HPV-positive women with
normal cytology, it was decided to split the collection of data
on women invited each year for primary screening into two
sections. The first section, including data on invitation and
participation to the HPV test, its result and results of triage
cytology, is collected in the year after the invitations. In Sep-
tember 2013, therefore, data were collected on women in-
vited for primary HPV testing during 2012 and tested by
April 2013. The second section, including 1-year repeats
and colposcopies resulting both from cytology and 1-year
repeat HPV tests, were collected during 2014 for women
invited for primary testing in 2012.
In addition to these data, information on the screening
protocol applied was also collected.
RESULTS
Extension of HPV-based screening
and participation
In 2012, 19 Italian programmes from 10 regions invited
women for HPV-based screening (table 1). Eleven of them
were from northern Italy, 3 from central Italy, and 5 from
southern Italy. Five programmes (Torino, Trento, Reggio
Emilia, Firenze, and Molise) invited both women to HPV-
based and cytology-based screening (the first 3 within a
randomized pilot project), while the remaining 14 invited
women just to HPV-based screening. Overall, 311,856
women aged 25-64 years were invited to HPV screening,
representing 8.0% of all women invited for cervical screen-
ing in Italy in 2012 (9.5%, 4.0%, and 8.8% of those in-
vited in northern, central, and southern Italy, respectively).
The regions with the largest number of women invited to
HPV were Veneto, where 6 programmes converted com-
pletely to HPV, and Abruzzo, where the entire region
moved to HPV testing. In addition, the region of Liguria,
where only a small area was previously covered by organ-
ized programmes, chose to extend coverage inviting to
HPV testing. In 2012, 61% of women invited for cervical
screening in Liguria were invited to HPV testing.
As the national guideline came out in 2013, all pro-
grammes active in 2012 were pilot projects. Among
them, 12 programmes started inviting women to HPV
testing at 25 years and 7 at 35. However, after publication
of the national guidelines, many programmes have
planned to shift age of first testing to 30 or 35 years. All
programmes used clinically-validated DNA-HPV tests
(mostly Digene Hybrid Capture2, and in few cases
Roche’s Cobas or the Abbott real-time PCR test) except
one which used an mRNA test. This programme was ex-
cluded from further analyses.
In 2012, 41.5% of all women invited to HPV DNA-
based screening complied. There was a strong variability
between centres. The lowest values, below 20%, were in
southern Italy and the highest, above 65%, in northern
Italy, reproducing outcomes observed with all invita-
tions13 (figure 2
, p.80
). Results were very similar when
restricted to women aged 35-64 years (mean 42.1%, data
by centre not shown).
Process indicators with HPV
All programmes used stand-alone HPV as primary test and
adopted cytological triage as recommended by national
HPV-based cervical cancer screening: first survey
79 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Region Number of programmes Target age Women invited Women screened
Abruzzo 4 25-64 108,739 34,094
Emilia-Romagna 1 35-64 5,192 3,280
Lazio 2 25-64 e 35-64 36,052 13,068
Liguria 1 35-64 14,164 6,453
Lombardia 1 25-64 8,317 5,294
Molise 1 35-64 2,000 251
Piemonte 1 35-64 24,289 12,419
Toscana 1 35-64 40 29
Trento 1 35-64 2,865 1,134
Veneto 6 25-64 110,198 55,147
Northern Italy 11 165,025 83,727
Central Italy 3 36,076 13,083
Southern Italy 5 110,739 34,345
Italy 19 311,840 131,155
Table 1. Organized pro-
grammes that invited women
to HPV-based screening in
Italy. 2012 activity.
Tabella 1. Programmi orga-
nizzati che hanno invitato a
uno screening basato sul test
HPV in Italia. Attività 2012.
guidelines and reported in figure 1. One programme
(Firenze) was excluded from calculations given the very low
number of women screened in 2012.
Since HPV infection prevalence is age-dependent and age
of start was different between programmes, we computed
the proportion of women positive to the primary HPV test
both for any age and restricted to age 35-64 years (figure 2).
Overall, 7.9% (range 4.3%-13.9%) of screened women of
any age and 6.6% (range 4.0%-12.4%) of those aged 35-
64 years (excluding Roma G because data by age were not
available) were HPV positive. Within programmes that in-
vited women aged 25-34 years, the overall prevalence was
1.5%-2.7% higher than the prevalence in the same pro-
grammes restricted to women aged 35-64 years. The low-
est value was observed in Trento (as was already the case in
the NTCC study14). High values were observed in Abruzzo
and Molise (southern Italy).
Overall, when including all ages, 34.8% of HPV-positive
women were judged to have ASC-US or more severe cy-
tology, with a very large variation, ranging from 11.1% in
Trento and 19.4% in Torino to 59.3% in a programme in
Veneto (figure 3). The proportion of HPV-positive women
classified as ASC-US or AGC was 5.3% (range 0.0%-
23.1%), that of women classified as L-SIL was 24.6%
HPV-based cervical cancer screening: first survey
80 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
18.4
Molise
Avezzano
Pescara
Roma G
Padova
Lanciano
Latina
Trento
Teramo
Savonese
Este
Veneziana
Torino
Emilia
Romagna
Rovigo
Adria
Valle
Camonica
Alta
Padovana
%
22.0
35.5 38.3 39.0 39.6 40.2
45.6 47.4 48.4 49.4 51.1 54.0
63.1 63.2 63.7 65.3
72.5
10
20
30
40
50
60
70
80
0
Figure 2. Compliance to invitation to HPV-based screening.All ages included. Italian organized programmes 2012 activity.
Figura 2. Compliance all’invito allo screening basato sul test HPV. Tutte le età incluse.Attività 2012 dei programmi organizzati.
NORTH CENTRE SOUTH and ISLANDS
Trento
%
2
4
6
8
10
12
14
0
Figure 3. Proportion of HPV-positive women. Italian organized programmes 2012 activity.
Figura 3. Proporzione di donne HPV-positive.Attività 2012 dei programmi organizzati.
Each bar rapresents one programme, 2012 activity.
NORTH CENTRE SOUTH and ISLANDS Each bar rapresents one programme, 2012 activity.
Left-hand bars refer to 35-64 year-old women, right-hand bars refer to all women.
4.0
*Roma G.
4.6 4.6
Padova
4.7
Veneziana
4.8
Rovigo
5.0
Adria
5.1
Valle
Camonica
5.6
Este
5.8
Avezzano
6.0
Torino
6.0
Latina
6.0
Reggio
Emilia
6.6
Teramo
7.8
Savonese
8.2
Lanciano
9.1
Molise
10.4
Pesacara
12.4
* age-stratified data are not available for this center
Alta
Padovana
(range 9.8%-41.7%), and of women classified as H-SIL or
ASC-H was 5.2% (range 1.7%-11.0%). Results were sim-
ilar when restricted to women aged 35-64 years: 34.8% of
HPV+ women were classified as ASC-US or higher.
When considering all ages, 2.9% of screened women were
referred to colposcopy on the basis of the primary HPV test
and simultaneous cytology. Variability was still very high,
ranging from 0.6% in Trento to 4.8% in Savona. Values
were below 2% in 5 programmes and below 3% in 11 (fig-
ure 4). When restricting data to women aged 35-64 years,
2.0% of women were referred. Trento (0.5%) and Savona
(4.8%) were again the programmes with the lowest and
highest values. Within the centres that invited women from
age 25, the referral rate including all women was 1.09 to
1.60 times the referral restricted to women aged 35-64 years.
DISCUSSION
In this first survey of HPV-based screening, only an in-
complete set of performance indicators can be presented.
Data dealing with the entire screening process on women
screened in 2012 will be presented next year. Nevertheless,
these are, to our knowledge, the first nationwide data on
routine HPV screening based on a large population.
The shift to HPV-based screening is becoming relevant in
Italy. In 2012, about 10% of women invited for primary
screening by Italian organized programmes were invited to
HPV testing. This proportion is expected to rapidly in-
crease after the publication of the guidelines of the Min-
istry of Health in January 2013. To our knowledge, in May
2014, 7/21 regions had decided to implement HPV-based
screening as the routine screening method to the entire fe-
male population in the recommended age range, although
this implementation will be progressive (3 to 5 years) in
most cases.15 For example, the region ofToscana started by
inviting the 55-64 age group in December 2012 and ex-
pects to complete accrual in three years by progressively
inviting younger women. Conversely, the region of
Piemonte plans to invite for HPV an increasing proportion
of randomly defined women over a span of three years, and
the entire target population starting from the fourth year.
One of the crucial issues with HPV-based screening is the
application of appropriate protocols, in order to avoid neg-
ative effects for women and increased costs. Indeed, rec-
ommendations on stand-alone HPV testing and cytological
triage were adopted by all centres. On the other hand,
guidelines were delivered by the Ministry after the period
of activity considered here. This explains the inclusion of
younger women in the target population. Due to the same
reason, screening intervals were still, officially, 3 years, but
are now being changed.
Compliance to invitation was slightly higher than nation-
wide when considering compliance to all invitations (to
HPV or cytology), which was 40.8% in 2012. Given the
high variability between centres and ages, a comparison of
this sort is not reliable, but at least suggests that invitation
to HPV testing is not a barrier to participation. Indeed,
an increased compliance to invitation to HPV when com-
pared to historical controls was observed in the pilot pro-
grammes in Veneto16,17 and Lazio.18
Variability in the proportion of women positive to HPV
testing was substantial even when restricting data to women
of the same age. However, it could well reflect true differ-
ences in screened populations. Substantial variability was
also observed in the NTCC study, where prevalence was
lowest in Trento.14 High prevalence was also previously ob-
served in Abruzzo.19
There was also a striking variability between programmes in
the proportion of HPV-positive women classified as having
abnormal cytology, resulting in strong variability of refer-
ral to colposcopy on the basis of cytological abnormalities.
HPV-based cervical cancer screening: first survey
81 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Trento
Torino
Pescara
Este
Latina
Roma G
Lanciano
Rovigo
Reggio
Emilia
Avezzano
Adria
Molise
Valle
Camonica
Teramo
Alta
Padovana
Savonese
Veneziana
Padova
%
10
20
30
40
50
60
0
Figure 4. Proportion of HPV-positive women with abnormal cytology. Women of any age. Italian organized programmes 2012 activity.
Figura 4. Proporzione di donne HPV-positive con citologia anomala. Donne di ogni età.Attività 2012 dei programmi organizzati.
ASC-US AGC (all ages) L-SIL (all ages) CMT H-SIL ASC-H (all ages)
The PPV of stand-alone HPV testing for high-grade CIN
was actually quite stable (except for an inverse correlation
to the previous screening activity) in different situations.20
Therefore, variability between areas is expected to be lower
than the variability in abnormal cytology in the entire pop-
ulation (which also reflects true variations in baseline risk).
Thus, the observed variability in cytology triage plausibly
reflects variability in the criteria of interpretation. Knowl-
edge that slides came from HPV-positive women probably
had a strong impact. Very high frequencies of cytological
abnormalities were also observed in early reports of pilot
projects.16,18 These data clearly show the need to train cy-
tologists and cytopathologists involved in the triage of
HPV-positive women.
Conflicts of interests: One of the authors, Paolo Giorgi Rossi, as
principal investigator in a study funded by the Italian Ministry
of Health, is in charge of leading negotiations with Hologic,
Roche Diagnostics, Qiagen, and Abbot in order to obtain rea-
gents for free or at lower costs.
HPV-based cervical cancer screening: first survey
82 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Trento
%
1
2
3
4
5
0
Figure 5. Proportion of women screened by HPV who are immediately referred to colposcopy because both HPV positive and judged to have abnormal cytology.
Figura 5. Proporzione di donne screenate per HPV che sono immediatamente inviate a colposcopia perché positive all’HPV e con citologia anomala.
NORTH CENTRE SOUTH and ISLANDS Each bar rapresents one programme, 2012 activity.
Left-hand bars refer to 35-64 year-old women, right-hand bars refer to all women.
0.5
*Roma G.
1.0
Este
1.3
Torino
1.4
Rovigo
1.6
Latina
1.7
Alta
Padovana
2.0
Adria
2.0
Avezzano
2.0
Lanciano
2.3
Valle
Camonica
2.3
Reggio
Emilia
2.4
Pescara
2.6
Padova
2.6
Molise
3.2
Teramo
3.2
Savonese
4.8
L. Gallo: ULSS 18 Rovigo
A. Del Sole: ULSS 19 Adria
L. Paterlini, C. Campari: AUSL Reggio Emilia
A. Iossa, M. Confortini: ISPO, Firenze
A. Barca: Regione Lazio
M.C. Tufi: ASL Roma G
P. Bellardini: ASL Latina
C. Fortunato: ASL Teramo
M. Minna: ASL Pescara
A. Macerola: ASL Av-Sulmona-L’Aquila
D. Caraceni: ASL Lanciano-Vasto-Chieti
N. Segnan, A. Gillio Tos: CPO Piemonte
L. Bonelli: IRCCS AOU San Martino-IST, Genova
A. Franco, E. Venturino: ASL2 Savonese
L. Pasquale, G. Luciano: ASL Valle Camonica Sebino
E. Polla, P. Dalla Palma: Ospedale S. Chiara, Trento
M. Zorzi: Registro tumori Veneto
C. Fedato: Regione Veneto
F. Zago: ULSS 12 Venezia
A. Pupo: ULSS 15 Alta Padovana
I. Simoncello: ULSS 16 Padova
A. Ferro: ULSS 17 Este-Monselice
HPV screening survey working group:
Gruppo di lavoro per la survey sullo screening con HPV:
HPV-based cervical cancer screening: first survey
83 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Naucler P, Ryd W, Tornberg S et al. Human papillomavirus and Pa-
panicolau tests to screen for cervical cancer. New Engl J Med
2007;357:1589-97.
2. Rijkaart DC, Berkhof J, Rozendaal L et al. Human papillomavirus test-
ing for the detection of high-grade cervical intraepithelial neopla-
sia and cancer: final results of the POBASCAM randomised con-
trolled trial. Lancet Oncol 2012;13(1):78-88.
3. Kitchener HC, Almonte M, Thomson C et al. HPV testing in combi-
nation with liquid-based cytology in primary cervicalscreening (ARTIS-
TIC): a randomised controlled trial. Lancet Oncol 2009;10:672-82.
4. Ronco G, Giorgi-Rossi P, Carozzi F et al., and the New Technologies
for Cervical Cancer screening (NTCC) Working group. Efficacy of hu-
man papillomavirus testing for the detection of invasive cervical can-
cers and cervical intraepithelial neoplasia: a randomised controlled
trial. Lancet Oncol 2010;11:249-57.
5. Sankaranarayanan R, Nene BM, Shastri SS et al. HPV screening for
cervical cancer in rural India. New Engl J Med 2009;360:1385-94.
6. Leinonen MK, Nieminen P, Lönnberg S et al. Detection rates of pre-
cancerous and cancerous cervical lesions within one screening
round of primary human papillomavirus DNA testing: prospective
randomised trial in Finland. BMJ 2012;345:e7789.
7. MayrandMH, Duarte-Franco E, Rodrigues I et al., and the Canadian
Cervical Cancer Screening Trial Study Group. Human papillomavirus
DNA versus Papanicolaou screening tests for cervical cancer. New
Engl J Med 2007;357:1579-88.
8. Ogilvie GS, Krajden M, van Niekerk DJ et al. Primary cervical cancer
screening with HPV testing compared with liquid-based cytology: re-
sults of round 1 of a randomised controlled trial – the HPV FOCAL
Study. Br J Cancer 2012;107:1917-24.
9. Ronco G, Dillner J, Elfström M et al. Efficacy of HPV-based screen-
ing for preventing invasive cervical cancer: follow-up of European
randomised controlled trials. Lancet 2014;383:524-32.
10. Ronco G, Biggeri A, Confortini M et al. HPV DNA based primary
screening for cervical camcer precursors. Epidemiol Prev 2012;3-
4(Suppl 1):e1-72.
11. Meijer CJ, Berkhof J, Castle PE et al. Guidelines for human papillo-
mavirus DNA test requirements for primary cervical cancer screen-
ing in women 30 years and older. Int J Cancer 2009; 124:516-20.
12. Ministero della salute. Piano nazionale della prevenzione 2010-
2012. Azione centrale prioritaria concernente la definizione di do-
cumenti tecnici di sintesi delle evidenze scientifiche a supporto
della programmazione, monitoraggio e valutazione degli interventi
di prevenzione oncologica nella popolazione a rischio. Documento
di indirizzo sull’utilizzo dell’HPV-DNA come test primario per lo
screening del cancro del collo dell’utero. [www.osservatoriona-
zionalescreening.it/sites/default/files/allegati/Screening.pdf]
13. Ronco G, Giubilato P, Carozzi F et al. and the Cancer Screening Sur-
vey working group. Extension of organized cervical cancer screen-
ing programmes in Italy and their process indicators, 2011-2012
activity. Epidemiol Prev 2015;3(Suppl 1):61-76.
14. Baussano I, Franceschi S, Gillio-Tos A et al. Difference in overall and
age-specific prevalence of high-risk human papilloma virus
infection in Italy: evidence from NTCC trial. BMC Infect Dis 2013;
13:238.
15. Giorgi Rossi P and the Middir Working Group. MIDDIR - Methods
for investments/disinvestments and distribution of health tech-
nologies in Italian Regions. 11th HTAi annual meeting, Washing-
ton DC, 16-18 June 2014.
16. Zorzi M, Del Mistro A, Farruggio A et al. Use of high-risk human
papillomavirus DNA test as the primary test in a cervical cancer
screening programme: a population-based cohort study. BJOG
2013;120:1260-67.
17. Del Mistro A, Frayle H, Ferro A et al. Cervical cancer screening by
high risk HPV testing in routine practice: results at one year recall
of high risk HPV-positive and cytology-negative women. J Med
Screen 2014;21:30-37.
18. Confortini M, Giorgi-Rossi P, Barbarino P et al. Screening for cer-
vical cancer with the human papilloma virus test in an area of cen-
tral Italy with no previous active cytological screening programme.
J Med Screen 2010;17:79-86.
19. Giorgi Rossi P, Bisanzi S, Paganini I et al. Prevalence of HPV high
and low risk types in cervical samples from the Italian general pop-
ulation: a population based study. BMC Infect Dis 2010; 10:214.
20. Giorgi-Rossi P, Franceschi S, Ronco G. HPV prevalence and accu-
racy of HPV testing to detect cervical intraepithelial neoplasia. Int
J Cancer 2012;130:1387-94.
84 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
hr-HPV testing in the management of women
with ASC-US+ and in the follow-up of women
with cytological abnormalities and negative colposcopy.
Recommendations of the Italian group for cervical
cancer screening (GISCi)
Test hr-HPV nella gestione delle donne con citologia ASC-US+
e nel follow-up delle donne con citologia anormale
e colposcopia negativa. Raccomandazioni del Gruppo italiano
per lo screening del carcinoma della cervice uterina (GISCi)
Francesca Maria Carozzi,1Anna Iossa,1Aurora Scalisi,2Mario Sideri,†Karin Louise Andersson,1Massimo Confortini,1
Annarosa Del Mistro,3Giovanni Maina,4Guglielmo Ronco,5Patrizio Raggi,6Maria Luisa Schiboni,7Marco Zappa,1Paolo Giorgi Rossi8,9
WWW.EPIPREV.IT
anno 39 (3) maggio-giugno 2015
e
&
p
1Istituto per lo studio e la prevenzione oncologica, Firenze
2Azienda sanitaria provinciale, Catania
3Istituto oncologico veneto IOV - IRCCS, Padova
4Azienda ospedaliera Città della scienza e della salute di Torino - Presidio S.Anna
5Centro di riferimento per l’epidemiologia e la prevenzione oncologica in Piemonte, Torino
6Azienda sanitaria locale, Viterbo
7Azienda ospedaliera San Giovanni - Addolorata, Roma
8Servizio interaziendale di epidemiologia, AUSL Reggio Emilia, IRCCS
9Arcispedale S. Maria Nuova, Reggio Emilia
Corresponding author: Francesca Maria Carozzi, f.carozzi@ispo.toscana.it
Management of women with abnormal screening tests: GISCi recommendations
85 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
No clear guidelines concerning the follow-up protocol after an
abnormal Pap test and negative colposcopy exist in Italy, and ac-
tual management of these cases is highly variable.
Purpose of this article is providing recommendations on the use
of hr-HPV testing in the management of colposcopies after an
abnormal cytology (either applied as primary screening test or
as triage of HPV-positive women) and in follow-up after col-
poscopy. The rational of recommendations is the high negative
predictive value for cervical intraepithelial neoplasia grade 2 or
more severe (CIN2+) of HPV testing for oncogenic human pa-
pillomavirus (hr-HPV).1,2 This makes it possible to reduce
and standardize follow-up controls.
Persistent infection with one of 12 high-risk HPV types is a ne-
cessary cause of invasive cervical cancer.3Thus, hr-HPV testing
can be used as a negative triage test to determine whether a wo-
man can be safely returned to routine screening4in the follow-
up of abnormal cytology and after a negative colposcopy.
MATERIALS AND METHODS
A literature review was carried out: European and American
guidelines were considered, along with good practice recom-
mendations from the most important scientific associations
and regulatory agencies.
Recommendations are based on the risk of harbouring a
CIN2+ (i.e., on PPV) by primary cytology result. The PPV of
cytology is highly variable between Italian screening pro-
grammes (2.8% to 52.7% for ASC-US or higher). Neverthe-
less, the difference between cytological categories is very large:
PPV for CIN2+ is <10% in women with ASC-US and L-SIL
cytology and >40% in women with ASC-H, and H-SIL cyto-
logy (figure 1).
For cytology classification we refer to the 2001 Bethesda sy-
stem.4«Second-level negative for CIN2+» means that no
CIN2+ was detected, either because histology was negative or
because no biopsy was taken as no colposcopic abnormality was
observed.
Abstract
Compared to spontaneous screening, an organized screening programme is characterized by the presence of protocols and rec-
ommendations for all stages including follow-up. Despite the availability of well-functioning screening programmes throughout
the country, the follow-up protocol after an abnormal Pap test and negative colposcopy is not clearly defined in Italy, and there
is no uniformity of indications.
HPV testing for oncogenic human papillomavirus (hr-HPV) has a high negative predictive value (NPV) and high positive predic-
tive value (PPV) for CIN2+ and its employment can reduce follow-up assessments.
In order to provide indications about the management of women with ASC-US+ and the follow-up of women with cytological
abnormalities and negative colposcopy, a literature analysis was carried out, taking into consideration European and American
guidelines and good practice recommendations from the most important scientific associations and regulatory agencies. GISCi
(Italian Group for Cervical Screening) drafted recommendations for the management of women with ASC-US, L-SIL, ASC-H,
AGC, and H-SIL until their return to the routine screening interval. This protocol can be applied not only in the management of
abnormal Pap smears in cytology-based programmes, but also in the management of abnormal Pap test triage after HPV posi-
tive test when HPV is the primary screening test. The protocols approved within the screening programmes must have an ex-
tensive consensus among all involved professionals, including any that women might meet outside the programme.
(Epidemiol Prev 2015; 39(3) Suppl 1: 84-90)
Keywords: cervical cancer screening, colposcopy, human papillomavirus, follow-up, Italy
Riassunto
Rispetto allo screening spontaneo, un programma di screening organizzato è caratterizzato dalla presenza di protocolli e rac-
comandazioni per tutte le fasi, incluso il follow-up. Nonostante l’ampia diffusione dei programmi di screening su tutto il terri-
torio, il protocollo di follow-up dopo Pap test anormale e colposcopia negativa in Italia non è chiaramente definito e non c’è
uniformità nelle indicazioni date dai programmi. Il test HPV per la ricerca di papillomavirus oncogeni ha un elevato valore pre-
dittivo negativo (NPV) e un elevato valore predittivo positivo (PPV) per CIN2+ e il suo utilizzo può ridurre i controlli di follow-
up. Al fine di fornire indicazioni sulla gestione delle donne con ASC-US+ e nel follow-up delle donne con citologia anormale e
colposcopia negativa è stata effettuata una analisi della letteratura, delle linee guida europee e americane e delle raccomanda-
zioni di buona pratica delle principali associazioni scientifiche. Il Gruppo italiano per lo screening del carcinoma della cervice ute-
rina (GISCi) ha prodotto le raccomandazioni per la gestione delle donne con ASC-US, L-SIL, ASC-H, AGC e H-SIL fino al loro ri-
torno al normale intervallo di screening. Questo protocollo può essere applicato non solo nella gestione del Pap test anormale
nello screening con Pap test primario, ma anche nella gestione del Pap test di triage anormale dopo hr-HPV test positivo, quando
HPV è il test di screening primario. I protocolli approvati nell'ambito dei programmi di screening devono avere un ampio con-
senso tra tutti i professionisti coinvolti, compresi coloro che potrebbero entrare in contatto con le donne al di fuori del programma.
(Epidemiol Prev 2015; 39(3) Suppl 1: 84-90)
Parole chiave: screening carcinoma cervice uterina, colposcopia, papillomavirus umano, follow-up, Italia
Management of women with abnormal screening tests: GISCi recommendations
86 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
MANAGEMENT OF WOMEN
WITH AN ABNORMAL SCREENING TEST
Atypical squamous cells of undetermined
significance (ASC-US)
ASC-US is the most common cytologic abnormality and entails
a low risk of CIN2+. In 2006, national guidelines and GISCi
recommended three possible management strategies: immediate
referral, repeat cytology at 1 year or hr-HPV triage.5
In 2011, a report of the English NHSCSP pilot study on hr-
HPV triage of women with ASC-US and L-SIL,6pointed out
that hr-HPV triage makes it possible to return about one third
of women with ASC-US to routine screening, with a conside-
rable reduction in colposcopies. The study also showed a good
acceptability of triaging to women.
The 2012 American Cancer Society Guidelines7recommend to
return women with ASC-US and negative hr-HPV to the nor-
mal screening interval, i.e., 3 years. For the management of
ASC-US/hr-HPV positive women who have a negative se-
cond-level assessment for CIN2+, American7and European gui-
delines8,9 provide two options: repeat an hr-HPV test after 12
months or repeat cytology after 6 and 12 months. American
guidelines7also recommend not to repeat hr-HPV testing ear-
lier than 12 months. Since 2005/2007, GISCi5has recom-
mended hr-HPV testing as one of the three possible options for
the management of ASC-US, and endorsed the use of hr-HPV
tests validated for screening. Data from the 2010 GISCi survey
showed that triage with hr-HPV has a PPV for CIN2+ greater
than the other two options (figure 1), and reduces variability
between centres.
Recommended management
If cytology is the primary test, triage by hr-HPV testing (HPV
triage) is recommended (figure 2). Women with ASC-US cy-
tology and negative hr-HPV should return to routine scree-
ning,10 while women with ASC-US and positive hr-HPV
should have colposcopy.
For women that are ASC-US/hr-HPV positive and second-
level negative for CIN2+, re-testing for hr-HPV test after 1
year is strongly recommended. If the hr-HPV test is negative,
return to normal screening is recommended. If it is positive,
colposcopy should be repeated. In the latter case, if the new se-
cond-level analysis comes out negative for CIN2+, women are
invited to repeat hr-HPV testing after 12 months. If the repeat
hr-HPV test is negative, women return to routine screening.
If the repeat hr-HPV test is positive, women are invited to re-
peat colposcopy and cytology.
Low-grade squamous intraepithelial lesions
hr-HPV triage is recommended for women with L-SIL cyto-
logy if age is ≥35 years, according to the GISCi 2005-2007 do-
cument.5Triage is not recommended in younger women. In
the Italian programmes that have adopted this approach, the
proportion of hr-HPV positive women is variable, and in
many cases high.11 This likely reflects different criteria for re-
porting cytology. Depending on the local situation, hr-HPV
triage could be proposed for older (i.e., above the age of 45 ye-
ars) women only.1
For women with L-SIL cytology and negative colposcopy, Eu-
ropean and American guidelines7-9 recommend repeating an
hr-HPV test after 1 year. If the test is negative, the woman re-
turns to routine screening; if it is positive the woman will be
referred to colposcopy.7English guidelines10 did not initially
provide protocols, pending the results of a pilot study. The pro-
tocol proposed at study6publication (2011) recommended, in
the case of negative 2nd level result for CIN, to return the wo-
man to routine screening while, in the case of CIN 1 without
treatment, repeating cytology in 12 months was recommended.
Recommended management
hr-HPV triage for L-SIL is recommended for programmes
where L-SIL cytology has a low PPV (<5-10 %) and after an
evaluation of the local proportion of hr-HPV-positive L-SIL
ASC-US
repeated
cytology
ASC-US
immediate
colposcopy
ASC-US HPV
triage
L-SIL ASC-H AGC H-SIL
%
10
20
30
40
50
60
70
0
Figure 1. ASC-US+ posi-
tive predictive value for
CIN2+. Survey ONS, 2009.
Figura 1. Valore predittivo
positivo per CIN2+ delle
ASC-US+. Survey ONS,2009.
Management of women with abnormal screening tests: GISCi recommendations
87 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
through a pilot study. For women ≥35 years of age and L-SIL-
hr-HPV negative, return to routine screening is recommended
(figure 2). For women ≥35 years and L-SIL-hr-HPV positive
and for women with L-SIL cytology and no HPV result, col-
poscopy is recommended.
If colposcopy is negative for CIN2+, the woman is invited for
an hr-HPV test after a year. If the test is negative, the woman
returns to routine screening. If it is positive, the woman is re-
ferred for colposcopy. If the second colposcopy is also negative
for CIN2+, the woman is invited to repeat an hr-HPV test at
12 months. In case this further hr-HPV test is positive, the wo-
man is referred to a new colposcopy and Pap test. This follow-
up protocol, which uses the hr-HPV test after an in-depth ana-
lysis of 2nd negative level for CIN2+, can be applied even
where there is no initial triage with hr-HPV.
Atypical glandular cells (AGC)
AGC is an uncommon cytology12 and is often associated with
benign conditions, such as reactive cellular changes or polyps.
In the literature, however, 9% to 38% of women with AGC are
reported to have CIN2+, and 3% to 17% to have an invasive
carcinoma.12
Atypia on glandular cells may affect endometrial as well as en-
docervical cells. European guidelines8,9 make a distinction
between «AGC, favour neoplasia or AIS» and «AGC not
otherwise specified (NOS)». For women older than 35 years,
in case of AGC favour neoplasia, a colposcopy with endo-
cervical sampling is indicated. Even if this colposcopy is ne-
gative for CIN2+, a diagnostic conisation is recommended in
this age group. In case of AGC NOS with negative colpo-
scopical findings, European guidelines recommend a Pap
test every 6 months for 2 years. American guidelines suggest
for both categories of AGC a colposcopy with endocervical
sampling. An endometrial sampling in all women over the age
of 35 years or those with clinical elements suggestive for
neoplastic pathology of the endometrium is also encouraged.
A negative hr-HPV test can be useful in identifying women
who have a greater risk of endometrial cancer rather than cer-
vical disease.7
Recommended management: initial workup
For women with AGC cytology, colposcopy is recommen-
ded; at the time of colposcopy hr-HPV testing is also recom-
mended: the hr-HPV test will assist in excluding an origin from
cervical glandular lesions in case of initial negative colposcopy
workup.
Figure 2. Management of
women with ASC-US cytology
and positive hr-HPV and L-SIL
with or without HPV triage.
Figura 2. Gestione delle don-
ne con citologia ASC-US e
HPV-hr positivo e L-SIL con o
senza HPV triage.
treatment
and follow-up
negative for CIN2+
colposcopy No. 1
negative for CIN2+
hr-HPV positive
colposcopy and Pap test
return to screening
hr-HPV negative
hr-HPV at 12 months
CIN 2+
hr-HPV positive
colposcopy No. 2
CIN 2+
ASC-US hr-HPV+ or L-SIL with or without triage
hr-HPV at 12 months
Management of women with abnormal screening tests: GISCi recommendations
88 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Atypical squamous cells, cannot exclude
high-grade squamous intraepithelial
lesion (ASC-H) and high-grade squamous
intraepithelial lesions (H-SIL)
Given their high positive predictive value (PPV), women with
ASC-H and H-SIL cytology should be referred directly to
colposcopy. While there is consensus for the management in
case of CIN2+ (excisional therapy), there is no uniformity of
indications in case of negative 2nd level workup for CIN2+.
American guidelines12 in case of H-SIL always recommend, in
addition to colposcopy, an examination of the cervical canal.
As for ASC-H with negative 2nd level for CIN2+, American
and European guidelines7-9 recommend a Pap test after 6 and
12 months or, alternatively, hr-HPV testing after 12 months.
If both Pap tests or the hr-HPV test are negative, guidelines
suggest a return to routine screening (1 or 2 years at the time
of the 2006 guidelines expanded to 2/3 in 2010). In case of Pap
test ≥ASC or positive hr-HPV a new colposcopy is recom-
mended. For H-SIL with negative 2nd level for CIN2+, gui-
delines offer three options:
■
combined cytology and colposcopy at 6 and 12 months with
return to screening in case of negativity of both tests;
■
excisional therapy;
■
review of cytology and histology (recommended by European
guidelines).
Recommended management for ASC-H and H-SIL
Women with ASC-H or H-SIL should be referred to colposcopy.
If colposcopy is positive for CIN2+, excisional treatment must
be provided. The management of women with ASC-H and
H-SIL and negative 2nd level for CIN2+ differs according to
squamo- columnar junction visibility. To exit follow-up and re-
turn to routine screening, in any case, two negative colposcopies,
two negative hr-HPV tests, and a negative cytology are needed.
Figure 3. Management of women with cytology ASC-H and H-SIL. / Figura 3. Gestione delle donne con citologia ASC-H e H-SIL.
treatment and post-treatment
follow-up
negative for CIN2+ and visible SCJ*
ASC-H: slide revision or LEEP
hr-HPV and Pap test
at 12 months
return to screening hr-HPV and Pap test
at 12 months colposcopy No. 3
after 6 months
colposcopy,
hr-HPV and Pap
test
LEEP
or
excisional
therapy
CIN 2+
ASC-H or H-SIL
H-SIL: - LEEP
- or repeat colpo at short interval
- or endocervical sampling
colposcopy No. 2 CIN2+
colposcopy n. 2 negative for CIN2+
and hr-HPV negative
and Pap test negative or ≤L-SIL
hr-HPV and Pap test at 12 months
negative hr-HPV and Pap
test ASC-US or L-SIL
negative hr-HPV
and Pap test negative
positive hr-HPV and Pap test
ASC-US or L-SIL
Pap test ASC-H, H-SIL or AGC
regardless hr-HPV result
colposcopy n. 2 negative for CIN2+
and hr-HPV positive or Pap test >L-SIL
after 6 months: colposcopy, hr-HPV
and Pap test
colposcopy No. 1
negative for CIN2+ and not visible SCJ*
*SCJ: squamo-columnar junction
Management of women with abnormal screening tests: GISCi recommendations
89 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
If, during the first colposcopy, the squamo-columnar junction
is visible and no CIN2+ is identified histologically, the woman
is invited after 6 months for a new colposcopy (figure 3, p. 88),
an hr-HPV test and a Pap test, recommended especially in case
of initial ASC-H cytology:
■
if after 6 months the 2nd level in-depth analysis turns out
to be positive for CIN2+, the woman should be referred to
treatment;
■
if after 6 months the 2nd level in-depth analysis remains ne-
gative, i.e., histology does not identify any CIN2+, colposco-
py does not locate suspicious areas on which to perform a biop-
sy, and hr-HPV test and Pap test are negative or ≤L-SIL, the wo-
man should be asked to repeat an hr-HPV test and a Pap test
after 12 months. If after 12 months there is H-SIL, ASC-H or
AGC cytology, the woman should be referred to colposcopy,
regardless of the hr-HPV test result. If after 12 months the hr-
HPV test remains negative and the Pap test is negative, the wo-
man can return to routine screening. If after 12 months the hr-
HPV test is confirmed negative but the Pap test shows ASC-US
or L-SIL, hr-HPV test and Pap test repeat at 12 months are re-
commended. If after 12 months the hr-HPV test turns out po-
sitive and the Pap test is negative or ASC-US or L-SIL cyto-
logy, the woman is recommended to repeat an hr-HPV test and
a Pap test after 12 months;
■
if after six months the 2nd level in-depth analysis is negati-
ve for CIN2+ but the hr-HPV test is positive, the gynaecolo-
gist can either (figure 3):
a) schedule a diagnostic LEEP (loop electrosurgical excision pro-
cedure) or excisional therapy
or:
b) repeat colposcopy, hr-HPV, and PAP test at 6 months.
If, during the first colposcopy, the squamo-columnar junction
is not visible and no CIN2+ is detected, different options can
be considered (figure 3) on the basis of initial cytology. For H-
SIL there are 3 options:
■
repeat colposcopy after a short interval;
■
perform endocervical sampling;
■
perform a diagnostic LEEP.
For initial ASC-H it is suggested to perform diagnostic LEEP
or to review the slide. If the review is negative, or ASC-US, or
L-SIL, an hr-HPV test should be repeated after one year. If the
review confirms ASC-H cytology, then endocervical sampling
is carried out.
DISCUSSION AND FUTURE PERSPECTIVES
Organized screening programmes are more effective than
opportunistic activity. The availability and quality of field and
laboratory facilities for screening and diagnostic follow-up, as
well as the available treatment facilities, are key elements of
any screening programme. Monitoring the management of
each patient with an abnormal screening result is of crucial
importance.13
Pap smear testing is widely available and has shown high effi-
cacy in reducing cervical cancer incidence. Nevertheless, every
year in Italy many new cervical cancers are diagnosed (2,200
in 2012), and 5-year relative survival rates have only slightly in-
creased, from 64% in 1990-1994 to 67% in 2000-2004.14
Reasons for Pap-test-based screening failure include lack of Pap
testing, failure of the Pap smear to detect an abnormality, and
lack of adequate follow-up after an abnormal Pap test.
Compared to spontaneous activity, organized screening is cha-
racterized by protocols and guidelines for all its stages, including
follow-up. Protocols to be applied within screening program-
mes must have an extensive consensus among all involved pro-
fessionals, including those that women might meet outside the
programme. It is of utmost importance to verify compliance to
follow-up protocols. In 2014, GISCi conducted a specific sur-
vey to evaluate the workload induced by follow-up after a ne-
gative colposcopy. Evidence to set forth the optimal manage-
ment of women with negative colposcopy after abnormal
cytology or with CIN1 is poor. A recent paper2confirmed that
hr-HPV testing is able to identify, among women with cytology
≥ASC-US and no evidence of high-grade disease, those at risk
of developing CIN2+. Performing hr-HPV testing within 1 year
could avoid 30% of follow-up colposcopies in women with
ASC-US and 33% in selected women with the remaining cy-
tological abnormalities (ASC-H, L-SIL, H-SIL, AGC).15 We
stress that only tests for the DNA of oncogenic HPV types, va-
lidated according to European guidelines as for sensitivity and
specificity for high-grade lesions, should be applied, even when
the test is used for follow-up.
Determining which hr-HPV-positive women are at future cli-
nical risk and identifying robust markers of disease progression
is the challenge for the future. Follow-up studies of women ma-
naged by HPV genotyping, p16 immunostaining, and me-
thylation markers are needed to establish their role in the ma-
nagement of cervical abnormalities. New HPV DNA tests,
including direct partial genotyping for types 16 and 18,15 or
p16INK4a,16 have also been shown to be promising triage test
methods. Hence, with the introduction of new biomarkers for
cervical cancer, more screening options will become available.
As the number and sophistication of tools applied to cervical
cancer prevention continue to increase, the complexity of ma-
nagement promises to grow.
Conflicts of interests: none declared
Management of women with abnormal screening tests: GISCi recommendations
90 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Ronco G, Cuzick J, Segnan N et al. HPV triage for Low grade (L-SIL)
cytology is appropriate for women over 35 in mass cervical cancer
screening using liquid based cytology. Eur J Cancer 2007; 43:476-80.
2. Carozzi F, Visioli CB, Confortini M et al. hr-HPV testing in the fol-
low-up of women with cytological abnormalities and negative col-
poscopy. Br J Cancer 2013;109(7):1766-74.
3. Bouvard V, Baan R, Straif K et al. WHO International Agency for
Research on Cancer Monograph Working Group. A review of hu-
man carcinogens-Part B: biological agents. Lancet Oncol 2009;
10:321-22.
4. GISCi Gruppo italiano screening del cervicocarcinoma. Documento
operativo GISCi per l’applicazione nei programmi di screening del
sistema Bethesda 2001. April 2006, updated June 2009.
[www.gisci.it/documenti-gisci]
5. GISCi Gruppo italiano screening del cervicocarcinoma. Utilizzo
del test HPV HR nel triage delle diagnosi citologiche di ASC-US e
delle diagnosi di LSIL in donne con più di 35 anni nonché nel fol-
low-up delle lesioni CIN2-3 come indicatore di rischio di recidiva.
April 2005, updated 2012. [www.gisci.it/documenti-gisci]
6. Kelly RS, Patnick J, Kitchener HC et al. HPV testing as a triage for
borderline or mild dyskaryosis on cervical cytology: results from the
Sentinel Sites study. Br J Cancer 2011;105:983-88.
7. Saslow D, Solomon D, Lawson HW et al. American Cancer Soci-
ety, American Society for Colposcopy and Cervical Pathology, and
American Society for Clinical Pathology. Screening guidelines for
the prevention and early detection of cervical cancer. Am J Clin
Pathol 2012;137:516-42.
8. Arbyn M, Anttila A, Jordan J et al. European Commission. European
guidelines for quality assurance in cervical cancer screening. Second
edition. European Communities, Luxembourg 2008.
9. Arbyn M, Anttila A, Jordan J et al. European Guidelines for
Quality Assurance in Cervical Cancer Screening. Second edition-
summary document. Ann Oncol 2010;21:448-58.
10. NHS Cancer Screening Programmes. Colposcopy and programme
management. Guidelines for the NHS Cervical screening program
(second edition). NHSCSP Publication No. 20, May 2010.
11. Giorgi Rossi P, Chini F, Bisanzi S et al. Distribution of high and low
risk HPV types by cytological status: a population based study from
Italy. Infect Agent Cancer 2011;6(1):2.
12. Wright TC Jr, Massad LS, Dunton CJ et al. 2006 consensus guide-
lines for the management of women with cervical cytological ab-
normalities. JAMA 2002;287:2120-29.
13. Nygård M. Screening for cervical cancer: when theory meets real-
ity. BMC Cancer 2011;11:240.
14. Altavilla G, Bernardo G, Bracarda S et al. I numeri del cancro in Ita-
lia. Brescia, Intermedia editore 2012.
15. Castle PE, Stoler MH, Wright Jr et al. Performance of carcinogenic
human papillomavirus (HPV) testing and HPV16 or HPV18 geno-
typing for cervical cancer screening of women aged 25 years and
older: a subanalysis of the ATHENA study. Lancet Oncol 2011;
12:880-890.
16 Carozzi F, Gillio-Tos A, Confortini M et al. Risk of high-grade cervi-
cal intraepithelial neoplasia during follow-up in HPV-positive
women according to baseline p16-INK4A results: a prospective
analysis of a nested substudy of the NTCC randomised controlled
trial. Lancet Oncol 2013;14(2):168-76.
Colorectal cancer
screening
Colorectal cancer screening: 2011-2012 survey
92 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Theoretical or potential or nominal extension: percentage of subjects involved in a screening programme out of the
total female population in the 50-69 age range resident in the area covered by an organized screening programme.
Actual extension or Extension of invitations: percentage of subjects involved in a screening programme out of the
total female population in the 50-69 age range who actually received an invitation to screening during the analyzed
period.
Compliance with invitation or Crude attendance: number of respondents out of the total number of invited subjects
minus undelivered invitations.
Adjusted attendance: number of respondents out of the total number of invited women excluding undelivered
invitations and subjects with a recent test (FIT or total colonoscopy, according to the local protocols for exclusion).
Positivity rate: percentage of subjects with a positive FIT (FS) out of the total number of attendees.
Attendance to colonoscopy assessment: number of subjects attending colonoscopy out of the total number of subjects
with a positive FIT (FS).
Complete colonoscopy rate: number of subjects with a complete colonoscopy (i.e., with caecal intubation), including
repeat colonoscopies, out of the number of subjects who underwent a colonoscopy.
Proportion of complications after colonoscopy: number of subjects with a complication that caused admission to
hospital within 30 days after colonoscopy out of the number of subjects who underwent a colonoscopy.
Detection rate: number of subjects with a screen-detected lesion out of 1,000 screened subjects. The detection rate is
calculated separately for carcinoma, advanced adenoma (i.e., an adenoma with a diameter ≥1 cm, with villous/tubulo-
villous type or with high-grade dysplasia) and non-advanced adenoma (an adenoma without the characteristics of
advanced adenomas).
Positive predictive value (PPV) of FIT+ (FS+) at colonoscopy: number of subjects with a diagnosis of carcinoma or
advanced adenoma, as a proportion of FIT+ (FS+) subjects that underwent colonoscopy.
Piemonte: Nereo Segnan, Centro prevenzione oncologia – CPO
Piemonte, Torino
Provincia Autonoma di Bolzano: Antonio Fanolla, Assessorato
alla sanità, Osservatorio epidemiologico, Provincia autonoma
di Bolzano, Bolzano
Provincia Autonoma di Trento: Sivano Piffer, Giovanni De Pretis,
Osservatorio epidemiologico, Azienda provinciale per i servizi
sanitari, Trento
Puglia: Vincenzo Pomo, Cinzia Annatea Germinario, Agenzia
regionale sanità, Regione Puglia, Bari
Sardegna: Pierina Thanchis, Assessorato dell’igiene e sanità
e dell’assistenza sociale, Regione Sardegna, Cagliari
Toscana: Paola Mantellini, Istituto per lo studio e la prevenzione
oncologica, Regione Toscana, Firenze
Umbria: Mariadonata Giaimo, Direzione regionale salute, coesione
sociale e società della conoscenza, Regione Umbria, Perugia
Valle D’Aosta: Gabriella Furfaro, Servizio dipendenze patologiche,
salute mentale e promozione della salute, Aosta
Veneto: Chiara Fedato, Registro tumori del Veneto, Padova
Abruzzo: Tamara Agostini, Direzione politiche della salute, Regione
Abruzzo, Pescara
Basilicata: Vincenzo Barile, Angelo Sigillito, Sergio Schettini, AO
San Carlo, Potenza
Calabria: Liliana Rizzo, Dipartimento Tutela della salute e politiche
sanitarie, Regione Calabria, Catanzaro
Campania: Renato Pizzuti, Osservatorio epidemiologico regionale,
Assessorato alla sanità, Regione Campania, Napoli
Emilia-Romagna: Carlo Naldoni, Assessorato alle politiche
per la salute, Regione Emilia-Romagna, Bologna
Friuli-Venezia Giulia: Nora Coppola, Direzione centrale salute,
integrazione socio sanitaria, politiche sociali e famiglia, Regione
Friuli-Venezia Giulia, Trieste
Lazio: Alessandra Barca, Lazio sanità, Agenzia di sanità pubblica,
Roma
Liguria:Luigina Bonelli, Gabriella Paoli, Istituto nazionale
per la ricerca sul cancro, Genova
Lombardia: Direzione generale salute, Regione Lombardia, Milano
Marche: Lucia Di Furia, Servizio salute, Regione Marche, Ancona
Molise: Ospedale Cardarelli, Regione Molise, Campobasso
Italian colorectal cancer screening survey group:
Gruppo italiano survey screening colorettale:
Glossary
Screening for colorectal cancer in Italy:
2011-2012 survey
Screening del cancro colorettale in Italia:
survey 2011-2012
Manuel Zorzi,1Filippo Da Re,2Paola Mantellini,3Carlo Naldoni,4Priscilla Sassoli de’ Bianchi,4
Carlo Senore,5Anna Turrin,6Carmen Beatriz Visioli,3Marco Zappa3and the Italian colorectal cancer
screening survey group
Abstract
We present the main results of the 2011-2012 survey of the Italian screening programmes for colorectal
cancer carried out by the National centre for screening monitoring (Osservatorio nazionale screening,
ONS) on behalf of the Ministry of Health.
By the end of 2012, 112 programmes were active, of which 11 had been activated during 2012 and
4 during 2011. The national theoretical extension increased from 66% of Italians aged 50-69 years
residing in areas covered by organized screening programmes in 2010 to 73.7% in 2012. The major-
ity of programmes employ the fecal immunochemical test (FIT), while some have adopted flexible sig-
moidoscopy (FS) once in a lifetime and FIT for non-responders to FS.
Overall, about 7,744,000 subjects were invited to undergo FIT, 53.1% of those to be invited within
the two years. The adjusted attendance rate was 47.1% and 3,531,937 subjects were screened. Large
differences in the attendance rate were observed among regions. Positivity rate of FIT programmes was
5.2% at first screening (range: 1.0-12.4%) and 4.0% at repeat screening (range: 3.4-6.4%). The av-
erage attendance rate to total colonoscopy (TC) was 81.2% and in two regions (Molise and Campa-
nia) it was lower than 70%. Completion rate for total colonoscopy (TC) was 91%. Among the
1,316,327 subjects attending screening for the first time, the detection rate (DR) per 1,000 screened
subjects was 2.0 for invasive cancer and 9.1‰ for advanced adenomas (AA, adenomas with a diam-
eter ≥1 cm, with villous/tubulo-villous type or high-grade dysplasia). As expected, the corresponding
figures in the 2,215,610 subjects at repeat screening were lower (1.0‰ and 6.8‰ for invasive cancer
and AA, respectively). Many programmes reported some difficulties in guaranteeing TC in the appro-
priate time frame to FIT+ subjects: in 15% of cases the waiting time was longer than two months.
Ten programmes in 2011 and eight in 2012 employed FS as the screening test: 24,549 subjects were
screened in the two years, with an attendance rate of 24.5%. Overall, 85.9% of FSs were classified
as complete. Overall, TC referral rate was 9.8% and the DR per 1,000 screened subjects was 3.0 and
48.2 for invasive cancer and AA, respectively.
(Epidemiol Prev 2015; 39(3) Suppl 1: 93-107)
Keywords: screening, colorectal cancer, national survey, faecal immunochemical test, flexible sigmoidoscopy, Italy
WWW.EPIPREV.IT
93 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
1Registro tumori del Veneto,
Padova
2Settore promozione
e sviluppo igiene e sanità
pubblica, Regione Veneto,
Venezia
3Istituto per lo studio
e la prevenzione oncologica
(ISPO), Firenze
4Assessorato alle politiche
per la salute, Regione
Emilia-Romagna, Bologna
5Centro per la prevenzione
oncologica, CPO Piemonte
e Ospedale San Giovanni
Battista, Torino
6Istituto oncologico veneto
IRCCS, Padova
Corrispondenza
Manuel Zorzi
manuel.zorzi@regione.veneto.it
Riassunto
Presentiamo i dati nazionali di attività dei programmi di screening del carcinoma colorettale rela-
tivi al biennio 2011-2012. A fine 2012 erano attivi in Italia 112 programmi, di cui undici attivati
nel corso del 2012 e quattro attivati nel 2011. In particolare, sono stati attivati: un programma in
Puglia e il programma della provincia autonoma di Bolzano, due nuovi programmi in Lazio, due in
Abruzzo, uno in Campania, cinque in Sicilia e tre in Sardegna. L’estensione teorica nazionale del-
Colorectal cancer screening: 2011-2012 survey
94 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
This paper presents the data from the survey carried out by the
National centre for screening monitoring (Osservatorio
nazionale screening, ONS) on behalf of the Ministry of Health,
regarding the activities performed by Italian screening pro-
grammes for colorectal cancer during 2011-2012. The previ-
ous surveys are available at the ONS website.1
Important differences prevail among colorectal cancer screen-
ing programmes in Italy. The main difference regards the type
of screening test performed. While the majority of programmes
employ the fecal immunochemical test (FIT), some (nearly re-
stricted to one region, Piemonte) have adopted flexible sig-
moidoscopy (FS) once in a lifetime and FIT for non-respon-
ders to FS (figure 1). Moreover, FIT programmes have different
targets as far as age is concerned. Invitation to attend screen-
ing starts at the age of 50 years; whereas the maximum age is
69 or 70 years in most programmes, in a number of pro-
grammes it is as high as 74 or 75 years. FS programmes invite
a single cohort of subjects aged 58-60.
All FIT programmes are set to invite their target population by
mail every 2 years to undergo a 1-time immunochemical FIT,
without any dietary restriction. Quantitative haemoglobin
analysis is performed by automated instruments using the 100
ng Hb/ml threshold to determine positivity (80 ng Hb/ml in
a few programmes). People with a negative FIT are notified of
their results by mail and they are advised to repeat screening
2 years later. Non responders to the first invitation are mailed
a reminder, usually within 6 months. Subjects with a positive
screening test are contacted to undergo a total colonoscopy
(TC) or, when a complete colonoscopy is not possible, a dou-
ble-contrast barium enema X-ray or a colonography (virtual
colonoscopy). Colonoscopies are usually performed at an en-
doscopic referral centre, during dedicated sessions. Patients
with screen-detected neoplasms are referred to surgery or en-
doscopy, and then enrolled in a follow-up programme.
In 2007, the Italian group for colorectal cancer screening
(Gruppo italiano screening mammografico, GISCoR) pub-
lished an Operative report of quality indicators for the evaluation
of colorectal cancer screening programmes. For each indicator
the reference standards (acceptable, desirable) are provided.
Table 1 (p. 96) shows the indicators and standards utilized in this
paper. The operative report is available at the ONS website.2
DATA COMPLETENESS
Only 44% of the 215 questionnaires collected in 2011-2012
provided complete data (31% in 2011; 56% in 2012). The
items with the lowest level of completeness were screen-de-
tected lesions and surgery: time to surgical treatment, stage at
diagnosis, kind of treatment (endoscopic vs surgical). How-
ever, some programmes (N=7) were unable to provide even
baseline data.
EXTENSION AND COMPLIANCE
Programmes activated as of 31.12.2012
During 2011-2012, 15 new programmes were launched, 12 of
which in the South of Italy and Islands (figure 1).
As of 31st December 2012, 112 programmes were active in all
regions (table 2, p. 97). The vast majority of programmes
(N=104) employ the fecal immunochemical test (FIT), while
eight have adopted flexible sigmoidoscopy (FS) once in a life-
time, and FIT for non-responders to FS. In 2012, 7 pro-
grammes, mainly from the South of Italy and Islands, were sus-
pended.
The results of FIT programmes are reported in the following sec-
tions; data of FS programmes are presented in a specific section.
In order to describe the national situation, it is necessary to
lo screening è passata dal 66% della popolazione eleggibile di età compresa tra i 50-69 anni nel 2010 al 72,3% a fine 2012.
Complessivamente, nel 2011 e 2012 sono state invitate allo screening con la ricerca del sangue occulto fecale immunochimico
(SOF) 7.744.295 persone, pari al 53,1% della popolazione target da invitare nel biennio. I soggetti che nel 2011-2012 hanno
eseguito il SOF sono stati 3.531.937, con un’adesione corretta all’invito del 47,1%, con notevoli differenze tra Regioni.
La proporzione di positivi è stata del 5.2% nei soggetti al primo esame di screening (range: 1,0-12,4%) e del 4,0% agli esami
successivi (range: 3,4-6,4%). L’adesione alla colonscopia delle persone con SOF+ è stata dell’81,2%, con valori inferiori al 70%
in sole due Regioni (Molise e Campania). Più del 95% dei soggetti ha avuto una colonscopia completa e/o completata da un ul-
teriore esame di approfondimento.
Tra i 1.316.327 soggetti al primo esame di screening, il tasso di identificazione dei carcinomi è stato del 2,2 ogni 1.000 screenati
e quello degli adenomi avanzati del 10,3‰. I tassi di identificazione sono maggiori nei maschi rispetto alle femmine e aumentano
progressivamente con l’età in entrambi i sessi. Come atteso, tassi di identificazione più bassi (1,0‰ e 6,8‰ per carcinomi e ade-
nomi avanzati, rispettivamente) sono stati registrati nei 2.215.610 soggetti presentatisi a episodi di screening successivi al primo.
Molti programmi hanno riportato serie difficoltà a garantire in tempi brevi la colonscopia in caso di positività al SOF: circa un
sesto delle persone ha dovuto attendere più di due mesi (15%).
Dieci programmi nel 2011 e otto nel 2012 hanno proposto come test di primo livello la rettosigmoidoscopia (RS) a singole coor-
ti di età (58/60enni). Nel biennio hanno esaminato complessivamente 24.549 persone, con un’adesione corretta all’invito del
24,5%. E’ stato classificato come completo l’85,9% delle RS. Sono stati inviati ad approfondimento colonscopico il 9,8% degli
screenati e sono stati diagnosticati 3,0 carcinomi e 48,2 adenomi avanzati ogni 1.000 screenati.
(Epidemiol Prev 2015; 39(3) Suppl 1: 93-107)
Parole chiave: screening, carcinoma colorettali, survey nazionale, sangue occulto fecale, rettosigmoidoscopia, Italia
simplify the variability of the target population among the pro-
grammes, by narrowing the analysis to a homogeneous age
group. Therefore, we provide the data related only to subjects
aged 50-69 years that are common to all FIT programmes and
constitute the real target population of most of them.
Theoretical extension
Theoretical extension refers to eligible subjects residing in ar-
eas covered by organized screening programmes.
According to the National institute of statistics (Istat), at the
beginning of 2012 approximately 14,718,125 people aged 50-
69 years were living in Italy.3The number of subjects residing
in areas where an organized screening programme was active
was 10,272,496, with a national theoretical extension of
73.3% (table 2), more than eight points higher than that ob-
served in 2011 (64.9%). Compared to the previous years, the
northern and central regions were almost completely covered
by screening programmes, while in the South of Italy and Is-
lands theoretical extension increased to 45.2% (compared to
29% in 2010), notwithstanding the discontinuation of some
programmes.
In particular, programmes on a regional-scale basis were acti-
vated in Emilia-Romagna, Friuli-Venezia Giulia, Liguria,
Lombardia, Marche, Molise, Piemonte, Toscana, Umbria,
Valle d’Aosta, Veneto, Trento, and Bolzano.
Extension of invitations
We define extension of invitations as the proportion of the res-
ident population who was sent a screening invitation during
the study period.
During 2011-2012, some 7,744,295 subjects were invited to
attend a screening programme, accounting for 53.1% of the
Italian resident population aged 50-69 years to be invited in
the biennium (table 3, p. 98). Extension showed a clear trend
across the country, with the highest value in the North
(82.5%) and the lowest in the South of Italy and Islands
(12.2%). While some regions confirmed the full capacity
reached in the previous years, other regions reported low
levels, due either to the recent activation of many programmes
or to the chronic difficulty of many programmes in ensuring
the necessary number of invitations.
If we restrict analysis to the areas with ongoing pro-
grammes, the extension of invitations was 77.7%, higher in
the North (92.0%), intermediate in the Centre (73.7%),
and lower in the South of Italy and Islands (35.2%). The
most recent programmes reported a lower performance
(46.4%; 10th percentile: 9.7%) than those that had been
activated before 2007 (94.1%; 10th percentile: 72.5%)
(table 4, p. 98).
Overall, 63.1% of programmes reached GISCoR’s acceptable
standard of >80% (85% of programmes that started by 2007,
42.6% of those that started by 2007-2009, and 27.3% of the
others). Intra-regional variability, illustrated in table 3 through
the percentiles for the regions with at least four programmes,
was high in all but a few regions, where all programmes
reached high levels.
Colorectal cancer screening: 2011-2012 survey
95 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Figure 1. Colorectal cancer screen-
ing programmes: first-level test,tar-
get population, and programmes
started in 2011-2012.
Figura 1. Programmi di screening
colorettale: test di primo livello, po-
polazione target e programmi atti-
vati nel 2011-2012.
FIT 50-69/74 years
FIT started in 2011-2012
FS 58/60 + FIT 59-69 years
Compliance with invitation
We report data on adjusted compliance, calculated as the pro-
portion of subjects invited to attend screening (minus those
with a wrong address and those excluded after invitation for a
recent test) who underwent a screening test.
Overall, about 3,351,937 people were screened with FIT in
2011-2012. Adjusted compliance (47.1%) slightly decreased
compared to the 48% rate observed in 2010 (table 3). Adjusted
compliance was higher in the northern (52%) and central re-
gions (40.6%), while in the South of Italy and Islands it was
lower (28.6%).
The analysis of compliance by region shows a high inter-regional
variability, with values ranging from 13.7% in Campania to
67.7% in Valle d’Aosta (table 3). Moreover, a high intra-regional
variability in almost all regions must be highlighted.
The 10th percentile (24%) is clearly insufficient to guarantee suit-
able coverage of the population and, consequently, efficiency of
a screening programme. Overall, 57.1% of programmes reached
the acceptable GISCoR standard (>45%) (table 4).
As was the case for extension, attendance was likewise greater
in programmes that started before 2007 (50.8%; 10th per-
centile: 40.3%) compared to those that started after 2009
(27.6%; 10th percentile: 11.8%), independently of geo-
graphical area.
This result in part depends on the higher proportion of sub-
jects that have never been invited that characterizes recent
programmes. The attendance rate of subjects invited for the
first time was 34.3%, that of those who had already responded
to previous invitations was 82.5%, while 17.8% of subjects
who had never responded to previous invitations responded to
a new invitation during 2011-2012.
DIAGNOSTIC INDICATORS
The most important diagnostic indicators (positivity rates,
detection rates, positive predictive values) are strongly influ-
enced by the underlying frequency of the disease in the
screened population. Colorectal cancer and pre-cancerous le-
sions are more frequent in males than females, and progres-
sively increase with age.4Moreover, the disease is more fre-
quently detected in subjects at first screening test (prevalence
round) than in those at repeat tests (incidence round).
Therefore, these indicators are presented separately for subjects
at first and repeat screening tests, as well as by gender and five-
year age group. Subjects screened in newly activated pro-
Colorectal cancer screening: 2011-2012 survey
96 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Indicator Standard
acceptable desirable
actual extension
>80% >90%
compliance to invitation >45% >65%
positivity rate FIT: first test: <6% FIT: first test: <5%
repeat tests: <4.5% repeat tests: <3.5%
FS: <8% FS: <6%
inadequate screening tests FIT: <1%
FS: <10% FS: <5%
attendance to further FIT: >85% FIT: >90%
assessment FS: >90% FS: >95%
complete FS rate >85% >90%
complete TC rate >85% >90%
detection rate FIT FIT
carcinoma first test: >2.0‰ carcinoma first test: >2.5‰
repeat tests: >1.0‰ repeat tests: >1.5‰
adv. adenoma first test: >7.5‰ adv. adenoma first test: >10‰
repeat tests: >5.0‰ repeat tests: >7.5‰
FS FS
carcinoma >3.0‰ carcinoma >4.0‰
adv. adenoma >35‰ adv. adenoma >40‰
detection rate of adenomas at FS males
>10%
males
>15%
females
>5%
females
>10%
PPV of FIT at colonoscopy first test >25% first test >30%
for advanced adenoma or carcinoma repeat tests >15% repeat tests >20%
PPV of FS at colonoscopy
>7% >10%
for proximal advanced adenoma
delay between FIT screening >90% within 21 calendar days
>
90% within 15 calendar days
and negative result
delay between the call >90% within 30 calendar days
>
95% within 30 calendar days
for assessment
and the assessment procedure
proportion of screen-detected <30% <20%
cancers in stage III+
FIT: faecal immunochemical test; FS: flexible sigmoidoscopy; TC: total colonoscopy; PPV: positive predictive value.
Adaptedfrom: Zorzi M et al. Indicatori di qualità perla valutazione dei programmi di screening dei tumori colorettali. Epidemiol Prev 2007;6 (Suppl 1):1-56.
Table 1. Indicators and refe-
rence standards.
Tabella 1. Indicatori e stan-
dard di riferimento.
grammes all undergo first screening, while in the older pro-
grammes the proportion of subjects at repeat screening pro-
gressively increases. Moreover, while subjects at first screening
test are younger (47.4% were 50-54 year old in 2012), those
at repeat screening are mainly distributed in the older age
classes (65-69 years old: 30%; 50-54 years old: 15.8%).
The mean values of these indicators by region are standardized
by age and gender, using the national mean as standard pop-
ulation. The data refer to 3,531,937 subjects screened during
2011-2012 for which data are available; of these 1,316,327
(37%) underwent first screening and 2,215,610 (63%) sub-
sequent examinations.
Positivity rates
In subjects at first screening, the proportion of positive FIT was
5.2%, with quite homogeneous values among the mean re-
gional values of the regions with a significant number of
screens (table 5, p. 99). The 10th and 90th percentile of pos-
itivity rates reported by the programmes were 3.7% and 6.6%,
respectively. Outlier values were observed in programmes with
a few number of screened subjects and in some of the recently-
activated programmes.
In subjects at repeat screening, the proportion of FIT+ was
4.0%, with a higher homogeneity between programmes (10th-
90th: 3.3%-5.1%). Seventy-six percent of programmes met the
acceptable standard at the first (<6%) exam and sixty-six at re-
peat examination (<4.5%).
As shown in figure 2 (p. 99), the proportions of positive results
were higher in males both at first and repeat examinations, and
progressively increased with age, particularly at first screening
test.
Inadequate tests
Inadequate tests are essentially due to an incorrect sampling by
the subject.
During 2011-2012, 95% of programmes reported a propor-
tion of inadequate FITs lower than the standard (<1%). Over-
all, the national mean value was 0.3%.
Colorectal cancer screening: 2011-2012 survey
97 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Table 2. Main data of FIT programmes, 50-69 year-old subjects, by region.Years 2011-2012.
Tabella 2. Dati principali dei programmi di screening colorettale, soggetti 50-69enni,per Regione. Anni 2011-2012.
Region Programmes1Total resident Subjects residing in Theoretical Theoretical Coverage
subjects areas covered by a extension extension 2011-2012
programme in 2012 2011 2012
(N)2(N) (%)3(%)3(%)4
Abruzzo 0 / 2 324.572 176.812 0.0 54.5 0.1
Alto Adige* 0 / 1 114.793 114.793 0.0 100 0.9
Basilicata 1 / 0 139.899 0 59.3 0.0 7.0
Calabria 2 470.890 129.729 14.7 27.5 1.8
Campania 3 / 2 1.333.753 299.315 26.8 22.4 1.6
Emilia-Romagna 11 1.083.295 1.083.295 100 100 60.4
Friuli-Venezia Giulia* 1 322.158 322.158 100 100 56.7
Lazio 6 / 7 1.366.176 783.637 55.0 57.4 5.4
Liguria 5 421.051 421.051 100 100 15.9
Lombardia 15 2.400.066 2.400.066 100 100 45.4
Marche 5 380.090 380.090 100 100 26.2
Molise* 1 78.110 78.110 100 100 29.7
Piemonte** 9 1.134.756 428.158 39.1 37.7 23.6#
Puglia 0 / 1 980.945 393.271 0.0 40.1 1.7
Sardegna 3 / 6 434.190 329.153 51.3 75.8 22.8
Sicilia 5 / 8 1.194.196 834.151 34.1 69.9 3.4
Toscana 12 944.371 944.371 100 100 45.0
Trentino* 1 129.509 129.509 100 100 57.6
Umbria* 1 222.785 222.785 100 100 48.2
Valle d’Aosta* 1 32.358 32.358 100 100 61.6
Veneto 21 1.210.162 1.132.237 93.7 93.6 59.5#
Italy 103 / 112 14.718.125 10.635.049 64.9 72.3 25.1
North 64 / 65 6.848.148 6.063.625 87.0 88.5 41.8
Centre 24 / 25 2.913.422 2.330.883 78.6 80.0 23.7
South/Islands 15 / 22 4.956.555 2.240.541 25.2 45.2 4.4
1pairs of values refer to 2011 / 2012
2residents 50-69 yrs old at 01.01.2012 (source: Istat)
3proportion of eligible subjects residing in areas covered by a screening programme
4proportion of eligible subjects that were screened in 2011-2012
* regional-based programmes
** programmes screen only subjects aged 58-69 years
#subjects who underwent a flexible sigmoidoscopy included
Colorectal cancer screening: 2011-2012 survey
98 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Table 3. FIT programmes: extension of invitations and adjusted compliance in 2011-2012, by region.
Tabella 3. Estensione degli inviti ed adesione aggiustata dei programmi SOF nel biennio 2011-2012, per Regione.
Region Invited Extension of invitations1Screened Adjusted compliance2
subjects (N) % 10th - 90th (N) % 10th - 90th
percentile3percentile3
Abruzzo 445 0.2 174 46.2
Alto Adige 2.549 2.2 1.020 40.1
Basilicata 26.868 19.2 9.524 36.8
Calabria 18.384 3.8 8.293 47.2
Campania 154.394 11.9 21.039 13.7
Emilia-Romagna 1.084.128 116.6 89.7 - 118.9 557.021 52.1 44.4 - 61.7
Friuli-Venezia Giulia 309.016 108.3 156.208 52.3
Lazio 321.952 22.5 6.3 - 66.8 73.757 24.0 11.7 - 41.9
Liguria 227.489 55.5 29.1 - 110.1 64.327 29.1 11.2 - 41.7
Lombardia 2.189.985 97.2 83.3 - 110.8 1.027.550 48.5 41.4 - 65.9
Marche 311.050 92.9 54.2 - 116.2 87.420 28.2 23.3 - 35.5
Molise 64.468 81.3 23.221 36.6
Piemonte° 299.236 26.3 24.3 - 103.4 132.428 44.7 34.2 - 49.7
Puglia 64.605 4.6 16.305 36.5
Sardegna 207.105 40.8 98.836 50.3
Sicilia 277.331 19.2 40.312 15.6
Toscana 842.794 90.7 68.1 - 104.7 409.649 50.1 38.6 - 59.4
Trentino 112.473 97.7 66.225 59.8
Umbria 213.225 106.4 95.939 45.8
Valle d’Aosta 29.632 89.8 19.869 67.7
Veneto 987.166 91.9 78.5 - 117.2 622.820 65.5 46.1 - 76.7
Italy 7.744.295 53.1 21.5 - 111.8 3.531.937 47.1 24.0 - 67.7
North 5.241.674 82.5 59.4 - 115.2 2.647.468 52.0 36.2 - 70.0
Centre 1.689.021 58.9 19.5 - 105.0 666.765 40.6 23.2 - 56.6
South/Islands 813.600 12.2 0.4 - 89.9 217.704 28.6 12.9 - 63.1
1proportion of the target population that was actually invited in 2011-2012
2subjects attending out of those invited, excluding from the denominator those reporting a recent test and those who did not receive the invitation letter
3only Regions with at least four programmes
° programmes screen only subjects aged 59-69 years
Table 4. FIT programmes: extension of invitations and adjusted compliance in 2011-2012, by year of programme start.
Tabella 4. Estensione degli inviti e adesione corretta dei programmi SOF nel biennio 2011-2012, per anno di attivazione del programma.
Start year
<2007 2007-2009 2010+ Total
Number of programmes
Total 60 27 22 109
North 46 14 5 65
Centre 12 6624
South/Islands 2711 20
Extension of invitations (%)* 94.1 68.6 46.4 77.7
10th-90th percentile 72.5 - 112.9 15.9 - 116.1 9.7 - 99.0 24.2 - 112.9
proportion of programmes with extension >80% 85.0 42.6 27.3 63.1
Adjusted compliance (%) 50.8 44.7 27.6 47.1
10th-90th percentile 40.3 - 68.4 27.5 - 62.3 11.8 - 50.7 26.4 - 67.2
proportion of programmes with adjusted compliance >45% 79.2 37.0 20.5 57.1
* proportion of the target population of the areas with a screening programme that was actually invited in 2011-2012
Colorectal cancer screening: 2011-2012 survey
99 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Region First screening episode Repeat screening episode
Positivity rates 10th - 90th Positivity rates 10th - 90th
(%) percentile1(%) percentile1
Abruzzo 1.0
Alto Adige° 7.1
Basilicata 12.4
Calabria 5.3 6.4
Campania 6.1 4.7
Emilia-Romagna 5.5 5.0 - 6.3 4.0 3.7 - 4.3
Friuli-Venezia Giulia 5.2 3.9
Lazio 4.9 2.4 - 5.9 4.7 3.3 - 9.9
Liguria 5.3 2.6 - 18.8 4.3
Lombardia 5.5 4.5 - 6.2 4.0 3.2 - 4.8
Marche 6.3 2.5 - 8.7 3.6
Molise 4.6 4.2
Piemonte° 6.6 4.6
Puglia 4.6
Sardegna 4.5 4.9
Sicilia 5.5
Toscana 5.1 4.3 - 7.5 3.9 3.7 - 4.5
Trentino 4.7 3.9
Umbria 4.8 4.0
Valle d’Aosta 4.3 3.4
Veneto 5.1 3.4 - 6.4 3.9 3.1 - 4.8
Italy 5.2 3.7 - 6.6 4.0 3.3 - 5.1
North 5.3 4.0 - 6.6 4.0 3.3 - 4.8
Centre 5.4 3.3 - 6.6 3.9 3.5 - 5.7
South/Islands 5.0 3.1 - 6.2 4.6 4.3 - 10.6
1only Regions with at least four programmes
° not standardized (Piemonte screened only subjects aged 59-69 years, Alto Adige in 2012 screened only subjects aged 65-69 years)
Table 5. FIT programmes:
positivity rates at first and re-
peat screening episodes in
2011-2012 standardized (by
age and gender, utilising the
overall screened population as
standard population) by re-
gion.
Tabella 5. Programmi SOF:
tassi di positività ai primi
esami e agli esami successivi
nel 2011-2012 standardizzati
(per età e sesso utilizzando
come riferimento l’intera po-
polazione screenata), per re-
gione.
55-54 55-59 60-64 65-69
age
x 100 screened
1
2
3
4
5
6
7
8
9
10
0
Figure 2. FIT programmes:
positivity rates by age and
gender at first and repeat
screening episodes. Years
2011-2012.
Figura 2. Positività al SOF per
età e sesso ai primi esami e
successivi. Anni 2011-2012.
first screening - males
first screening - females
repeat screening - males
repeat screening - females
4.6
6.4
7.8
9.2
3.7
4.1
5.1
3.3
2.7
4.0
2.9
4.8
6.1
5.6
3.5
4.2
Attendance to colonoscopy assessment
Attendance to colonoscopy assessment is essential for screen-
ing programmes to achieve colorectal cancer mortality reduc-
tion. Overall, 81.1% of FIT+ subjects attended colonoscopy
in 2011-2012, a figure which is similar to those observed in
2010 (81.4%) and 2009 (82.5%). Attendance rate was higher
in the North (83.0%) and progressively decreased in the Cen-
tre (79.6%) and South and Islands (67.0%).
Only 19.5% of programmes met the desired standard (>90%),
while 7.8% was under the cut-off of 70%.
As already reported in the previous years, attendance was
higher in males (80.6%) than in females (78.9%).
Complete colonoscopies
Besides compliance to colonoscopy, a cornerstone element in
measuring the effectiveness of a screening programme is the
completeness of the endoscopic examination. Overall, 91.5%
of the colonoscopies carried out in 2011-2012 were classified
as complete, a highly satisfactory result (table 6). Eighty-one
percent of programmes met the acceptable (>85%) and 61.5%
the desired standard (>90%).
Mean regional values ranged from 80.2% in Liguria to 97%
in Trentino. The values of single programmes ranged from
53.8 and 100% and the lowest values were due to a small
number of outliers (10th percentile: 82.6%). Programmes
generally reported higher proportions of complete exams in
males compared to females (overall 91.6% vs 89.3%, respec-
tively), as reported in the literature.5
Since a proportion of subjects complete the second-level as-
sessment by repeating colonoscopy or undergoing other exams,
we also calculated the rate of completion of the diagnostic
workup. Overall in 2011-2012, second-level assessment was
completed by 95.5% of subjects with a positive first-level test.
Complications at colonoscopy
Two hundred and nine cases of bleeding were reported, 165
of which were during operative TCs, with a rate of 0.065% for
non-operative and 0.29% for operative TCs; both values are
in accordance with GISCoR standards (<0.5% and <2.5%, re-
spectively). Sixty-five perforations were recorded (52 during
operative TCs), with a rate of 0.02% for non-operative and
0.09% for operative TCs, in line with GISCoR standards
(<0.5% and <2.5%, respectively).
Overall these results are good; however, a high variability in the
collection and recording of criteria was observed. Most pro-
grammes do not provide a systematic data collection within a
fixed interval of time after the examination (e.g., 30 days), pos-
sibly resulting in an underestimation of complications, in-
cluding the most serious ones. On the other hand, the data
about bleeding might refer to self-limiting episodes that did
not require any intervention such as hospitalisation, blood
transfusion, or endoscopic interventions. In that case, the in-
dicator would be overestimated.
Colorectal cancer screening: 2011-2012 survey
100 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Region Complete colonoscopy Complete workup
rate1(%) rate2(%)
Basilicata 80.6 81.0
Campania 96.5 96.7
Emilia-Romagna 91.5 96.5
Friuli-Venezia Giulia 93.7 97.0
Lazio 88.7 98.4
Liguria 80.2 85.4
Lombardia 91.7 95.7
Marche 92.1 95.5
Piemonte 91.1 91.1
Sardegna 96.6 99.7
Sicilia 83.2 87.5
Toscana 87.9 94.5
Trentio 97.0 98.9
Umbria 88.0 91.9
Valle d’Aosta 95.0 95.0
Veneto 93.0 97.5
Italy 2011-2012 91.5 95.5
1proportionof first total colonoscopies following a positive screening test thatreached cae-
cal intubation
2proportion of subjects who underwent a second-level workup who had a complete as-
sessment (a complete total colonoscopy and/or other exams)
Table 6. Complete colonoscopy rate and complete workup rate in 2011-2012, by
region.
Tabella 6. Tasso di colonscopie complete e tasso di approfondimenti completi,per
Regione. Periodo 2011-2012.
first screening repeat screening
x 1,000 screened
2
4
6
8
10
12
0
Figure 3. FIT programmes: detec-
tion rates of carcinoma, advanced
adenoma, and non advanced ade-
noma at first and repeat screening
episodes in 2011-2012.
Figura 3. Programmi SOF: tassi di
identificazione di carcinoma, ade-
noma avanzato e adenoma iniziale
ai primi esami e successivi. Anni
2011-2012.
2.2
10.3
7.1
1.0
6.8 6.1
carcinoma adv. adenoma non adv. adenoma
Detection rates
We describe the detection rates (DR) of invasive carcinomas,
advanced adenomas (i.e., adenomas with a diameter ≥1 cm,
villous/tubulo-villous type, or high-grade dysplasia) and
non-advanced adenomas (smaller in size, tubular type, and
low-grade dysplasia). DRs are defined as the number of his-
tologically-confirmed lesions detected per 1,000 screened
subjects.
Overall, in subjects screened for the first time, 2,916 carci-
nomas, 13,578 advanced adenomas, and 9,320 non-advanced
adenomas were detected.Therefore, the DR was 2.2‰ for car-
cinoma, 10.3‰ for advanced adenomas and 7.1‰ for non-
advanced adenomas (figure 3). Sixty-three percent of pro-
grammes reached the acceptable standard for carcinoma
(>2‰), and 75% for advanced adenoma (>7.5‰).
In subjects undergoing repeat testing, 2,306 carcinomas,
15,001 advanced adenomas, and 13,427 non-advanced ade-
nomas were detected. As expected, the DRs were lower than
the corresponding figure at first exams (figure 3). Sixty-
nine percent of programmes reached the acceptable standard
for carcinoma (>1‰), and 63% for advanced adenoma
(>5‰).
The ratio between the DRs of advanced and non-advanced
adenomas does not reflect the underlying prevalence of the two
groups of lesions in the screened population, the frequency of
non-advanced adenomas being higher than that of advanced
adenomas. The DR of advanced adenomas is higher, since FIT
appears to be highly selective for these lesions, which tend to
bleed more easily than non-advanced adenomas, as described
in the literature.6However, we observed a high variability
among programmes in the ratio between advanced and non-
advanced adenomas.This result suggests a low standardisation
of the diagnostic criteria used by the different programmes to
classify adenomas.
At first exams, we observed a high variability among the re-
gional mean values of DRs of carcinoma (from 1.7‰ in Cal-
abria to 7.8‰ in Bolzano, both non-standardized values), ad-
vanced adenomas (from 1.9‰ in Puglia to 13.7‰ in Marche
and Emilia-Romagna; in Piemonte, with its 19.4‰, pro-
grammes screened only subjects aged 58-69 years) and non-
advanced adenomas (from 3.3‰ in Puglia to 14.7‰ in
Friuli-Venezia Giulia and Bolzano) (figure 4).
We did not observe any geographical North-South trend in the
detection rates of carcinoma and advanced adenoma, as ex-
pected according to the underlying epidemiological figures
(carcinoma: North 2.3‰, Centre 2.2‰, South-Islands
2.2‰; advanced adenoma: North 11.2‰, Centre 10.6‰,
South/Islands 7.1‰; non-advanced adenoma: North 7.6‰,
Centre 7.5‰, South/Islands 4.8‰). At repeat examinations,
a higher homogeneity was reported among regions for the DR
of carcinoma (from 0.6‰ in Marche to 2.3‰ in Calabria),
but not for advanced adenoma (from 2.3‰ in Calabria to
10.3‰ in Sardegna) nor non-advanced adenoma (from 3.1‰
in Valle d’Aosta to 11.6‰ in Trentino) (figure 5, p. 102).
Colorectal cancer screening: 2011-2012 survey
101 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Figure 4. FIT programmes: standardized (by age and gender, utilising the overall screened population as standard population) detection rates for carcinoma, advanced
adenoma and non-advanced adenoma at first screening, by region.Years 2011-2012.
Figura 4. Programmi SOF: tassi di identificazione di carcinoma, adenoma avanzato e adenoma iniziale ai primi esami, standardizzati (per età e sesso, utilizzando come
riferimento l’intera popolazione screenata), per regione.Anni 2011-2012.
x 1,000 screened
5
10
15
20
25
30
0
carcinoma adv. adenoma non adv. adenoma
* not standardized (Calabria did not provide data by age class; AltoAdige, Marche, and Liguria screened only some age classes; Piemonte screened only subjects aged 59-69 years)
10.3
7.6
2.3
11.2
2.2
6.7
9.7
Puglia
Calabria*
Sicilia
Sardegna
Lazio
Umbria
Toscana
Valle d’Aosta
Liguria*
Campania
Lombardia
Emilia-Romagna
Veneto
Friuli-Venezia Giulia
Trentino
Piemonte*
Marche*
Alto Adige*
Italy 2010
Italy 2011
Italy 2012
8.8
2.4
As expected, on the basis of underlying epidemiological fig-
ures, the DRs of carcinoma were higher in males and pro-
gressively increased with age in both genders (figure 6). This
trend may be observed both in subjects screened for the first
time and in those at repeat screening.
The reduction in DRs between first and repeat exams was
larger in males and in the older groups: this could be due to
a proportionally higher impact in these subjects of the
polyps’ removal that takes place in the prevalence round, and
it is in agreement with the data about positivity rates of FIT
(figure 2).
Positive predictive value
Positive predictive value (PPV) of FIT+ at colonoscopy is de-
fined as the number of subjects with a diagnosis of carcinoma
or advanced adenoma, as a proportion of FIT+ subjects that
underwent colonoscopy.
In 2011-2012, the FIT showed a noteworthy capability of se-
lecting subjects with a high risk of invasive carcinoma or ad-
vanced adenoma, as already reported in the previous years.
Among the 55,419 subjects at first screening round who un-
derwent a colonoscopy after a FIT+, a diagnosis of carcinoma was
formulated in 5.3% and advanced adenoma in a further 24.5%.
Colorectal cancer screening: 2011-2012 survey
102 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Figure 5. FIT programmes: standardized (by age and gender, utilising the overall screened population as standard population) detection rates for carcinoma, advanced
adenoma and non-advanced adenoma at repeat screening episodes,by region. Years 2011-2012.
Figura 5. ProgrammiSOF: tassi di identificazione di carcinoma, adenoma avanzato e adenoma iniziale agli esami successivi, standardizzati (per età e sesso,utilizzando come
riferimento l’intera popolazione screenata), per regione.Anni 2011-2012.
x 1,000 screened
5
10
15
20
25
0
carcinoma adv. adenoma non adv. adenoma
* not standardized (Calabria did not provide data by age class; Marche and Liguria screened only some age classes; Piemonte screened only subjects aged 59-69 years)
7.6
5.6
1.1
6.7
0.9
6.3
7.0
Liguria*
Valle d’Aosta
Umbria
Toscana
Lazio
Campania
Emilia-Romagna
Lombardia
Veneto
Calabria*
Marche*
Friuli-Venezia Giulia
Piemonte*
Sardegna
Trentino
Italy 2010
Italy 2011
Italy 2012
6.1
1.2
55-54 55-59 60-64 65-69
age
x 1,000 screened
1
2
3
4
5
6
0
Figure 6. FIT programmes: detection
rates of carcinoma by age and sex at
first and repeat screening.Years 2011-
2012.
Figura 6. Programmi SOF: tassi di
identificazione di carcinoma per età e
sesso e classe d’età ai primi esami e
agli esami successivi.Anni 2011-2012.
1.3
2.6
4.3
5.9
0.9
0.5
0.5
1.7
0.6
2.1
1.5
2.8
2.0
0.9 1.1
0,8
first screening - males
first screening - females
repeat screening - males
repeat screening - females
Among the 74,810 subjects at repeat screening, the correspon-
ding values were respectively 3.1% for carcinoma and 20.1% for
advanced adenoma.
Seventy-five percent of programmes reached the acceptable
standard for subjects at first screening (>25%) and 85% for
those at repeat screening (>15%).
Once again, males showed constantly higher values than fe-
males (31.0% vs 22.6% for carcinoma and advanced ade-
noma altogether) and an increasing PPV trend was observed
with age (from 24.2% in subjects 50-54 years old to 29.4% in
those aged 65-69).
Waiting times
In order to reduce the anxiety of screened subjects, the delay
between the test and mailing of a negative result or the carry-
ing out of a further assessment for those positive must be kept
as short as possible. Since FIT is a laboratory test, it can be car-
ried out quite quickly (as compared to the reading of mam-
mograms and Pap smears), therefore the delay between the test
and the mailing of a negative result is generally short. In fact,
about 94% of letters after a negative result were mailed within
15 days and a further 3% within 21 days.
On the contrary, all regions recorded serious difficulties in guar-
anteeing a colonoscopy to FIT+ subjects within a short period
of time. Overall, colonoscopy was carried out within 30 days
after FIT only in 53.3% of cases and only nine programmes
met the acceptable standard (>90% within 30 days). Fifteen
percent of subjects had to wait for more than two months.
Finally, surgery was performed within 30 days after diagnosis
in 52% of cases, and in a further 33% within two months.
FS SCREENING PROGRAMMES
FS is proposed as a first level test by 9 programmes in Piemonte
and 1 in Veneto (in 2012 two programmes were suspended).
These programmes also offer FIT to subjects refusing FS
screening and to those up to 69 years of age.The principal data
are presented in table 7.
Overall, these programmes invited 53,668 subjects in 2011,
corresponding to an 88.2% actual coverage of their target
population (N= 61,973) and 47,499 subjects in 2012 (84% of
55,871 subjects in the target population).
Overall, 12,825 subjects were screened in 2011 and 11,724 in
2012. Uptake of invitation was 24.5% (range: 6.9-36.8%). In
almost all programmes, uptake was higher for males in com-
parison to females (overall: 25.8% vs 23.2%), as reported in the
literature. Compliance to FS screening was lower than for
FIT. However, the comparison is related to different geo-
graphical areas.
The programmes offer FIT to subjects refusing FS screening.
This strategy makes it possible to increase overall coverage and
reduce gender differences, as reported where this strategy has
been ongoing for a number of years. In fact, the proportion of
subjects that underwent one of the two tests was 36.4% and was
higher among females (37.5%) than males (35.3%) (table 7).
Since FS is performed on a once-in-a-lifetime basis, the pro-
portion of complete exams should be as high as possible. On
the other hand, caution must be taken to avoid perforations,
bleeding, or other complications. Overall, 85.9% of FS were
classified as complete, with higher levels in males (88.6%) than
in females (82.5%). This result is in line with GISCoR’s ac-
ceptable standard (>85%). A considerably high variability be-
tween programmes was recorded (range: 74.4-94%).
Generally, the programmes referred to colonoscopy assess-
ment 9.8% of screened subjects (12.0% of males and 7.4% of
females). Only in 57% of these cases was the reason prompt-
ing colonoscopy an advanced adenoma, which, according to
the literature, is associated with an increased probability of neo-
plasia in the proximal colon.
The overall attendance rate of the assessment was 93.5% in
2011 and dropped to 81.9% in 2012, probably due to a loss
of data. The colonoscopy completeness rate was 91.9%,
with values of single programmes ranging from 86.2% to
100%.
Among the subjects referred to colonoscopy, the prevalence of
Colorectal cancer screening: 2011-2012 survey
103 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Males Females Total
screened 2012 (N) 5,983 5,741 11,724
screened 2011 (N) 6,646 6,179 12,825
compliance with invitation to FS (%) 25.8 23.2 24.5
compliance with invitation to FS+FIT (%) 35.3 37.5 36.4
reason prompting colonoscopy (%)
advanced adenoma* 7.4 3.6 5.6
other 4.6 3.8 4.2
detection rate (‰)
carcinoma 4.4 1.5 3.0
advanced adenoma 63.9 31.1 48.2
non advanced adenoma 96.8 61.6 79.9
PPV (%) for proximal neoplasia** 12.0 9.0 11.0
* at least one advanced adenoma (with a diameter ≥10 mm, villous/tubulo-villous type,or high-grade dysplasia); 3 or more adenomas with diameter <10 mm, tubular
type, and low grade dysplasia
** carcinoma or advanced adenoma
Table 7. Main results of FS
programmes. Years 2011-
2012.
Tabella 7. Risultati principali
dei programmi RS.Anni 2011-
2012.
proximal advanced lesions (advanced adenomas plus cancers)
ranged between 2.7% and 14.9%.
Overall, FS programmes detected 71 carcinomas, of which 67
in the distal tract of the colon, and 1,129 advanced adenomas,
with a DR of 3.0 and 48.2‰, respectively.
Stage at diagnosis
Overall, 2,916 cancers were detected in subjects at first screen-
ing and 2,306 at repeat screening. Invasive malignant polyps
represented 27.6% of cancers at first screening and 22.3% at
repeat screening. FS programmes detected 71 cancers, 14 of
which were invasive malignant polyps.
As already observed in the previous years, many programmes
did not collect any data about stage at diagnosis, while infor-
mation provided by others was incomplete. Therefore, stage
is available only for 3,733 cases (71.5%) of the 5,222 carci-
nomas, similar to 2010 and 2009 (73.5% and 71.7% respec-
tively).The incompleteness of this information was one of the
most critical issues encountered by Italian programmes dur-
ing 2010.
Table 8 shows the distribution by stage at diagnosis of cases
screen-detected by FIT and FS programmes. The distribution
of cases diagnosed at first vs repeat FIT are similar, with more
than half of cases at stage I and a considerable proportion of
cases treated only by endoscopic resection.
Overall, 27.3% of cases were in stage III+ at diagnosis, in ac-
cordance with the acceptable standard (<30%). As for the pro-
portion of cases in stage III-IV, small differences were re-
ported between cases at first and repeat screening.
Sixty percent of cases diagnosed by FS programmes were at
stage I; of these, 22.6% were invasive (pT1) malignant polyps
that underwent endoscopic resection alone.
Surgery
This survey collects data about the kind of therapy performed
on carcinomas, invasive malignant polyps and advanced ade-
nomas, and distinguishes between surgical intervention and
endoscopic resection alone. Overall, data were provided for
81.4% of carcinomas and 91.8% of advanced adenomas.
Eighty-five percent of carcinomas underwent surgery, while in
15% of cases treatment was limited to endoscopic resection.
This percentage increased to 40.8% considering only pT1
cases. As for advanced adenomas, treatment was exclusively en-
doscopic in 96.7% of cases.
Post-colonoscopy follow- up
The national survey collected information about recommen-
dations given at the end of the diagnostic workup by type of
diagnosis, and distribution of the colonoscopies carried out by
the screening programmes, by type: second-level assessments,
repetition, follow-up, etc.
■
Recommendations after a clean colon
Most subjects with a negative colonoscopy were invited to per-
form a FIT after 5 years (79.6%), in line with the European
guidelines7(table 9). Thirteen percent of the cases were rec-
ommended to undergo a further colonoscopy, at different in-
tervals, without any relevant difference between geographical
areas. The European guidelines recommend to return subjects
to screening even in case of a diagnosis of non-advanced ade-
noma. This recommendation was respected only by 10.7%
cases, while the indication in the vast majority of cases was a
further colonoscopy, at longer intervals in the North (53% af-
ter 5 years and 20% after 3) compared to the Centre (37% af-
ter 5 years, 34% after 3) and the South of Italy and Islands
(23% and 25%, respectively).
Advanced adenomas should be recalled to colonoscopy after
1 or 3 years (depending on the number and dimension of the
adenomas). This recommendation was given in 73% of cases,
Colorectal cancer screening: 2011-2012 survey
104 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Recommendation Negative Low-risk High-risk Cancerized
(%) adenoma (%) adenoma* (%) adenoma (%)
FIT after 5 years 79.6 8.1 1.4 1.9
FIT after 2 years 4.4 2.6 0.6 0.0
colonoscopy after 5 years 7.3 50.8 5.8 0.6
colonoscopy after 3 years 3.3 22.2 48.4 4.0
colonoscopy after 6 months/1 year 2.4 7.9 33.9 16.9
surgery 0.6 1.9 5.7 68.9
other 2.4 6.5 4.2 7.7
* high-risk adenoma: at least one advanced adenoma (with a diameter ≥10 mm, villous/tubulo-villous type, or high-grade dysplasia); 3 or more adenomas with diameter
<10 mm, tubular type, and low-grade dysplasia
Table 9. Distribution of rec-
ommendations after
clean
colon
, by diagnosis at colono-
scopy in 2011-2012.
Tabella 9. Distribuzione per-
centuale delle raccomanda-
zioni dopo
clean colon
per
diagnosi istologica nel bien-
nio 2011-2012.
Stage FIT programmes FS
first screening repeat screening programmes
(N=1,910) (%) (N=1,823) (%) (N=62) (%)
I 41.9 42.9 37.1
I* 10.8 9.8 22.6
II 20.4 19.5 12.9
III-IV 26.8 27.9 27.4
Stage I: T1 or T2. N0. M0
Stage I*: T1. NX
Stage II: T3 or T4. N0. M0
Stage III-IV: lymph-node involvement or distant metastases
Table 8. Stage distribution of screen-detected cancers in 2011-2012.Cases with
known stage (3,733 out of 5,222 carcinomas).
Tabella 8. Distribuzione per stadio alla diagnosi dei carcinomi diagnosticati allo
screening nei programmi SOF e RS nel biennio 2011-2012 (%).Casi con stadio noto
(3.733 su 5.222 carcinomi totali).
while in 9.6% of cases colonoscopy was anticipated after 6
months and 2% of cases were recalled to FIT.
Sixty-nine percent of the cases of invasive malignant polyps
were sent to surgery, a further 8.3% to repeat colonoscopy af-
ter 6 months.
■
Distribution by reason prompting colonoscopy
Seventy-four percent of the colonoscopies performed in 2011-
2012 were second-level assessments in subjects with a positive
screening test (table 10), 20.3% were post-colonoscopy follow-
up and 4.8% completion or repetitions of a previous
colonoscopy.
The proportion of follow-up colonoscopies was very low
(1.6%) in the programmes without an active invitation to
follow-up, while it rose to 26.5% in those with an active fol-
low-up. Among the latter, the proportion of follow-up colono-
scopies was highest in programmes older than 6 years (27.9%).
DISCUSSION
During 2011 and 2012, colorectal cancer screening pro-
grammes continued to spread gradually, and by the end of
the period they covered 74% of the national target
population.
About 7.7 million subjects were invited to screening, half of
whom underwent a screening test; 5,222 carcinomas and
28,579 advanced adenomas were diagnosed, making the Ital-
ian experience one of the most advanced in the world.
Fifteen new programmes were started, 12 of which were in the
South of Italy and Islands, which maintained a delay in com-
parison with the North and Centre, in part because a number
of programmes was suspended.
Overall, 78% of the annual target population residing in areas
with a programme were invited.
The extension of invitations of the programmes that had been
activated before 2007 was optimal, while the more recent pro-
grammes showed much lower performances (on average, 46%).
It seems that the new programmes are meeting more problems
in reaching adequate numbers of invitations. We recommend
a careful monitoring of this indicator to all programmes.
Compliance with invitation is in line with the previous years.
However, the very low values that affect many programmes,
particularly when associated with a limited extension of invi-
tations, are of particular concern, as in some cases the com-
bined effect of these two elements makes the proportion of the
target population that has been effectively screened marginal.
Intra-regional attendance showed high levels of variability,
which suggests the possibility of increasing the performance of
many programmes.
Overall, 82% of the subjects that had attended a screening
episode responded to the subsequent invitation. No differences
according to age or gender were observed, suggesting that the
experience of the previous screening episode becomes the main
driver for subsequent attendance, as already described in the lit-
erature.8Thus, the effect of other factors, which influence re-
sponse to the first invitation, decreases. It is therefore impor-
tant for programmes to identify the limitations that may have
determined a lack of satisfaction in the screened population, es-
pecially if the attendance rate is low, because attendance in sub-
sequent rounds is necessary to obtain the expected protection.
Attendance among subjects that had already been invited but
never attended was 18%. This reflects the possibility to enrol
subjects at higher risk (because they have never been screened)
and the importance of continuing to regularly invite this group
of people that might seem reluctant to participate in screening.
These data suggest that the screened population changes over
the years. This means that:
■
the test coverage of the target population is higher than the
number of screened subjects;
■
for subjects who do not regularly undergo screening, the pro-
tective effect of screening will be lower than expected.
This aspect should be taken into consideration when com-
paring the impact of FIT vs FS programmes, because the lat-
ter provides a protection that lasts for at least 12 years to all
screenees. On the other hand, the protection afforded by FIT
will be extended to a greater number of subjects than those an-
nually recorded in the survey.
The available data are not enough to estimate the length of the
protection of FIT and hence the interval between two tests that
still confers a consistent risk reduction.
The evaluation of diagnostic indicators is difficult because
many programmes produced incomplete data and this may be
misleading when interpreting the results on a regional basis:
some indicators depend on many factors (e.g., DRs are influ-
enced by the distribution of the screenee by age and sex, by FIT
positivity, and by compliance to colonoscopy) and they should
be interpreted according to their intra-regional composition.
For each indicator we had to select the programmes that sent
complete data, with a possible selection bias. Unfortunately, the
less complete questionnaires came from the regions with the
lower number of programmes, leading to an even greater bias.
FIT showed to be an excellent first-level test for colorectal
screening in terms of homogeneity of positivity rates both at
first and subsequent episodes, with high PPVs and short de-
lay between the test and the mailing of a negative result. Other
evidence is still sparse, such as evaluation of the sensitivity of
FIT-based programmes through interval cancers. GISCoR
produced an Operative report on the collection of interval can-
cers and the estimate of sensitivity, for the purpose of making
Colorectal cancer screening: 2011-2012 survey
105 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Type of colonoscopy Total Programmes with active
follow-up (%)
start date start date
2000-2006 2007-2010
second level assessment after
a positive first-level exam 74.5 66.7 74.3
repetition, etc 4.8 5.0 5.6
follow-up (after
clean colon
) 20.3 27.9 19.3
other 0.4 0.4 0.8
Table 10. Distribution of colonoscopies performed in 2011-2012, by type (%).
Tabella 10. Distribuzione percentuale delle colonscopie per motivo di esecuzio-
ne nel biennio 2011-2012.
the monitoring of this fundamental aspect of screening pro-
grammes easier and more homogenous.
Particular attention should be given to attendance to
colonoscopy (81.1%). This is a critical point of FIT pro-
grammes which has been observed in the last 5 years without
any sign of improvement.The actual proportion of FIT+ sub-
jects that did not undergo any further assessment was proba-
bly lower, since many programmes did not collect data about
assessments performed in non-screening settings. According to
a multicentric Italian study, about 3% of FIT+ subjects un-
derwent TC outside the screening programme.9
However, it must be stressed that the duty of screening pro-
grammes is not only that of reaching high levels of attendance
to colonoscopy, but also making sure that FIT+ subjects have
undergone assessment, even if outside the programme. The
data reported suggest that many programmes did not concern
themselves with this aspect.
A further issue that needs to be analyzed locally is the rela-
tionship between attendance to colonoscopy and the use of se-
dation and waiting time for assessment. During 2011-2012, we
observed a generalized difficulty for endoscopic services in
dealing with the workload deriving from screening positives,
as the burden of colonoscopies for the follow-up of adenomas
progressively increases.
Italian data are similar to those reported in the literature.10-
12 Some Italian experiences, which recorded attendance rates
higher than 90%, underlined the relationship between a high
compliance to colonoscopy and the diagnostic yield of screen-
ing programmes.13,14 A multicentric study recently showed
that different modalities of invitation may be used to increase
compliance with colonoscopy.9
The analysis of PPV of FIT+ at colonoscopy confirms the high
values reported in the previous years. According to these find-
ings, it is essential that screening programmes adopt strategies
in order to maximise colonoscopy attendance, or to ensure that
subjects with a positive FIT undergo further diagnostic as-
sessment in non-screening structures.
Compared to the last years, the overall DRs of carcinoma and
advanced adenoma were stable, even though many pro-
grammes showed a lower DRs at first screening. This is not
worrisome, since for programmes at subsequent rounds, a
high proportion of the population that undergoes the screen-
ing test for the first time is represented by fifty-year-old sub-
jects, which are at lower risk of disease.
Since DRs are calculated dividing the diagnosed lesions by the
screened population, they are inversely associated with loss of
attendance to colonoscopy. In fact, when adjusting the DRs by
attendance to colonoscopy, we observed a levelling off of the
differences between regional means (data not shown).
The fluctuations of DRs between programmes and regions sug-
gest, beyond different underlying prevalence rates, the presence
of other factors responsible for this aspect influencing the di-
agnostic sensitivity of the screening programme, such as the
quality of endoscopy and the different criteria locally used to
classify adenomas as advanced or non-advanced. The high
variability among programmes of the ratio between advanced
and non-advanced adenomas seems to confirm the importance
of the latter factor.
Adenoma detection rate is one of the most important indica-
tors to monitor the quality of colonoscopy.7The data obtained
from programmes show a good quality of colonoscopies in
terms of completeness (91% of caecal intubation rates) and
complication rates, both for surgical and non-surgical TCs.
The National centre for screening monitoring, together with
GISCoR and with the major Italian scientific societies of en-
doscopy, carried out an assessment, the Equipe study, in order
to evaluate the performance of colonoscopies at the level of in-
dividual endoscopists and endoscopy services. The results of the
study are in line with those produced by the national survey.
In particular, the analysis of 75,569 total colonoscopies carried
out in 44 screening programmes showed that policies ad-
dressing organizational issues, such as sedation and the avail-
ability of screening sessions, may improve adenoma detection
rate and overall quality of colonoscopy.15
As for treatment, we collected information about the use of sur-
gical intervention versus endoscopic resection alone. Overall,
15% of carcinomas underwent endoscopic resection alone, re-
sulting in improved patient quality of life and cost reduction.
This percentage increased only to 41% in pT1 cases, which
mostly involve invasive malignant polyps. A possible overtreat-
ment of these subjects should be accounted for. Overall, 97%
of advanced adenomas were treated through endoscopic re-
section alone.
An important step that requires evaluation is post-colonoscopy
follow-up, which represents a relevant share of the total en-
doscopic workload of programmes that actively invite subjects
to follow-up. Application of the European guidelines proto-
cols would reduce the burden of these exams substantially, be-
cause the observed recommendations mainly result in an
over-prescription of endoscopic follow-up. We encourage
screening programmes to locally evaluate the indicators that
are reported in this survey, in order to verify compliance with
the European guidelines, both of endoscopists and endoscopy
services, especially if the waiting time for colonoscopy is par-
ticularly long.
This survey could not evaluate the outcomes of follow-up: this
would require an individual collection of information about the
timing and diagnosis of the index colonoscopy. We recommend
that programmes and regions that have adequate historical
databases carry out these analyses, which are expected to con-
firm the evidence underlying the recommendations of the Eu-
ropean guidelines and would be useful to support the spread
of their application.
Finally, the results of this survey may be used by new pro-
grammes to estimate the burden of colonoscopic workload they
may expect as time goes by.
Conflicts of interests: none declared
Colorectal cancer screening: 2011-2012 survey
106 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Colorectal cancer screening: 2011-2012 survey
107 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
(Reggio Emilia); R. Sassatelli (AOSP Reggio Emilia); R. Corradini,
C. Goldoni (Modena); A. Pasquini, M. Manfredi, P. Baldazzi (Bologna);
R. Nannini, L. Caprara (Imola); M.C. Carpanelli, G. Zoli, (Ferrara);
V. Matarese (AOSP Ferrara); O. Triossi, M. Serafini, B. Vitali (Ravenna);
F. Falcini, A. Colamartini, O. Giuliani, R. Vattiato (Forlì); M. Palazzi,
C. Imolesi, P. Pazzi (Cesena); D. Canuti, C. Casale, M. Giovanardi,
G. Monticelli (Rimini)
Toscana: C. Nicolai, P. Vivani (Massa e Carrara); D. Giorgi, G. Finucci
(Lucca); M. Rapanà (Pistoia); C. Epifani, L. Abdelghani (Prato);
G. Allegrini (Pisa); C. Maffei (Livorno); R. Turillazzi (Siena); F. Mirri,
P. Ceccatelli (Arezzo); R. Rosati, P. Piacentini (Grosseto); C.B. Visioli,
P. Falini (Firenze); P. Amico (Empoli); C. Ciabattoni (Viareggio)
Umbria: M. Giaimo, S. Prandini (Regione Umbria); G. Vinti (Città
di Castello); A. Di Marco (Foligno); M. Malaspina (Perugia); R. Corvetti
(Terni)
Marche: L. Di Furia
Lazio: A. Barca, D. Baiocchi, F. Quadrino
Abruzzo: M. Di Giacomo (Regione Abruzzo); F.M. Lattanzio (ASL
Lanciano-Vasto-Chieti); M. Minna (ASL Pescara)
Molise: A. Di Credico
Campania: R. Pizzuti
Basilicata: A. Sigillito
Calabria: M.P. Montesi (Lamezia Terme); T. Landro (Vibo Valentia);
A. Giorno (Cosenza)
Sicilia: M. Santino (Caltanissetta); G. Magrì (Catania); G. Ferrara
(Ragusa)
Sardegna: R. Masala
Valle D’Aosta: S. Crotta
Piemonte: C. Senore (Torino), S. Polizzi (Moncalieri), M. Sartori
(Rivoli-Collegno), M.P. Alibrandi (Ivrea), F. Germinetti (Biella-Vercelli),
P. Bestagini (Novara), L. Orione (Cuneo), T. Miroglio (Asti), G. Faragli
(Alessandria)
Lombardia: D. Cereda, L. Coppola, L. Zerbi, M. Gramegna (Regione);
L. Tessandri, P. Imbrogno, G. Rocca, B. Pesenti (ASL Bergamo);
M. Schivardi, M. Crisetig, E. Grassi, F. Speziani (ASL Brescia); G. Gola
(ASL Como); S. Gotti, M. Dal Soldà, L. Boldori (ASL Cremona);
G. Moretti, A. Ilardo (ASL Lecco); A. Belloni, E. Rossetti, G. Marazza
(ASL Lodi); E. Anghinoni, (ASL Mantova); A. Silvestri, E. Tidone,
B. Frammartino, N. Leonardo, S. Deandrea (ASL Milano); P. Ceresa,
G. Beghi (ASL Milano 1); R. Lucchini, L. Acerbi (ASL Milano 2);
F. Lo Buono, L. Cavalieri d'Oro (ASL MB); G. Magenes, L. Camana
(ASL Pavia); A.M. Cioccarelli, A.C. Fanetti, L. Cecconami (ASL
Sondrio); R. Bardelli, M. Violini, F. Sambo (ASL Varese);
S. Domenighini, G. Pieracci (ASL Valle Camonica Sebino)
Trentino: R. Pertile, S.Piffer
Veneto: C. Fedato
Friuli-Venezia Giulia: A. Franzo, J. Fabro, M. Gobbato, L. Zanier
Liguria:L. Bonelli (IRCCS AOU San Martino-IST); M. Orlando,
D. Vaccari (ASL 1 Imperiese); A. Franxo, M. Scotto (ASL 2 Savonese);
I. Valle (ASL 3 Genovese); M. Ferrari Bravo, C. Sticchi (ASL 4
Chiavarese); F. Maddalo, F. Pensa (ASL 5 Spezzino)
Emilia-Romagna: C. Naldoni, P. Sassoli de’ Bianchi, P. Landi (Regione
Emilia-Romagna); E. Borciani, F. Fornari, G. Gatti (Piacenza); C. Zurlini,
M. Zatelli (Parma); F. Maradini (AOSP Parma); L. Paterlini, C. Campari
Data for the 2011-2012 survey was provided by:
Hanno fornito i dati per la survey 2011-2012:
References/Bibliografia
1. Zorzi M, Fedato C, Grazzini G et al. Screening for colorectal cancer
in Italy, 2010 survey. Epidemiol Prev 2012 Nov;36(6 Suppl 1):55-77.
2. Zorzi M, Sassoli de’ Bianchi P, Grazzini G, Senore C e Gruppo di la-
voro sugli indicatori del GISCoR. Quality indicators for the evalua-
tion of colorectal cancer screening programmes. Epidemiol Prev
2007;6(Suppl 1):6-56.
[
www.osservatorionazionalescreening.it/ ons/
npubblicazioni/ altre_ons.htm].
3. http://demo.istat.it/pop2012/index.html. Accessed 19 march 2014.
4. AIRTUMWorking group.Italian cancer figures. Report 2006: incidence,
mortality and estimates. Epidemiol Prev 2006;1(Suppl 2):38-41.
5. Segnan N, Senore C, Andreoni B et al. SCORE2 Working group-Italy.
Randomized trial of different screening strategies for colorectal
cancer: patient response and detection rates. J Natl Cancer Inst
2005;97(5):347-57.
6. Ciatto S, Martinelli F, Castiglione G et al. Association of FOBT-
assessed faecal Hb content with colonic lesions detected in the
Florence screening programme. Br J Cancer 2007;96(2):218-21.
7. Segnan N, Patnick J, von Karsa L (eds.). European guidelines for
quality assurance in colorectal cancer screening and diagnosis.
First edition. European commission, Publications office of the Eu-
ropean Union, Luxembourg 2010.
8. Vernon SW. Participation in colorectal cancer screening: a review.
J Natl Cancer Inst 1997;89:1406-22.
9. Zorzi M, Giorgi Rossi P, Cogo C et al; PARC Working Group. A
comparison of different strategies used to invite subjects with a
positive faecal occult blood test to a colonoscopy assessment. A
randomised controlled trial in population-based screening pro-
grammes. Prev Med 2014;65:70-76.
10. Moss SM, Campbell C, Melia J et al. Performance measures in
three rounds of the English bowel cancer screening pilot. Gut
2011 doi:10.1136/gut.2010.236430
11. Van Roon AHC, Hol L, van Vuuren AJ et al. Are Fecal Immuno-
chemical Test Characteristics influenced by Sample Return Time?
A Population-Based Colorectal Cancer Screening Trial. Am J Gas-
troenterol 2012;107:99-107.
12. Peris M, Espinàs JA, Muñoz L et al. Lessons learnt from a popula-
tion-based pilot programme for colorectal cancer screening in Cat-
alonia (Spain). J Med Screen 2007;14(2):81-86.
13. Crotta S, Segnan N, Paganin S et al. High rate of advanced ade-
noma detection in 4 rounds of colorectal cancer screening with
the fecal immunochemical test. Clin Gastroenterol Hepatol 2012;
10(6):633-38.
14. Parente F, Marino B, DeVecchi N, Moretti R; Lecco Colorectal Can-
cer Screening Group. Faecal occult blood test-based screening
programme with high compliance for colonoscopy has a strong
clinical impact on colorectal cancer. Br J Surg 2009;96(5):533-40.
15. Zorzi M, Senore C, Da Re F et al; the Equipe Working Group. Qual-
ity of colonoscopy in an organised colorectal cancer screening pro-
gramme with immunochemical faecal occult blood test: the
EQuIPE study (Evaluating Quality Indicators of the Performance of
Endoscopy). Gut 2014 Sep 16. pii: gutjnl-2014-307954. doi:
10.1136/gutjnl-2014-307954.
16. Segnan N, Armaroli P, Bonelli L et al. Once-only sigmoidoscopy in
colorectal cancer screening: follow-up findings of the Italian ran-
domized controlled trial – SCORE. J Natl Cancer Inst 2011;103:
1310-22.
17. Atkin WS, Edwards R, Kralj-Hans I et al. Once-only flexible sig-
moidoscopy screening in prevention of colorectal cancer: a multi-
centre randomised controlled trial. Lancet 2010;375:1624-33.
18. Schoen RE, Pinsky PF, Weissfeld JL et al; the PLCO Project Team.
Colorectal-Cancer Incidence and Mortality with Screening Flexi-
ble Sigmoidoscopy. N Engl J Med 2012;366:2345-57.
108 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Characteristics of the colorectal cancers
diagnosed in the early 2000s in Italy.
Figures from the IMPATTO study on colorectal
cancer screening
Caratteristiche dei tumori del colon retto
diagnosticati in Italia nei primi anni Duemila.
Dati dello studio IMPATTO sdelo screening colorettale
Manuel Zorzi,1Lucia Mangone,2,3 Emanuela Anghinoni,4,5 Susanna Baracco,1Elisabetta Borciani,6
Adele Caldarella,7Fabio Falcini,8Anna Clara Fanetti,9Stefano Ferretti,10 Paolo Giorgi Rossi,11,12
Maria Michiara,13 Giorgia Randi,14 Fabrizio Stracci,15 Massimo Vicentini,11 Antonella Zucchetto,16
Marco Zappa7,17 and IMPATTO COLONRETTO working group
Abstract
The impact of organized screening programmes on colorectal cancer (CRC) can be observed at a pop-
ulation level only several years after the implementation of screening. We compared CRC character-
istics by diagnostic modality (screen-detected, non-screen-detected) as an early outcome to monitor
screening programme effectiveness.
Data on CRCs diagnosed in Italy from 2000 to 2008 were collected by several cancer registries. Link-
age with screening datasets made it possible to divide the cases by geographic area, implementation
of screening, and modality of diagnosis (screen-detected, non-screen-detected). We compared the main
characteristics of the different subgroups of CRCs through multivariate logistic regression models.
The study included 23,668 CRCs diagnosed in subjects aged 50-69 years, of which 11.9% were screen-
detected (N=2,806), all from the North-Centre of Italy. Among screen-detected CRCs, we observed a
higher proportion of males, of cases in the distal colon, and a higher mean age of the patients. Com-
pared with pre-screening cases, screen-detected CRCs showed a better distribution by stage at diag-
nosis (OR for stage III or IV: 0.40, 95%CI: 0.36-0.44) and grading (OR for poorly differentiated CRCs
was 0.86, 95%CI: 0.75-1.00).
Screen-detected CRCs have more favourable prognostic characteristics than non-screen-detected cases.
A renewed effort to implement screening programmes throughout the entire country is recommended.
(Epidemiol Prev 2015; 39(3) Suppl 1: 108-114)
Keywords: colorectal cancer screening, colorectal cancer, Italy
1Registro tumori del Veneto,
Padova
2Registro tumori di Reggio
Emilia, Reggio Emilia
3Associazione italiana
registri tumori (AIRTUM)
4Servizio medicina
preventiva nelle comunità,
ASL di Mantova
5Gruppo italiano screening
colorettale (GISCoR)
6UO Epidemiologia
e comunicazione del rischio
AUSL Piacenza
7Istituto per lo studio
e la prevenzione oncologica
(ISPO), Firenze
8Registro tumori
dell’Emilia-Romagna,
Meldola, Forlì
9Osservatorio epidemiologico,
Registro tumori
della provincia di Sondrio,
Sondrio
10Registro tumori
di Ferrara, Ferrara
11Servizio interaziendale
di epidemiologia, AUSL
Reggio Emilia
12IRCCS-Arcispedale
S. Maria Nuova,
Reggio Emilia
13Registro tumori
di Parma, Parma
14Registro tumori
di Milano, Milano
15Registro tumori umbro
di popolazione, Scuola
di specializzazione in igiene
e medicina preventiva
Università di Perugia,
Perugia
16Istituto nazionale tumori,
CRO, Aviano (Pn)
17Osservatorio nazionale
screening (ONS)
Corrispondenza
Manuel Zorzi
manuel.zorzi@regione.veneto.it
Riassunto
L’impatto dei programmi di screening del tumore del colon retto (CRC) può essere osservato a li-
vello di popolazione solo alcuni anni dopo l’attivazione degli stessi. Abbiamo confrontato le ca-
ratteristiche dei CRC, suddivisi per modalità diagnostica (screen-detected, non-screen-
detected), come indicatore precoce di efficacia dei programmi di screening.
Sono stati raccolti da diversi Registri tumori i dati sui CRC diagnosticati in Italia dal 2000 al 2008.
Tramite linkage con gli archivi di screening è stata raccolta la modalità diagnostica dei casi, oltre al-
WWW.EPIPREV.IT
Colorectal cancers characteristics: IMPATTO study
109 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
Colorectal cancer (CRC) survival is strictly related to the stage
at diagnosis, with a better prognosis for stage I compared to stage
III and IV.1CRC screening with a biennial faecal occult blood
test (FOBT) has been shown to reduce mortality through the
early detection and treatment of cancer in large population-
based trials.2Routine, organized screening programmes (SPs)
based on the faecal immunochemical test (FIT) have been
shown to achieve even better outcomes on mortality.3Further-
more, there is some evidence that screening can also reduce in-
vasive colorectal cancer through the identification and treatment
of adenomas, preventing their transformation into cancer.4-6
The effects of screening can be observed at a population level
only several years after the implementation of screening and
only if SP participation is high. The stage at diagnosis of
screen-detected cancers is an interesting early outcome to
monitor screening programme effectiveness and predict the im-
pact on mortality, since a necessary condition to achieving a re-
duction in mortality in the short term is to detect cancer at an
earlier stage than clinically detected cancers.
With few exceptions, CRC SPs are aimed at Italian residents
aged 50 to 69 or 74 years who receive a mailed invitation to
undergo a single FIT every two years. Subjects with positive
screening tests are contacted to undergo a total colonoscopy at
an endoscopic referral centre. In only one region (Piemonte)
has a different programme been established, with either a flex-
ible sigmoidoscopy at the age of 58 or a FIT invitation every
2 years in the 59-69 years age range. The implementation of
CRC SPs started gradually in 2005-2006, and has been more
rapid in northern and central Italy than in the South. In 2008,
theoretical extension, (i.e., the proportion of the resident pop-
ulation aged 50-69 years living in areas covered by an SP), was
73.7% in the North, 56.3% in the Centre, and 21.4% in the
South and on the Islands (Sicilia and Sardegna).7
In order to describe the impact of implementing the CRC SPs
in Italy, the Italian Ministry of Health financed the IMPATTO
study, a research project that collected and linked information
from both SP archives and cancer registries.
In this paper, we used the IMPATTO study’s archives to
compare the characteristics of CRCs diagnosed in Italy from
2000 to 2008 by diagnostic modality (screen-detected, non-
screen-detected).
MATERIALS AND METHODS
Data
The data collected in the IMPATTO study database have been
described in the associated paper of this article.8Briefly, for the
purpose of this paper, CRCs diagnosed in patients aged 50-69
years were selected and characterized according to the following
patterns of diagnosis:
■
CRCs diagnosed in areas where an SP has been implemented:
■
pre-screening (i.e., diagnosed before the onset of the SP);
■
screen-detected;
■
not screen-detected, diagnosed after the onset of the SP;
■
CRCs diagnosed in areas where no SP has been implemented.
Analysis
The Chi square test was used to compare the distribution of
the main CRC characteristics included in the study by pattern
of diagnosis: anatomic sub-site, stage at diagnosis, grading,
number of lymph nodes examined and positive lymph nodes.
The association between pattern of diagnosis and CRC char-
acteristics was evaluated using logistic regression models which
included the variables that resulted significantly associated at
univariate analysis. In particular, we explored which factors
were associated with stage and grading, including the pattern
of diagnosis among the explanatory variables.
RESULTS
Overall, the study included 23,668 invasive cases of CRCs di-
agnosed in subjects aged 50-69 years between 2000 and 2008.
The cancer registries took part in the study with cases from dif-
ferent periods. Moreover, the SPs were introduced in different
years. In particular, the SPs were implemented in most areas
during 2005-2006, as opposed to Veneto (2002) and Firenze-
Prato, where SPs were already in place at the beginning of this
study. Finally, there were no SPs in the South and on the Islands
during the study period.
Table 2 (p. 110) shows the main characteristics by macro-area
and period: the North-Centre in 2000-2005 (before the SPs
became widespread), the North-Centre with SPs (2006-2008),
and the South and the Islands. The cases from the latter
macro-area represented about one-sixth of the overall study
(15.8%). As expected, the largest proportion of cases was
males (59%) from the older age group.
l’area geografica e alla presenza di un programma di screening organizzato. Abbiamo confrontato le principali caratteristiche
dei diversi sottogruppi di CRC tramite modelli di regressione logistica multivariata.
Lo studio riguarda 23.668 CRC diagnosticati in soggetti di età 50-69 anni, l’11,9% dei quali screen-detected (N=2.806),
tutti di aree del Nord o Centro Italia. Tra i casi screen-detected abbiamo osservato una maggiore proporzione di maschi, di
casi a carico del colon distale e un’età media più alta. Rispetto ai casi diagnosticati prima dell’attivazione degli screening, i
casi screen-detected avevano una migliore distribuzioneper stadio alla diagnosi (odds ratio per stadio III o IV: 0,40; IC95%:
0,36-0,44) e grading (OR per grading scarsamente differenziato: 0,88; IC95%: 0,75-1,00).
I casi screen-detected avevano caratteristiche prognostiche migliori anche rispetto ai casi non-screen-detected. Si raccomanda
uno sforzo rinnovato per attivare programmi di screening colorettale in tutto il territorio nazionale.
(Epidemiol Prev 2015; 39(3) Suppl 1: 108-114)
Keywords: screening colorettale, tumore del colon retto, Italia
Colorectal cancers characteristics: IMPATTO study
110 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Macro-area Cancer registry Number of cases
2000 2001 2002 2003 2004 2005 2006 2007 2008 Total
Northwest Genova 312 294 283 889
Milano 449 456 415 427 399 367 452 2,965
Sondrio 47 41 54 50 48 54 73 77 67 511
Biella 64 54 64 57 67 54 63 423
Northeast Trentino 120 129 117 138 140 644
Veneto 138 155 166 188 179 162 988
Friuli-Venezia Giulia 418 395 369 1,182
Emilia-Romagna 137 240 629 954 1,334 1,519 1,994 1,565 385 8,757
Centre Firenze-Prato 341 343 319 322 323 319 1,967
Umbria 287 296 266 383 372 1,604
South-Islands Latina 126 129 108 132 495
Napoli 88 90 111 116 123 528
Siracusa 78 80 89 88 82 417
Palermo 268 261 258 787
Catania-Messina 318 366 368 1,052
Sassari 114 117 117 111 459
Figures in color represent the years when a screening programme was active
Table 1. Number of colorectal cancer cases by cancer registry and calendar year. Ages 50-69 years.
Tabella 1. Casi di tumore del colon retto per registro tumori e anno. Età 50-69 anni.
North-Centre 2000-2005 North-Centre 2006-2008 South and the Islands 2000-2008
N%N%p-value1N%p-value1
Total 13,275 100 6,655 100 3,738 100
Gender
male 7,817 58.9 4,075 61.2 0.002 2,151 57.5 0.14
female 5,458 41.1 2,580 38.8 1,587 42.5
Age (years)
50-59 4,291 32.3 2,115 31.8 0.44 1335 35.7 <0.001
60-69 8,984 67.7 4,540 68.2 2,403 64.3
Pattern of diagnosis
screen-detected 569 4.3 2,237 33.6 <0.001 00-
not screen-detected* 12,706 95.7 4,418 66.4 3,738 100
Anatomic site
proximal colon 3,557 26.8 1,776 26.7 <0.001 1,001 26.8 <0.001
distal colon 4,820 36.3 2,631 39.5 1,152 30.8
rectum 4,276 32.2 1,890 28.4 1,321 35.3
colon NOS** 622 4.7 358 5.4 7.1
Stage at diagnosis
I 2,146 16.2 1,878 28.2 <0.001 471 12.6 <0.001
II 3,299 24.9 1,518 22.8 905 24.2
III 3,817 28.8 1,598 24.0 852 22.8
IV 2,461 18.5 1,087 16.3 841 22.5
not available/missing 1,552 11.7 574 8.6 669 17.9
Grading
well-differentiated 1,165 8.8 935 14.0 <0.001 232 6.2 <0.001
moderately differentiated 7,792 58.7 3,740 56.2 2,290 61.3
poorly differentiated 1,981 15.0 1,178 17.7 531 14.2
not available/missing 2,337 17.6 802 12.1 685 18.3
1p-value of Chi square test comparing the distribution by each variable in the table with the reference group = North-Centre, 2000-2005
* it includes pre-screening, not screen-detected in areas with screening, diagnosed in areas with no screening
** NOS:not otherwise specified
Table 2. Distribution of colorectal cancer cases according to main characteristics, by macro-area and period.
Tabella 2. Distribuzione dei casi di tumore del colon retto per varie caratteristiche, per macroarea e periodo.
Colorectal cancers characteristics: IMPATTO study
111 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
There were 2,806 screen-detected cases, or 11.9% of the
whole.This percentage rose to 33.6% in areas with an SP. One-
third of the cases (31.6%) were localized in the rectum.
The proportion of stage I cases and of cases with grade I was
highest in the North-Centre in 2006-2008 and lowest in the
South and on the Islands. Overall, the proportion of cases with
a stage missing at diagnosis was 11.8%.This was highest in the
South and on the Islands, and lowest in the North-Centre in
2006-2008.
Table 3 shows the characteristics of cases by pattern of diag-
nosis. Compared to the CRCs diagnosed during the pre-
screening period, screen-detected CRCs showed a different dis-
tribution for most variables. In particular, the proportion of
subjects aged 65-69 years was greater than 40% (41.7%), as
compared to 38.3%. Screen-detected cases were more fre-
quently located in the distal colon than pre-screening cancers
(50.6% vs 36.8%). Grading was more favourable, with 20.1%
of screen-detected cases being well-differentiated and only
11% poorly differentiated, compared respectively to 9.6%
and 15.6% in the pre-screening period. Also stage at diagno-
sis was less advanced: 42.8% of screen-detected cases were di-
agnosed at stage I (vs 16.2%) and only 6.2% at stage IV (vs
19.8%).
Finally, the number of lymph nodes examined in screen-detected
CRCs was similar to pre-screening cases (15.6 in both groups),
while the mean number of positive lymph nodes overall and for
cases stages III/IV was significantly lower in the former (1.0 vs
2.1 and 3.4 vs 4.2, respectively).
Both not screen-detected CRCs and CRCs diagnosed in areas
without screening showed similar distributions to those of
CRC in the pre-screening period, according to major charac-
teristics (except macro-area and number of lymph nodes).
Compared with the CRCs diagnosed before implementation of
the screening programmes, the probability of stage III or IV at
Total Areas with a screening programme Areas
period with screening without
N%pre-screening screen- not screen- a screening
period detected detected programme
Total (N) 23,668 100 6,710 2,806 6,759 7,393
Macro-area
Northwest 4,788 20.2 39.1 6.7 16.1 12.0
Northeast 11,571 48.9 52.2 74.4 47.6 37.4
Centre 4,066 17.2 8.7 18.9 36.4 6.7
South-Islands 3,243 13.7 0.0 0.0 0.0 43.9
Gender
male 4,043 59.3 58.8 61.7 60.0 158.3
female 9,625 40.7 41.2 38.4 40.0 41.7
Mean age (years) (SD) 61.8 (5.3) 61.8 (5.3) 62.3 (5.2) 61.7 (5.4) 61.7 (5.3)
Mean age (years)
50-54 2,954 12.5 12.5 9.9 13.6 12.5
55-59 4,787 20.2 19.8 20.2 19.6 21.2
60-64 6,821 28.8 29.4 28.2 28.3 29.0
65-69 9,106 38.5 38.3 41.7 38.5 37.4
Anatomic site
proximal colon 6,334 26.8 27.2 24.3 27.8 26.4
distal colon 8,603 36.4 37.8 50.6 35.7 30.2
rectum 7,487 31.6 31.2 23.2 32.5 34.5
colon NOS 1,244 5.3 3.9 2.0 4.0 8.9
Grading
well-differentiated 2,332 9.9 9.6 20.1 9.4 6.5
moderately differentiated 13,822 58.4 57.2 56.9 56.5 61.8
poorly differentiated 3,690 15.6 15.6 11.0 16.7 16.3
not available/missing 3,824 16.2 17.5 12.0 17.4 15.4
Lymph nodes examined
mean number (SD) 16.1 (9.9) 15.6 (9.3) 15.6 (9.7) 18.0 (11.0) 14.5 (8.7)
Positive lymph nodes
mean number (SD) 2.0 (4.2) 2.1 (4.0) 1.0 (2.7) 2.4 (4.9) 2.1 (4.2)
Positive lymph nodes in stage III/IV cases
mean number (SD) 4.3 (5.3) 4.2 (4.9) 3.4 (4.0) 4.6 (6.0) 4.3 (5.0)
Stage at diagnosis
I 4,495 19.0 16.2 42.8 17.2 14.2
II 5,722 24.2 24.9 19.1 24.4 25.2
III 6,267 26.5 27.5 20.1 28.3 26.3
IV 4,389 18.5 19.8 6.2 20.3 20.5
not available/missing 2,795 11.8 11.6 11.9 9.8 13.8
Table 3. Distribution of colo-
rectal cancer cases according to
main characteristics, by pattern
of diagnosis.
Tabella 3. Distribuzione dei
casi di tumore del colon retto
per varie caratteristiche, per mo-
dalità di diagnosi.
Colorectal cancers characteristics: IMPATTO study
112 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
diagnosis was reduced by 60% among screen-detected cases
(table 4, p. 112). Instead, there were no significant differences
regarding cases that were not screen-detected and cases diag-
nosed in areas with no screening.The risk of an advanced stage
decreased with age and was lower for cases of cancer located in
the distal colon and the rectum.
The probability of a poorly differentiated grading was signif-
icantly lower (14%) among screen-detected cases (table 5) as
compared to the period prior to screening, even when adjust-
ing for stage at diagnosis (OR not adjusted for stage was 0.62,
95%CI 0.54-0.71), while non-screen-detected cases and cases
diagnosed in areas with no screening did not show a different
risk from pre-screening CRCs. The risk of poorly differentiated
grading was higher in cases with an advanced stage at diagno-
sis and for cases located in the proximal colon.
DISCUSSION
Using data collected from the large number of CRCs diag-
nosed from 2000 to 2008, this study found that screen-de-
tected CRCs significantly differ from non-screen-detected
ones. In particular, the study confirms what is expected by the
diagnostic anticipation of screenings, i.e., more favourable
prognostic characteristics of screen-detected CRCs: a better
distribution by stage at diagnosis and by grading, and a
lower number of positive lymph nodes overall and for stage
III/IV cases.
Compared to non-screen-detected cases, the proportion of
screen-detected CRCs in males was higher, as was the mean age
of the patients.
We also observed a higher proportion of CRCs in the distal
colon. This figure could be due to the FIT’s higher sensitivity
to lesions of the left colon9-11 and hence to a higher impact on
the prevalence round of screening at this anatomic site. Most
screen-detected cases included in this study were diagnosed in
the first or second screening round, when many of the preva-
lent pre-clinical lesions are detected, thus producing a transient
increase in incidence rates. Another reason why screen-de-
tected CRCs are more frequent in the distal colon could de-
N stage III-IV* N stage I-II Odds ratio** 95%CI
Gender
male 6,249 6,104 1* -
female 4,407 4,113 1.02 0.97-1.08
Age (5-year linear increase) 0.91 0.89-0.94
Anatomic site
proximal colon 3,166 2,770 1* -
distal colon 3,647 4,019 0.83 0.77-0.88
rectum 3,226 3,026 0.91 0.84-0.97
colon NOS 617 402 1.19 1.04-1.37
N examined lymph nodes 1.002 1.001-1.003
Pattern of diagnosis
pre-screening 3,182 2,774 1* -
screen-detected 737 1,734 0.40 0.36-0.44
not screen-detected 3,566 3,090 1.04 0.97-1.12
areas with no screening 3,171 2,619 1.05 0.97-1.13
* reference
** estimated using logistic regression model (response variable stage III-IV vs. stage I-II), adjusted by all the variables in the table
Table 4. Odds ratios of colo-
rectal cancers diagnosed at
stage III or IV (as compared to
stage I-II), according to selec-
ted variables.
Tabella 4. Odds ratio di sta-
dio avanzato (III o IV), per di-
verse variabli.
N poorly N well mod. Odds ratio** 95%CI
differentiated* differentiated
Gender
male 2,120 9,627 1* -
female 1,570 6,527 1.07 1.00-1.16
Age (5-year linear increase) 1.00 0.96-1.03
Anatomic site
proximal colon 1,401 4,114 1* -
distal colon 1,104 6,411 0.54 0.49-0.59
rectum 980 4,973 0.61 0.56-0.68
colon NOS 205 656 0.93 0.78-1.11
N examined lymph nodes 1.00 0.998-1.00
Pattern of diagnosis
pre-screening 1,075 4,491 1* -
screen-detected 308 2,160 0.86 0.75-1.00
not screen-detected 1,298 4,859 1.06 0.96-1.16
areas with no screening 1,009 4,644 0.96 0.87-1.06
* reference
** estimated using logistic regression model (response variable stage III-IV vs. stage I-II), adjusted by all the variables in the table
Table 5. Odds ratios of poorly
differentiated grading colo-
rectal cancers (as compared
to well/moderately differen-
tiated), according to selected
variables.
Tabella 5. Odds ratio di gra-
ding scarsamente differenzia-
to, per diverse variabli.
Colorectal cancers characteristics: IMPATTO study
113 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
pend on the different biology of these lesions, that have been
associated to a slow natural history with a long pre-clinical
phase.12 This would increase the difference in diagnostic yield
of the distal versus the proximal colon.
Age was inversely correlated to the probability of stage III or
IV at diagnosis. The prevalence round of screening (which oc-
curs at a younger age) could play a role in this effect.
The number of lymph nodes examined was positively associ-
ated with a more advanced stage. The interpretation of this ef-
fect is controversial: on the one hand, the higher likelihood of
a staging upgrade the more lymph nodes are examined; on the
other hand, it could be hypothesized that more lymph nodes
are examined in more advanced cancers.
In areas with an SP, the proportion of screen-detected cases was
about one-third of the total. Besides the diagnostic sensitivity
of the first-level test and second-level assessment, this figure de-
pends on the extension of invitations and compliance with in-
vitation to screening. Even though this study monitored the
impact of screening in the first years after SP implementation
(when the spread of screening over the target population is rea-
sonably lower than expected in well-established programmes),
we observed a relevant impact of screening even when evalu-
ating all the CRCs diagnosed in the entire population.
Compared to the North-Centre, cases in the South and Islands
showed a worse distribution by stage at diagnosis and by grad-
ing. These figures suggest a diagnostic delay in this macro-area
that was worsened by the increase in the number of SPs in the
North-Centre. This hypothesis is in line with the results from
the latest report of the Italian association of cancer registries
(AIRTUM) on cancer patient survival. CRCs diagnosed dur-
ing 2001-2004 in the South and on the Islands showed a
lower 5-year survival rate compared to cases diagnosed in
other areas of Italy.13
Another important result of this study was that, after screening
was implemented, the cases diagnosed before the onset of an SP
and those not screen-detected in the same areas were very sim-
ilar in terms of distribution by age and anatomic site, stage at di-
agnosis, and grading. The only exception was the number of ex-
amined and positive lymph nodes, which was higher in the latter
group. However, this figure could be due to a period effect.
The cases that were diagnosed outside the SPs were not differ-
ent from the cases detected beforethe onset of screening.There-
fore, they do not seem to have been significantly affected by SP
implementation. This fact has at least three consequences:
■
the presence of an SP does not seem to generate a “halo” ef-
fect (i.e., an increase in the spontaneous, extra-screening, up-
take of FIT and/or total colonoscopy) to produce a visible di-
agnostic anticipation; this hypothesis needs to be confirmed in
areas where SPs have been active for more years;
■
non-screen-detected cases are representative of the cases that
were diagnosed in the absence of SPs, therefore they can be safe-
ly used as a comparison group for screen-detected CRCs;
■
the differences that we observed in the screen-detected cas-
es may be entirely attributed to the specific pattern of diagnosis.
The risk of selection bias (i.e., compliance with the screening
invitation being higher among healthier subjects, who would
have a more favourable pattern of disease even without an SP)
seems unlikely. Otherwise, non-screen-detected cases would have
shown worse characteristics than cases diagnosed before the on-
set of screening.
This is in line with data from a national survey on preventive
behaviours and service utilization, which showed that in Italy
spontaneous screening for CRC is very low and the coverage
in regions with well-implemented population-based SPs is
higher among subjects with a lower educational level.14
However, this picture could be modified as SPs age and fol-
lowing changes in compliance with invitation.
CONCLUSION
Screen-detected CRCs showed a favourable distribution by dif-
ferent prognostic factors, while cases diagnosed in the South
and on the Islands reported the worst figures.
A renewed effort to implement screening programmes through-
out the entire country, and particularly in the South and on the
Islands, is therefore warranted, filling the prognostic gap
among geographic areas, to increase the equity of access to a
public health programme that is proving to be highly protec-
tive of the population.
Conflicts of interests: none declared
Colorectal cancers characteristics: IMPATTO study
114 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. Ward KC, Young JL, Ries LAG . Cancers of the colon and rectum.
In: Ries LAG, Young JL, Keel GE et al. eds. SEER survival mono-
graph: cancer survival amongst adults: U.S. SEER Program, 1998-
2001, patient and tumor characteristics. NIH Pub. No. 07-6215.
National Institutes of Health, Bethesda, Md, 2007:33-42.
2. Hewitson P, Glasziou P et al. Cochrane systematic review of col-
orectal cancer screening using the fecal occult blood test (hemo-
ccult): an update. Am J Gastroenterol 2008;103(6):1541-49.
3. Zorzi M, Fedeli U, Schievano E et al. Impact on colorectal cancer
mortality of screening programmes based on the faecal immuno-
chemical test. Gut 2014 Sep 1. pii: gutjnl-2014-307508. doi:
10.1136/gutjnl-2014-307508.
4. Kronborg O, Jorgensen OD, Fenger C et al. Randomized study of
biennial screening with a faecal occult blood test: Results after nine
screening rounds. Scand J Gastroenterol 2004;39:846-51.
5. Hardcastle JD, Chamberlain JO, Robinson MH et al. Randomised
controlled trial of faecal-occult-blood screening for colorectal can-
cer. Lancet 1996;348:1472-77.
6. Ventura L, Mantellini P, Grazzini G et al. The impact of immuno-
chemical faecal occult blood testing on colorectal cancer inci-
dence. Dig Liver Dis 2014;46(1):82-86. doi:pii: S1590-8658
(13)00280-6. 10.1016/j.dld.2013.07.017.
7. Zorzi M, Baracco S, Fedato C et al. Screening for colorectal can-
cer in Italy: 2008 survey. Epidemiol Prev 2010;34 (Suppl 4):53-72.
8. Zorzi M, Mangone L, Sassatelli R et al. Incidence trends of col-
orectal cancer in the early 2000s in Italy. Figures from the
IMPATTO study on colorectal cancer screening. Epid Prev
2015;3(Suppl 1):115-25.
9. Haug U, Kuntz KM, Knudsen AB et al. Sensitivity of immuno-
chemical faecal occult blood testing for detecting left- vs right-
sided colorectal neoplasia. BJC 2011b;104:1779-85.
10. Morikawa T, Kato J, Yamaji Y et al. Sensitivity of immunochemical
fecal occult blood test to small colorectal adenomas. Am J Gas-
troenterol 2007;102:2259–64.
11. Ahlquist DA, Sargent DJ, Loprinzi CL et al. Stool DNA and occult
blood testing for screen detection of colorectal neoplasia. Ann
Intern Med 2008;149:441-50.
12. Missiaglia E, Jacobs B, D'Ario G et al. Distal and proximal colon
cancers differ in terms of molecular, pathological, and clinical
features. Ann Oncol 2014;25(10):1995-2001. doi:10.1093/
annonc/mdu275. Epub 2014 Jul 23. PubMed PMID: 25057166.
13. AIRTUM Working group. Italian cancer figures, report 2011: Survival
of cancer patients in Italy. Epidemiol Prev 2011;5-6(Suppl 3):85.
14. Rapporto nazionale PASSI 2013: screening per il cancro del colon
retto. Available at: http://www.epicentro.iss.it/passi/rapporto2013/
Colonretto.asp [Accessed November 24, 2014].
Parma: P. Sgarzi, F. Bozzani (Registro tumori di Parma); M. Zatelli,
C. Zurlini (Centro screening oncologici AUSL Parma); P. Caruana
(AOU Parma)
Piacenza: P. Seghini, G. Gatti, R. Prazzoli (UO Epidemiologia
e comunicazione del rischio, AUSL Piacenza)
Reggio Emilia: C. Campari, L. Paterlini (Centro screening, AUSL
Reggio Emilia); T. Cassetti (Registro tumori di Reggio Emilia);
R. Sassatelli (UO Gastroenterologia ed endoscopia digestiva,
Arcispedale Santa Maria Nuova – IRCCS, Reggio Emilia)
Romagna: C. Imolesi (Centro screening, OCM Bufalini, Cesena);
C. Casale (Centro Screening, O. Infermi, Rimini); M. Serafini (Centro
Screening, OC S.M. delle Croci); R. Vattiato, O. Giuliani (Registro
tumori della Romagna, Istituto dei tumori della Romagna IRCCS,
Meldola)
Sassari: R. Cesaraccio, O. Sechi, M. Budroni (Registro tumori
di Sassari)
Siracusa: A. Madeddu, M.L. Contrino, A. Colanino Ziino, M. Russo,
F. Tisano (Registro tumori di Siracusa)
Sondrio: A.C. Fanetti, S. Maspero, E. Moroni, I. Cometti
(Osservatorio epidemiologico - Registro tumori della Provincia
di Sondrio)
Trentino: M. Gentilini, S. Piffer (Registro tumori di Trento);
G. De Pretis (UO Gastroenterologia, Ospedale S. Chiara, Trento);
P. Caciagli (Dipartimento laboratorio e servizi, APSS, Trento); R. Pertile
(Servizio epidemiologia clinica e valutativa, APSS, Trento)
Umbria: F. Bianconi (Registro tumori Umbro di Popolazione / Scuola
di specializzazione in Igiene e medicina preventiva, Università
di Perugia); D. Bucchi, M.E. Galeotti (Scuola di specializzazione
in Igiene e medicina preventiva, Università di Perugia); M. Malaspina
(Servizio di screening USL 1 Umbria)
Veneto: A. Greco, A.R. Fiore, C.F. Stocco (Registro tumori del Veneto,
Padova); C. Fedato (Coordinamento regionale screening, Regione
Veneto, Venezia)
Biella: A. Giacomin (Registro tumori del Piemonte, Provincia
di Biella), A. Azzoni (S.S. Gastroenterologia, ASL Biella)
Bologna: P. Baldazzi, N. Collina, P. Pandolfi, P. Biavati, G. Gualandi
(Registro tumori di Bologna)
Catania-Messina: S. Sciacca, P. Pesce, A. Torrisi, C. Sciacchitano,
M. Fidelbo (Registro tumori integrato di Catania-Messina, Catania)
Emilia-Romagna: A.C. Finarelli, C. Naldoni, P. Sassoli de’ BIanchi,
P. Landi (Assessorato alle politiche per la salute, Regione Emilia-
Romagna, Bologna)
Ferrara: V. Matarese (UO Gastroenterologia, AOU S. Anna,
Cona-Ferrara), A. De Togni, C. Palmonari (Centro screening
oncologici AUSL Ferrara)
Firenze: E. Crocetti, G. Grazzini, G. Manneschi, P. Mantellini
(Istituto per lo studio e la prevenzione oncologica – ISPO, Firenze)
Friuli-Venezia Giulia: D. Serraino, E. Bidoli, M. Taborelli, A. Gini,
S. Virdone (Istituto nazionale dei tumori – CRO, Aviano)
Genova: A. Puppo, C. Casella, M. Celesia, R. Cogno, E. Marani
(Registro tumori Regione Liguria, UO Epidemiologia clinica, IRCCS
AOU San Martino – IST, Genova)
Latina: E. Bugliarello, S. Fattoruso, L. Tamburo, S. Tamburrino
(Registro tumori di Latina, Latina); P. Bellardini (Coordinamento
screening ASL Latina)
Milano: M. Autelitano, B. Frammartino, L. Bisanti, S. Ghilardi,
R. Leone (Registro tumori di Milano)
Modena: R. Corradini, F. De Girolamo (Centro screening oncologici
AUSL di Modena, Italy); K Valla (Registro tumori di Modena)
Napoli: R. Palombino, L. Gigli, S. Russo Spena (Servizio
epidemiologia e prevenzione, ASL Napoli 3 Sud); M.F. Vitale (Registro
tumori di popolazione Regione Campania, Napoli)
Palermo: M.A. Cascio, R. Mannino, W. Mazzucco, A. Mistretta,
B. Ravazzolo (Registro tumori di Palermo e Provincia, AOU Policlinico
“Paolo Giaccone” di Palermo, UOC di epidemiologia clinica
con Registro tumori, Palermo)
Members of the IMPATTO COLONRETTO working group:
Membri del gruppo di studio IMPATTO COLONRETTO:
Incidence trends of colorectal cancer
in the early 2000s in Italy.
Figures from the IMPATTO study on colorectal
cancer screening
Trend di incidenza del tumore del colon retto
nei primi anni Duemila in Italia.
Dati dello studio IMPATTO dello screening colorettale
Manuel Zorzi,1Lucia Mangone,2,3 Romano Sassatelli,4,5 Susanna Baracco,1Mario Budroni,6
Marine Castaing,7Claudia Cirilli,8Rosanna Cusimano,9Mario Fusco,10 Adriano Giacomin,11
Paolo Giorgi Rossi,12,13 Carlo Naldoni,14 Fabio Pannozzo,15 Silvano Piffer,16 Antonella Puppo,17
Francesco Tisano,18 Marco Zappa19,20 and IMPATTO COLONRETTO working group
Abstract
We utilised the IMPATTO study’s archives to describe the 2000-2008 colorectal cancer (CRC) incidence
rate trends in Italy, once screening programmes based on the faecal immunochemical test were im-
plemented in different areas.
Data on CRCs diagnosed in Italy from 2000 to 2008 in subjects aged 40-79 years were collected by
23 cancer registries. Incidence rate trends were evaluated as a whole and by macro-area (North-Cen-
tre and South-Islands), presence of a screening programme, sex, ten-year age class, anatomic site, stage
at diagnosis, and pattern of diagnosis (screen-detected, non-screen-detected). The annual percent
change (APC) of incidence rate trends, with 95% confidence intervals (95%CI), were computed.
The study included 46,857 CRCs diagnosed in subjects aged 40-79 years, of which 2,806 were screen-
detected. The incidence rates in the North-Centre were higher than in the South and on the Islands.
During the study period, screening programmes had been implemented only in the North-Centre and
had a significant effect on incidence rates, with an initial sharp increase in incidence, followed by a
decrease that started in the 3rd-4th years of screening. These incidence rate trends were exclusively
due to modifications in the rates of stage I cases. After screening programmes started, incidence in-
creased in all anatomic sites, particularly in the distal colon.
The differential figures introduced by the implementation of screening programmes warrant a con-
tinuous surveillance of CRC incidence and mortality trends to monitor the impact of screening at a na-
tional level.
(Epidemiol Prev 2015; 39(3) Suppl 1: 115-125)
Keywords: colorectal cancer, screening, incidence rates, fecal immunochemical test, Italy
Riassunto
E’ stato utilizzato l’archivio dello studio IMPATTO per descrivere i trend di incidenza del tumore
del colon retto (CCR) in Italia nel periodo 2000-2008, quando sono stati avviati programmi di scree-
ning colorettale basati sul test per la ricerca del sangue occulto fecale in diverse aree.
23 Registri tumori hanno fornito i dati relativi ai CCR diagnosticati nel periodo 2000-2008 in sog-
WWW.EPIPREV.IT
115 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
1Registro tumori del Veneto,
Padova
2Registro tumori di Reggio
Emilia, Reggio Emilia
3Associazione italiana registri
tumori (AIRTUM)
4UO Gastroenterologia
ed endoscopia digestiva,
IRCCS Arcispedale Santa Maria
Nuova, Reggio Emilia
5Gruppo italiano screening
colorettale (GISCoR)
6Registro tumori
di Sassari, Sassari
7Registro tumori integrato
di Catania-Messina, Catania
8Registro tumori
di Modena, Modena
9Registro tumori
di Palermo, Palermo
10Registro tumori
di popolazione Regione
Campania, Napoli
11Registro tumori
del Piemonte,
Provincia di Biella, Biella
12Servizio interaziendale
di epidemiologia, AUSL
Reggio Emilia
13IRCCS-Arcispedale
S. Maria Nuova,
Reggio Emilia
14Assessorato alle politiche
per la salute, Regione
Emilia-Romagna, Bologna
15Registro tumori di Latina,
Latina
16Registro tumori di Trento,
Trento
17Registro tumori Regione
Liguria, Genova
18Registro tumori di Siracusa,
Siracusa
19Istituto per lo studio
e la prevenzione oncologica
(ISPO), Firenze
20Osservatorio nazionale
screening (ONS)
Corrispondenza
Manuel Zorzi
manuel.zorzi@regione.veneto.it
Colorectal cancers incidence trends: IMPATTO study
116 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
INTRODUCTION
Colorectal cancer (CRC) is a major public health problem. In
Italy it represents the most frequent tumour in terms of incidence
with more than 50,000 new cases and is the second cause of
death among cancers, with about 19,000 deaths per year.1
According to estimates by the Italian Association of Cancer
Registries (AIRTUM), mortality rates showed a reduction in
both genders starting in the early 1990s, while incidence rates
increased, particularly in males.2
A number of case series from cancer registries in the North of
Italy showed that at the end of the 1990s the proportion of cases
that were TNM stage III or IV at diagnosis still ranged between
39% and 51% of the total.3-5 Stage at diagnosis is well known
to be closely related to prognosis: a case series from the SEER
study showed a 5-year survival of 93% for cases at AJCC stage
I, 80% for those at stage II, 58% at stage III, and only 7% at
stage IV.6The SEER study compared series of cases diagnosed
in different periods (from 1973 to 1997) and showed that the
increase in stage-specific survival had been very limited.7
Therefore, the reported increase in survival, from 51% in 1990-
1992 to 64% in 2005-2007,8is plausibly associated with a
more favourable distribution of stage at diagnosis, which derived
from the spread of the uptake of exams for early diagnosis, first
spontaneously and then within organized screening programmes
(SP). The distribution by stage at diagnosis of screen-detected
CRCs is better than that of clinically diagnosed CRCs, with
more than 50% of cases at stage I, while those at stage III or IV
are about one-fourth of the total.9-11
In Italy, CRC SPs were progressively implemented in most re-
gions starting in the early 2000s. By the end of 2010, 66% of
the Italian population lived in areas with active CRC screening
programmes, but strong geographical differences were present:
the corresponding figures were 87% in the North, 79% in the
Centre and only 29% in the South and Islands.9
Four randomized controlled trials showed that SPs based on the
guaiac faecal occult blood test (gFOBT) reduce mortality by
16%,12,15-19 which rises to 23% in the per-protocol analysis.12
Results from a gFOBT population-based SP showed similar fig-
ures,13 while early evidence from faecal immunochemical test
(FIT)-based programmes reported a greater reduction in mor-
tality that began earlier compared to the trials, i.e., in the 5th year
after screening started.14
A population-based SP is expected to initially increase incidence
rates, thanks to the diagnostic anticipation of cancers that would
otherwise be diagnosed later. In the medium and long term, a
progressive reduction of incidence rates is expected, deriving from
the prevention of new CRCs as a result of the detection and re-
moval of a large number of precancerous lesions (i.e., advanced
adenomas). As a matter of fact, the four trials showed contrast-
ing effects on the incidence rates, with a 17-20% reduction in one
of them15-16 but no effect in the other three.17-19 The latter re-
ported low compliance with the study by the enrolled subjects
(respectively 67%, 60%, and 63%).
A recent paper showed a reduction of incidence rates in the
medium term (22% 11 years after screening started).20 Many
studies have shown that FIT sensitivity for advanced adenoma
and cancer is higher than that of gFOBT.21-26 Thus the effect
on incidence observed in screening programmes and not in tri-
als could be due to FIT having a higher sensitivity for adenomas
than gFOBT.
In Italy, CRC SPs are aimed at residents aged 50-69 or 74 years,
who are invited via mail every 2 years to perform a single FIT.
Subjects with a positive screening test are contacted to undergo
a total colonoscopy performed at an endoscopic referral centre.
In only one region (Piemonte) has a different programme been
established, with either one sigmoidoscopy at the age of 58, or a
FIT invitation every 2 years in the age interval of 59-69 years. The
average detection rate of advanced adenomas in organized pro-
grammes in Italy is high, compared to that of guaiac trials,18,27
reaching 13 x 1,000 at the prevalence round and 8 x 1,000 at the
incidence round, respectively.11 Consequently, the impact of
screening programmes on incidence is an open question.
To describe the impact that implementing CRC screening pro-
grammes has had in Italy, a research project, the IMPATTO
study, was financed by the Italian Ministry of Health; the study
collects and links information from both screening programme
archives and cancer registries.
This paper utilizes the IMPATTO study’s archives to describe the
CRC incidence rate trends in Italy during 2000-2008, when sev-
eral SPs were implemented in different areas.
getti di età compresa fra 40 e 79 anni. Sono stati calcolati i trend di incidenza complessivi e per macroarea (Centro-Nord e Sud-
Isole), presenza di un programma di screening, sesso, età, localizzazione anatomica, stadio alla diagnosi e modalità diagnostica
(screen-detected, non-screen-detected). Sono riportati gli APC (annual percent change) con intervalli di confidenza al 95%.
L’archivio riguarda 46,857 CRC, di cui 2,806 screen-detected. I tassi di incidenza nel Centro-Nord erano maggiori rispetto
al Sud-Isole. Nel periodo di studio sono stati avviati programmi di screening solo in aree del Centro-Nord, con un effetto
significativo sui tassi di incidenza, con un ripido incremento iniziale seguito da una riduzione a partire dal 3°-4° anno dal-
l’avvio dei programmi. L’effetto degli screening era a carico esclusivamente dei CCR in stadio I alla diagnosi. Dopo l’avvio
degli screening, l’incidenza è aumentata per tutte le sottosedi anatomiche del colon, in particolare per il colon distale.
L’avvio dei programmi di screening colorettale in Italia ha avuto un forte impatto portando a un aumento dell’incidenza
e delle forme precoci. E’ necessario un continuo monitoraggio delle aree italiane per capire gli effetti dello screening su
tutta la popolazione.
(Epidemiol Prev 2015; 39(3) Suppl 1: 115-125)
Parole chiave: tumore del colon retto, screening colorettale, tassi di incidenza, test per la ricerca del sangue occult fecale, Italia
MATERIALS AND METHODS
Data
The IMPATTO study collected data from CRC cases (Inter-
national Classification of Diseases, 10th revision: C18–C20)
in subjects aged 40-79 years that were diagnosed between
2000 and 2008 in the populations covered by 23 population-
based cancer registries (CR) in 13 Italian regions (Piemonte,
Liguria, Lombardia, Veneto, Trentino, Friuli-Venezia Giulia,
Emilia-Romagna, Toscana, Umbria, Lazio, Campania, Sicilia,
Sardegna). These areas included about 36%, 17%, and 24%
of the resident population in northern, central, and southern
Italy, respectively.
Cases based on death certificates only, autopsies without his-
tology, or autopsies with histology and incidence data equal to
date of death were excluded. All multiple metachronous cases
were included.
Collected data included incidence date, morphology and to-
pography, stage at diagnosis (according to Dukes’ classification
as modified by Astler and Coller39) and grading, surgical in-
tervention, lymph nodes examined and positive lymph nodes.
Multiple synchronous cases (incidence date within six months
from the index case) were recorded if located in different anatomic
sub-sites (fourth digit of the ICD-10 topography code) and only
the most advanced were staged. If more cancers were located in
the same sub-site, only the most advanced was recorded, main-
taining the recording rules of different morphologies.
Vital status was recorded for all cases up to either 31.12.2008
or 31.12.2010, according to the CR. Information about the
cause of death was collected for deceased subjects, according to
the International Classification of Diseases, 9th revision.
Tumour histological type was recorded according to the In-
ternational Classification of Diseases for Oncology, 3rd edition.
CRs carried out a record-linkage with the local SPs to retrieve
individual data on the screening history of patients before the
incidence date by collecting the date of the first invitation and
the dates of screening tests. Patients were then classified ac-
cording to the following screening patterns:
■
screen-detected at the first screening episode;
■
screen-detected at a repeat screening episode;
■
screen-detected at follow-up;
■
not compliant with diagnostic work-up after a positive screen-
ing test;
■
subjects with at least one negative screening test before in-
cidence;
■
never compliant (i.e., invited, but not tested within the SP);
■
never invited to screening.
Two categories were then created according to the diagnostic
modality: screen-detected cases, including the first three classes,
and non-screen-detected cases, including the last four.
Finally, age- and sex-specific data on the resident population
in the study period for each CR were collected.
Analysis
Cases were classified by geographic macro-area according to the
Istat (Italian National Statistics Agency) classification: North-
west, Northeast, Centre, and South and Islands. They were
then grouped into two epidemiologically homogeneous areas,
North-Centre and South-Islands, apart from Latina, in the
southern part of the Lazio region (the centre of Italy), which
was included in the South-Islands according to its epidemio-
logical pattern.
During the study period, the CR included in the study only
SPs active in the North-Centre. The number of SPs increased
particularly in 2006, when the actual extension of invitations
rose to 51% of the target population (subjects aged 50-69
years) compared to 16% in 2005.11 In the IMPATTO study,
the proportion of screen-detected cases in subjects aged 50-69
years in the North-Centre rose from 9.1% in 2005 to 30.3%
in 2006 and reached 45.9% in 2008. Therefore, two periods
were identified, pre-2005 and from 2006 onward. Period-spe-
cific indicators were reported for areas where SPs were present.
Incidence rate trends (standardized on the 2001 European
population) were evaluated as a whole and by macro-area
(North-Centre and South-Islands), sex, ten-year age class,
anatomic site (proximal colon: C18.0-C18.4; distal colon:
C18.5-C18.8; colon NOS: C18.9; and rectum: C19-C20),
stage at diagnosis (according to Dukes’ classification), and
pattern of diagnosis (screen-detected, non-screen-detected).
The annual percent change (APC) of incidence rate trends,
with 95% confidence intervals (95%CI), were computed.
RESULTS
We collected data on 47,830 CRCs, of which 973 were ex-
cluded (775 anus and anal canal, 129 lymphomas, sarcoma, or
melanoma, and 69 for other reasons).The study archives used
in this paper are therefore the 46,857 CRCs diagnosed between
2000 and 2008 in subjects aged 40-79 years.
About one-sixth of the cases (15.7%) were from the South and
the Islands (table 1, p. 118). Most cases were male (58%) and
in the upper age class (70-79 years, 44.8%).
There were 3,164 screen-detected cases (6.8% of the total; the
proportion increased to 16.6% when considering only cases of
50- to 69-year-olds from areas with an SP).
One third of the cases were in the rectum. The stage was
available for 87.6% of the cases.
Overall, the incidence rate was 133.7 and 83 per 100,000 in
males and females, respectively.
The incidence in the North-Centre was higher than in the
South-Islands: 141 vs 103.9 x 100,000 in men and 86.4 vs 69.4
x 100,000 in women.
As shown in table 2 (p. 118), the CRs took part in the study
with cases from different periods. Moreover, the SPs were in-
troduced in different years.
Standardized incidence rates of single CRs over the entire study
period were between 153.4 in Genova and 99.1 in Sassari in men
and between 94.8 in Genova and 65.6 in Sassari in women.
In the North-Centre, incidence rose more steeply from 2006
in both genders, the year that many SPs were implemented in
this macro-area (figure 1, p. 119). In the South and on the Is-
lands, the figure was stable for both genders.
Colorectal cancers incidence trends: IMPATTO study
117 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Colorectal cancers incidence trends: IMPATTO study
118 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Geographic Cancer registry Cases Incidence rate (x 100,000)
area (N) 2000 2001 2002 2003 2004 2005 2006 2007 2008
Northwest Genova 2,014 112.1 105.7 102.1
Milano 6,019 104.8 108.1 102.0 101.9 102.5 90.0 107.4
Sondrio 875 91.7 88.8 101.1 95.1 89.7 90.6 130.4 120.7 110.2
Biella 893 104.3 88.4 111.6 99.9 115.3 100.5 107.3
Northeast Trentino 1,215 91.5 92.2 88.1 90.0 95.8
Veneto 1,894 97.9 109.5 107.7 119.5 118.8 110.8
Friuli-Venezia Giulia 2,336 105.6 104.8 98.9
Emilia-Romagna 17,017 120.2 104.1 112.4 108.0 109.5 117.4 139.5 117.3 113.4
Centre Firenze-Prato 3,935 111.9 113.0 112.5 106.4 106.3 107.2
Umbria 3,289 111.7 115.6 111.6 142.3 128.2
South/Islands Latina 932 89.7 86.4 85.3 87.3
Napoli 945 83.3 77.2 98.2 84.8 90.5
Siracusa 821 81.1 79.7 83.9 83.6 79.1
Palermo 1,628 86.3 87.6 86.9
Catania-Messina 2,236 78.6 87.3 84.4
Sassari 808 87.9 86.8 82.6 79.1
Numbers in color represent the years when a screening programme was active
Table 2. Number of colorectal cancer cases and incidence rates (standardized Eu 2001) by cancer registry and year. Males and females aged 40-79 years.
Tabella 2. Casi di tumore del colon retto e tassi standardizzati di incidenza (popolazione europea 2001) per Registro tumori e anno. Uomini e donne, età 40-79 anni.
N %
Total 46,857 100
Macro-area
North-Centre 39,487 84.3
South-Islands 7,370 15.7
Gender
male 27,195 58.0
female 19,662 42.0
Age (years)
40-49 2,180 4.7
50-59 7,741 16.5
60-69 15,927 34.0
70-79 21,009 44.8
Pattern of diagnosis (all areas, age 40-79 years)
screen-detected at the first screening episode 2,897 6.2
screen-detected at a repeat screening episode 220 0.5
screen-detected at follow-up 47 0.1
not compliant with work-up after a positive screening test 116 0.2
subjects with a negative screening test before incidence 862 1.8
never compliant (i.e., invited but without a screening test) 2,102 4.5
never invited to screening 40,613 86.7
Pattern of diagnosis (areas with a screening programme, age 50-69 years)
screen-detected 2,805 16.6
non-screen-detected 14,061 83.4
Anatomic site
proximal colon 13,772 29.4
distal colon 16,278 34.7
rectum 14,278 30.5
colon NOS 2,529 5.4
Stage at diagnosis (Dukes)
I 8,218 17.5
II 12,051 25.7
III 12,206 26.0
IV 8,577 18.3
unknown 5,805 12.4
Table 1. Main characteristics
of the study subjects.
Tabella 1. Principali caratte-
ristiche dei soggetti studiati.
The increase observed in the North-Centre regarded only
those areas where SPs were implemented, with the APC in ar-
eas without SP being -0.7 (95%CI -4.3 to 3.1) for males and
-2.5 (95%CI -5.6 to 0.7) for females (table 3).
The trends in the South and on the Islands showed a non-sig-
nificant increase in males (APC 2.5; 95%CI -0.6 to 5.7) and
a decrease in females (APC -0.4; 95%CI -3.0 to 2.2).
In the North-Centre with an SP present, we recorded a non-
significant increase in males 50-69 years old (APC 3.6; 95%CI
-0.1 to 7.4) and in females 50-69 years old (APC 2.3; 95%CI
-0.8 to 5.4), while the 40-49 and 70-79 year age classes showed
small, non-significant decreases.
In the North-Centre without SPs no significant trends were ob-
served in the age class of 50-69 years, while in the South and
on the Islands incidence increased in males (APC 4.4; 95%CI
0.4 to 8.5) and overall (APC 3.3; 95%CI 1.3 to 5.3).
Figure 2 (p. 120) shows incidence rates by age in areas with an
SP, on a time scale centred on the year of implementation of
screening.The pre-screening incidence rates of the four 10-year
age classes were stable. During the first two years after screen-
ing started, incidence rates increased in all age groups, apart
from the youngest, and then decreased. The increase was
higher in subjects aged 60-69 years, whose incidence rates
shifted from 169 to 249 cases per 100,000 (+47.3%) as op-
posed to subjects 50-59 years old (+21.7%). The decrease in
incidence after year 2 was evident both in subjects aged 60-69
years (APC -5.7; 95%CI -28.3 to 24.2) and in those older than
70 years (APC -7.4; 95%CI -22.7 to 11.0).
In subjects aged 50-69 years, the pre-screening incidence rates
were similar to those of areas without SPs in the North-Centre,
and became significantly higher after the implementation of SPs
(table 4, p. 120). Incidence rates in the South and on the Islands
were lower. In particular, pre-screening incidence rates in the
North-Centre were generally comparable to North-Centre
Colorectal cancers incidence trends: IMPATTO study
119 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
Males Females Total
APC 95%CI APC 95%CI APC 95%CI
North-Centre with SP
40-79 years 2.5 -0.3 ; 5.4 1.3 -0.8 ; 3.5 2.2 -0.1 ; 4.5
50-69 years 3.6 -0.1 ; 7.4 2.3 -0.8 ; 5.4 3.1 -0.1 ; 6.5
North-Centre with SP
pre-screening
40-79 years 0.2 -1.7 ; 2.2 1.7 -1.9 ; 5.4 1.1 -0.1 ; 2.4
50-69 years 0.4 -2.3 ; 3.1 1.8 -1.9 ; 5.7 1.2 -0.7 ; 3.1
North-Centre with SP
post-screening
40-79 years 2.5 -0.3 ; 5.4 1.3 -0.8 ; 3.5 1.8 -0.8 ; 4.4
50-69 years 3.6 -0.1 ; 7.4 2.3 -0.8 ; 5.4 2.5 -1.1 ; 6.2
North-Centre without SP
40-79 years -0.7 -4.3 ; 3.1 -2.5 -5.6 ; 0.7 -1.3 -3.6 ; 1.1
50-69 years -2.0 -6.4 ; 2.6 -2.7 -9.1 ; 4.2 -2.0 -5.1 ; 1.2
South and the Islands
40-79 years 2.5 -0.6 ; 5.7 -0.4 -3.0 ; 2.2 1.3 -0.4 ; 3.0
50-69 years 4.4 0.4 ; 8.5 1.5 -2.2 ; 5.3 3.3 1.3 ; 5.3
Table 3. Annual percent
change (APC), with 95% con-
fidence intervals, of incidence
rates by macro-area, imple-
mentation of screening pro-
gramme (SP), age class, and
gender. Years 2000-2008.
Tabella 3. Annual percent
change (APC) dei tassi di inci-
denza (con intervalli di confi-
denza al 95%) per macroarea,
presenza di programmi di
screening (SP), classe d’età e
genere. Anni 2000-2008.
2000 2001 2002 2003 2004 2005 2006 2007 2008
x 100,000
20
40
60
80
100
120
140
160
180
0
Figure 1. Trends of incidence rates
(standardized Eu 2001) by macro-
area and gender. Ages 40-79 years.
Figura 1. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
macroarea e genere. Età 40-79
anni.
APC (95%CI)
NORTH-CENTRE
males (2000-2008) 2.1 (0.1 ; 4.1)
females (2000-2008) 1.3 (-0.2 ; 2.7)
SOUTH-ISLANDS
males (2001-2007) 2.5 (-0.6 ; 5.7)
females (2001-2007) -0.4 (-3.0 ; 2.2)
males North-Centre
females North-Centre
males South-Islands
females South-Islands
without screening for all categories of the variables studied, apart
from females (+7.5%) and those younger in age (+5.2%). In-
stead, the respective incidence rates were higher in the North-
Centre post-screening and lower in the South and on the Islands
for all variables, except for stage IV at diagnosis.
We could not compare the incidence rates by anatomic site of
the different areas, because the proportion of colon NOS in the
North-Centre without active SPs was too high and unevenly dis-
tributed during the years of the study.
Analysis by stage at diagnosis
In the North-Centre without active SPs, and in the South and
on the Islands, incidence rates by stage were stable (apart from
some fluctuations in the North-Centre during the early years
(figures 3 and 4). The APCs for both macro-areas were not
significant.
Instead, stage-specific incidence rates in the North-Centre
with active SPs showed two different phases (figure 5). Before
screening, the incidence rates of stage II, III, and IV cases were
stable while those of stage I increased. In fact, during the years
before screening the proportion of cases for which the stage was
not available decreased. We therefore carried out a sensitivity
analysis attributing to such cases the distribution by stage ob-
served among the cases whose stage was known. The pre-
screening incidence rates obtained in this way showed a smaller
but still significant increase for stage I cases (+5.7 x 100,000
over the entire period).
During the screening period, the incidence rates of stage I cases in-
Colorectal cancers incidence trends: IMPATTO study
120 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
-4 -3 -2 -1 0 1 2 3 4
x 100,000
50
100
150
200
250
300
350
0
Figure 2. Trends in incidence rates
by age class. Cancer registries of ar-
eas with screening programme
only. The time scale is centred on
the year the programme was im-
plemented.
Figura 2. Andamento temporale
dei tassi di incidenza per classe
d’età. Solo registri tumori di aree
dove è attivo un programma di
screening. La scala temporale è
centrata sull’anno in cui il pro-
gramma è stato attivato.
70-79 years
60-69 years
50-59 years
40-49 years
screening start
North-Centre North-Centre South-Islands
without with screening programme
screening pre-screening post-screening
programme
incidence incidence p-value* incidence p-value* incidence p-value*
rates rates rates rates rates
Overall 112.1 116.3 0.10 137.1 <0.001 93.7 <0.001
Gender
male 142.2 144.0 0.66 170.5 <0.001 112.3 <0.001
female 84.2 91.7 0.02 106.8 <0.001 76.9 <0.001
Age (years)
50-59 72.9 78.1 0.07 89.7 <0.001 62.2 <0.001
60-69 168.9 171.4 0.59 205.5 <0.001 139.1 <0.001
Pattern of diagnosis
screen-detected ---37.4 ---
non-screen-detected 112.1 116.3 0.10 99.7 <0.001 93.7 <0.001
Stage at diagnosis
(Dukes)
I 18.0 18.6 0.57 32.8 <0.001 11.7 <0.001
II 26.9 29.2 0.08 31.7 <0.001 22.5 <0.001
III 31.4 32.4 0.45 36.6 <0.001 21.4 <0.001
IV 23.2 23.1 0.89 22.0 0.28 21.1 0.18
unknown 12.6 13.1 0.64 13.9 0.14 16.9 <0.001
* compared to reference = North-Centre without screening programme
Table 4. Incidence rates
(standardized Eu 2001) by
macro-area, implementation
of screening programme and
period with respect to diffe-
rent characteristics, x 100,000.
Ages 50-69 years.
Tabella 4. Tassi standardiz-
zati di incidenza (popolazione
europea 2001) per macroa-
rea, con e senza screening,
per periodo, x 100.000. Età
50-69 anni.
x 100,000
20
40
60
80
100
120
140
160
Figure 5. Trends in incidence rates
(standardized Eu 2001) by stage
at diagnosis. Only areas with a
screening programme. Ages 50-69
years. The time scale is centred on
the year of implementation of the
screening programme.
Figura 5. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
stadio alla diagnosi. Solo aree con
programmi di screening attivi. Età
50-69 anni. La scala temporale è
centrata sull’anno in cui il pro-
gramma è stato attivato.
I II III IV unknown
Colorectal cancers incidence trends: IMPATTO study
121 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
2002 2003 2004 2005 2006 2007
x 100,000
20
40
60
80
100
120
140
160
0
Figure 3. Trends in incidence rates
(standardized Eu 2001) by stage
at diagnosis. North-Centre with no
screening programme. Ages 40-79
years.
Figura 3. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
stadio alla diagnosi. Centro-Nord
senza programmi di screening. Età
40-79 anni.
I II III IV unknown
2003 2004 2005 2006 2007
x 100,000
20
40
60
80
100
120
140
160
0
Figure 4. Trends in incidence rates
(standardized Eu 2001) by stage
at diagnosis. South-Islands. Ages
40-79 years.
Figura 4. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
stadio alla diagnosi. Sud-Isole. Età
40-79 anni.
I II III IV unknown
0
-4 -3 -2 -1 0 1 2 3 4
screening start
creased from 19.5 to 44.7 x 100,000 in the 2nd year, those of stage
II from 30.3 to 35.8 x 100,000, while stage III and IV cases were
quite stable, their sum ranging between 68 and 73 cases x 100,000.
In the North-Centre with no active SP during the final years
of the study, stage I cases increased by 7.5 percent points and
stage IV cases by 4.6 points, while stage III cases decreased by
7.4 points and the proportion of cases with unavailable stage
also declined (table 5).
In the South and on the Islands, no variation occurred during
the study period. A relevant proportion of cases were stage IV
(21.3%), while stage I cases were 12.8%, lower than in the
North-Centre. The proportion of cases with an unknown
stage was around 20%.
In areas with an SP, the proportion of stage I cases in subjects
aged 50-69 years increased from 16% before SP implementa-
tion to 26.7% after, while stage III and stage IV cases decreased
respectively by 2.0 and 4.2 percent points.
Analysis by anatomic site
The proportion of colon NOS in the North-Centre with no
active SP was too high to produce the incidence rates without
SP by site for the North-Centre.
In the South-Islands macro-area, incidence rates in the proximal
colon decreased, while those in the distal colon were stable and
those in the rectum increased (figure 6). Only the latter trend
was statistically significant (APC 3.0; 95%CI 0.3 to 5.7).
In areas where SPs were implemented, the pre-screening trend
for all sites was stable (figure 7). When the SPs started, we
recorded a steep increase of incidence rates in the distal colon
(from 43.9 to 69.3 x 100,000 in the 2nd year) and, to a lesser
extent, in the proximal colon (from 32.3 to 40.9 x 100,000)
and the rectum (from 36.7 to 45.5 x 100,000). This increase
ended two years after the implementation of screening and was
followed by a reduction in the rates for all three sites.
DISCUSSION
We evaluated CRC incidence rates in Italy from the early
2000s, with particular regard to the effects of the implemen-
tation of the SPs introduced during that period in several ar-
eas of the country.
Overall, we observed a remarkable difference between the
North-Centre and the South and Islands, with the incidence
rates in the former macro-area being much higher than in the
latter. A different risk of CRC throughout the country, mainly
attributed to different exposure to risk factors (e.g., diet), had
already been reported.28
Of the areas included in the study, SPs had been implemented
only in the North-Centre and showed a significant effect on in-
cidence rates. As expected, a sharp increase in incidence was ob-
served in the first years of screening, the prevalence round,29
followed by a decrease that started quite soon, i.e., within 3-4
years of screening start. For subjects aged 70-79 years, the in-
Colorectal cancers incidence trends: IMPATTO study
122 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
2000 2001 2002 2003 2004 2005 2006 2007 2008
x 100,000
5
10
15
20
25
30
35
40
0
Figure 6. Trends in incidence
rates (standardized Eu 2001) by
anatomic site. Sout h-Islands. Ages
40-79 years.
Figura 6. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
sede anatomica. Sud-Isole. Età 40-
79 anni.
colon NOS
proximal colon
distal colon
rectum
North-Centre without screening South/Islands North-Centre with screening
programme 40-79 years 40-79 years programme 40-69 years
Stage 2000-2006 2006-2008 2001-2007 pre-screening post-screening
(N=4,488) (N=1,805) (N=7,370) (N=6,713) (N=8,186)
I 13.7 21.2 12.8 16.0 26.7
II 25.6 25.5 23.9 25.3 23.2
III 29.1 21.7 23.1 27.8 25.8
IV 18.1 22.7 21.3 19.6 15.4
unknown 13.4 8.9 18.8 11.3 8.8
Table 5. Distribution by stage
at diagnosis, by macro-area
with and without a screening
programme,and by period (%).
Tabella 5. Distribuzione per
stadio alla diagnosi, per ma-
croarea, con e senza un pro-
gramma di screnning, per pe-
riodo (%).
cidence trend after the prevalence round is suggestive of a de-
crease to values lower than the pre-screening level.
The overall increase is evident even when considering the
whole age range included in the study (40-79 years), which
exceeds the specific target population of screening, as well as
in the national statistics regarding all ages (0-85+ years).2It
is of utmost importance that such trends be correctly inter-
preted in terms of any transient effect related to the imple-
mentation of screening and not as an increased risk of CRC
in the population.
SPs increased the incidence gap between macro-areas: in the
South and on the Islands, no significant trend was observed,
in either gender, nor in the pre-screening in those areas where
screening was implemented. After the introduction of SPs, the
increase in incidence was more evident in the 60-69 year age
class than in the 50-59 year one. A differential effect of FIT
with age has been described.30 We also observed a small, non-
significant but consistent increase in both genders and across
centres, in the age classes above 70 years (probably related to
a significant proportion of screen-detected cases in that age
group: 15.1% in the 70-74 year class – the SP of Umbria is
aimed at residents aged from 50 to 74 years). We do not have
enough power to observe even considerably strong trends in the
youngest age class because incidence is quite low. Nonetheless,
our data suggest a decreasing trend.
These figures are highly suggestive of the expected increase in
incidence rates that the introduction of SPs produces through
the anticipated diagnosis of cases that otherwise would emerge
later and, in part, through a (hard to quantify) number of
over-diagnoses.
One relevant aspect analyzed regards the impact of screening
on incidence trends by stage at diagnosis. In the areas where an
SP was implemented, we recorded a pre-screening trend only
for stage I cases. This could be related to a spontaneous (i.e.,
in the absence of a population-based SP) increased spread of
colonoscopies in the population. The implementation of SPs
modified exclusively the rates of stage I cases, with the “clas-
sic” pattern of initial increase and subsequent reduction in in-
cidence. None of the other stages were affected by screening.
This suggests that diagnostic anticipation takes place mainly for
cases at an initial stage.
Our data do not allow us to assess the issue of over-diagnosis,
mainly because the follow-up period for SPs is too short to de-
termine whether the decrease in incidence observed beginning
in the 3rd year will reach the level of pre-screening incidence
or drop even lower. However, it has been argued that over-di-
agnosis of invasive CRC is not a worrisome phenomenon in
CRC screening, because the removal of precancerous lesions
(i.e., advanced adenomas) determines a relevant incidence
reduction.31-34
Screening is expected to reduce the incidence rates of advanced
stages. We did not notice such an effect, probably because the
slow implementation of SPs is still delaying the end of the preva-
lence round. In fact, only a few programmes have invited the
entire target population within the first two years, and all Ital-
ian programmes have seen quite low participation rates. Con-
sequently, the proportion of first screening tests is very high even
3 or 4 years after programme start, due to people being invited
for the first time or those who did not respond to the first in-
vitation and decided to respond to a second one. Only a longer
follow-up period and a more detailed analysis of the cohorts ac-
tually invited or participating will make it possible to confirm
any effect of screening on the incidence of advanced cancers and
incidence as a whole. It is worth underlining that none of the
studies evaluating the impact of colorectal screening on inci-
dence rates have found a cumulative reduction of incidence
within 5 years of starting to screen,16,20 including those based
on flexible sigmoidoscopy.35-36
Differently from areas with SPs, both the North-Centre with-
out SPs and the South and Islands did not record any signifi-
Colorectal cancers incidence trends: IMPATTO study
123 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
-4 -3 -2 -1 0 1 2 3 4
x 100,000
10
20
30
40
50
60
70
80
0
Figure 7. Trends in incidence rates
(standardized Eu 2001) by anatomic
site. Cancer registries only of areas
with a screening programme.Ages
50-69 years. The time scale is cen-
tred on the year of implementation
of the screening programme.
Figura 7. Andamento temporale
dei tassi standardizzati di incidenza
(popolazione europea 2001) per
sede anatomica. Solo registri tu-
mori di aree dove è attivo un pro-
gramma di screening. Età 50-69
anni. La scala temporale è centrata
sull’anno in cui il programma è
stato attivato.
rectum
distal colon
proximal colon
colon NOS
screening start
cant trend at any stage of diagnosis. However, notwithstand-
ing the lower overall incidence rates, the specific rates of stage
IV in the South and on the Islands were comparable to the
other areas of the country. This figure may be attributed to a
delay of diagnosis in this macro-area. The lack of any decrease
during the study period suggests that no improvements took
place to enhance the anticipation of CRC diagnosis.Thus, the
widespread implementation of SPs in this macro-area seems
particularly relevant.
Unfortunately, we could not evaluate incidence trends by
anatomic site in the northern-central areas without SPs, due to
a high percentage of missing data. In the South and on the Is-
lands, incidence trends by site showed a significant increase for
the rectum.
In areas with SPs, the pre-screening rates in the proximal
colon, distal colon, and rectum were stable. After SP started,
incidence increased in all anatomic sites, particularly in the dis-
tal colon. This figure is in line with the results of many stud-
ies that have shown a higher sensitivity for advanced neopla-
sia in the left versus right colon with faecal occult blood
testing37 and colonoscopy.38
The major strength of this study is the large number of cases
included in the analysis and the quality of the data collected.
The study is based on almost 47,000 CRCs collected by a large
number of cancer registries throughout the entire country,
and thus offers the best available representation of CRC epi-
demiology in Italy in relation to SP implementation. On the
other hand, the areas included in this study represent a relevant
proportion (27% overall) of the national population, but the
various macro-areas are unevenly represented. Therefore, pro-
jecting our results to the whole country should be done with
caution.
This study also has several limits. First, the results of this
study do not exclusively reflect the performance of the screen-
ing protocols utilised by SPs (first level test and further assess-
ment), but were very much influenced by the spread of screen-
ing in the target population, a result associated with the
effective extension of invitations and compliance with the in-
vitation to a first-level test, as well as diagnostic workup for sub-
jects with a positive test, etc. These figures are quite different
among programmes and make generalizations difficult. This
implies that our results should be regarded as purely indicative
of what can be expected when implementing an SP, but the fig-
ures obtained in a different setting may be very different.
Second, the study only included the few years since screening
started. Therefore it could not show how long the decrease in
incidence rates, following the initial peak, might last and the
size of the reduction that could be achieved.
CONCLUSION
We described the trends of CRC incidence rates in Italy from
2000 to 2008, when several SPs were implemented in differ-
ent areas. The differential figures introduced by the imple-
mentation of SPs warrant a continuous surveillance of CRC in-
cidence and mortality trends to monitor the impact of
screening at a national level.
Conflicts of interests: none declared
Colorectal cancers incidence trends: IMPATTO study
124 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
“Paolo Giaccone” di Palermo, Epidemiologia clinica con Registro
tumori, Palermo)
Parma: P. Sgarzi, F. Bozzani (Registro tumori di Parma); M. Zatelli,
C. Zurlini (Centro screening oncologici AUSL Parma); P. Caruana
(AOU Parma)
Piacenza: E. Borciani, G. Gatti, R. Prazzoli, P. Seghini
(UO Epidemiologia e comunicazione del rischio, AUSL Piacenza)
Reggio Emilia: C. Campari (Centro screening, AUSL Reggio Emilia);
E. Di Felice, C. Sacchettini, T. Cassetti (Registro tumori di Reggio Emilia)
Romagna: F. Falcini, S. Mancini, R. Vattiato, O. Giuliani (Registro
tumori della Romagna, Istituto dei tumori della Romagna IRCCS,
Meldola); L. Caprara (Anatomia patoligica OC S.M. della Scaletta,
Imola)
Sassari: R. Cesaraccio, O. Sechi (Registro tumori di Sassari)
Siracusa: A. Madeddu, M.L. Contrino, A. Colanino Ziino, M. Russo
(Registro tumori di Siracusa)
Sondrio: A.C. Fanetti, S. Maspero, E. Moroni, I. Cometti (Osservatorio
epidemiologico - Registro tumori della Provincia di Sondrio)
Trentino: M. Gentilini (Registro tumori di Trento); G. De Pretis
(UO Gastroenterologia, Ospedale S. Chiara, Trento); P. Caciagli
(Dipartimento laboratorio e servizi, APSS, Trento); R. Pertile (Servizio
epidemiologia clinica e valutativa, APSS, Trento)
Umbria: F. Bianconi, G.M. Masanotti (Registro tumori umbro
di popolazione / Scuola di specializzazione in Igiene e medicina
preventiva, Università di Perugia); M.C. De Lio (Scuola di
specializzazione in Igiene e medicina preventiva, Università
di Perugia); S. Leite (Registro tumori umbro di popolazione);
B.U. Passamonti (Servizio di screening USL 1 Umbria)
Veneto: A.P. Dei Tos, A. Rosano, D. Monetti, S.Guzzinati,
S. Tognazzo (Registro tumori del Veneto, Padova)
Biella: A. Azzoni (S.S. Gastroenterologia, ASL Biella)
Bologna: P. Baldazzi, N. Collina, P. Pandolfi, C. Petrucci, G. Gualandi
(Registro tumori di Bologna)
Catania-Messina: S. Sciacchitano, A. Ieni, F. Bella, A. Torrisi,
M. Varvarà (Registro tumori integrato di Catania-Messina, Catania)
Emilia-Romagna: A.C. Finarelli, P. Sassoli de’ BIanchi, P. Landi
(Assessorato alle politiche per la salute, Regione Emilia-Romagna,
Bologna)
Ferrara: S.Ferretti (Rewgistro tumori di Ferrara); V. Matarese
(UO Gastroenterologia, AOU S. Anna, Cona-Ferrara, Italy);
A. De Togni, C. Palmonari (Centro screening oncologici AUSL Ferrara)
Firenze: A. Caldarella, E. Crocetti, G. Grazzini, G. Manneschi,
P. Mantellini (Istituto per lo studio e la prevenzione oncologica – ISPO,
Firenze)
Friuli-Venezia Giulia: D. Serraino, E. Bidoli, M. Taborelli, A. Gini,
S. Virdone (Istituto nazionale dei tumori – CRO, Aviano, PN)
Genova: C. Casella, M. Celesia, R. Cogno, E. Marani (Registro
tumori Regione Liguria, UO Epidemiologia clinica, IRCCS AOU
San Martino–IST, Genova)
Latina: S. Busco, E. Bernazza, S. Curatella, L. Macci, M. Rossi
(Registro tumori di Latina, Latina)
Milano: G. Randi, B. Frammartino, A. Bonini, L. Filippazzi, C. Giubelli
(Registro tumori di Milano, Milano)
Modena: R. Corradini, F. De Girolamo (Centro screening oncologici
AUSL di Modena, Italy); K. Valla (Registro tumori di Modena, Modena)
Napoli: R. Palombino, M. D’Orsi, M. Isernia (Servizio epidemiologia
e prevenzione, ASL Napoli 3 Sud); M. Fusco, M.F. Vitale (Registro
tumori di popolazione Regione Campania, Napoli)
Palermo: M.S. Adamo, R. Amodio, A. Brucculeri, A. Guttadauro,
W. Mazzucco (Registro tumori di Palermo e Provincia, AOU Policlinico
Members of the IMPATTO COLONRETTO working group:
Membri del gruppo di studio IMPATTO COLONRETTO:
Colorectal cancers incidence trends: IMPATTO study
125 NATIONAL CENTRE SCREENING MONITORING 11TH REPORT
Epidemiol Prev 2015; 39(3) Suppl 1: 1-125
anno 39 (3) maggio-giugno 2015
e
&
p
References/Bibliografia
1. AIOM-AIRTUM. I numeri del cancro in Italia 2014. Brescia, Inter-
media Editore 2014.
2. ITACAN: Tumori in Italia, Versione 2.0. AIRTUM. Available at:
http://www.registri-tumori.it [Accessed November 24, 2014].
3. Baracco M, Barizza F, Bovo E et al. Fatti e cifre dei tumori in Veneto.
Colon retto. Padova 2007. Available at: http://www.registrotu-
moriveneto.it/registro/analisi/file/FATTI_E_CIFRE_DEI_TUMORI_colon
_retto_definitivo.pdf [Accessed November 24, 2014].
4. Mangone L et al. I tumori in provincia di Reggio Emilia nel periodo
1996-2003. Collana Progetto Salute 9, AUSL Reggio Emilia.
5. Ponz de Leon M, Benatti P et al. Epidemiologia dei Tumori del
colon-retto. Incidenza, mortalità, familiarità e sopravvivenza nella
ex USL di Modena, 1984-1998. Università di Modena, Modena
2001.
6. Ward KC, Young JL , Ries LAG. Cancers of the colon and rectum.
In: Ries LAG, Young JL, Keel GE et al eds. SEER survival monograph:
cancer survival amongst adults: U.S. SEER Program, 1998-2001,
patient and tumor characteristics. NIH Pub. No. 07-6215.
Bethesda, Md, National Institutes of Health 2007:33-42.
7. Ries LAG, Wingo PA, Miller DS et al. The annual report to the na-
tion on the status of cancer, 1973–1997, with a special section on
colorectal cancer. Cancer 2000;88:2398-424.
8. AIRTUM Working Group. Italian cancer figures – Report 2011. Sur-
vival of cancer patients in Italy. Epidemiol Prev 2011;5-6(Suppl
3):85.
9. Zorzi M, Fedato C, Grazzini G et al. Screening for colorectal cancer
in Italy, 2010 survey. Epidemiol Prev 2012 ;6 (Suppl 1):55-77.
10. Zorzi M, Baracco S, Fedato C et al. Screening for colorectal cancer in
Italy, 2009 survey. Epidemiol Prev 2011;5-6 (Suppl 5):55-77.
11. Zorzi M, Baracco S, Fedato C et al. Screening for colorectal cancer in
Italy: 2006 survey. Epidemiol Prev 2008;2 (Suppl 1):55-68.
12. Hewitson P, Glasziou P et al. Cochrane systematic review of colorectal
cancer screening using the fecal occult blood test (hemoccult): an up-
date. Am J Gastroenterol 2008;103(6):1541-49.
13. Costantini AS, Martini A, Puliti D et al. Colorectal cancer mortal-
ity in two areas of Tuscany with different screening exposures. J
Natl Cancer Inst 2008;100(24):1818-21.
14. Zorzi M, Fedeli U, Schievano E et al. Impact on colorectal cancer
mortality of screening programmes based on the faecal immuno-
chemical test. Gut 2014 Sep 1. pii: gutjnl-2014-307508. doi:
10.1136/gutjnl-2014-307508.
15. Mandel JS, Bond JH, Church TR et al. Reducing mortality from col-
orectal cancer by screening for fecal occult blood. Minnesota
Colon Cancer Control Study. N Engl J Med 1993;328:1365-71.
16. Mandel JS, Church TR, Bond JH et al. The effect of fecal occult-
blood screening on the incidence of colorectal cancer. N Engl J Med
2000;343:1603-607.
17. Kronborg O, Jorgensen OD, Fenger C et al. Randomized study of
biennial screening with a faecal occult blood test: Results after nine
screening rounds. Scand J Gastroenterol 2004;39:846-51.
18. Kewenter J, Brevinge H, Engaras B et al. Results of screening, re-
screening, and follow-up in a prospective, randomized study for
the detection of colorectal cancer by fecal occult blood testing. Re-
sults for 68308 subjects. Scand J Gastroenterol 1994;29:468-73.
19. Hardcastle JD, Chamberlain JO, Robinson MH et al. Randomised
controlled trial of faecal-occult-blood screening for colorectal can-
cer. Lancet 1996;348:1472-77.
20. Ventura L, Mantellini P, Grazzini G et al. The impact of immuno-
chemical faecal occult blood testing on colorectal cancer incidence.
Dig Liver Dis 2013 Sep 3. doi:pii: S1590-8658(13)00280-6.
10.1016/j.dld.2013.07.017.
21. Allison JE, Tekawa IS, Ransom LJ et al. A comparison of fecal occult
-blood tests for colorectal cancer screening. N Engl J Med 1996;
334(3):155-59.
22. Allison JE, Sakoda LC, Levin TR et al. Screening for colorectal neo-
plasms with new fecal occult blood tests: update on performance
characteristics. J Natl Cancer Inst 2007;99(19):1462-70.
23. Guittet L, Bouvier V, Mariotte N et al. Comparison of a guaiac
based and an immunochemical faecal occult blood test in screen-
ing for colorectal cancer in a general average risk population. Gut
2007;56(2):210-14.
24. Hol L, van Leerdam ME, van Ballegooijen M et al. Screening for col-
orectal cancer: randomised trial comparing guaiac-based and im-
munochemical faecal occult blood testing and flexible sigmoi-
doscopy. Gut 2010;59(1):62-68.
25. Hol L, Wilschut JA, van BM et al. Screening for colorectal cancer: ran-
dom comparison of guaiac and immunochemical faecal occult blood
testing at different cut-off levels. Br J Cancer 2009;100(7):1103-10.
26. van Rossum LG, van Rijn AF, Laheij RJ et al. Random comparison
of guaiac and immunochemical fecal occult blood tests for col-
orectal cancer in a screening population. Gastroenterology
2008;135(1):82-90.
27. Kronborg O, Fenger C, Olsen J et al. Randomised study of screen-
ing for colorectal cancer with faecal-occult-blood test. Lancet
1996;348(9040):1467-71.
28. AIRTUM Working Group. I trend dei tumori in Italia negli anni
duemila (1998-2005). Cancer trend (1998-2005). Epidemiol Prev
2009;4-5(Suppl. 1):1-165
29. Morrison AS. Screening in Chronic Disease, 2nd ed. New York,
Oxford University Press 1992.
30. Wong MC, Ching JY, Chan VC et al. Factors associated with false-
positive and false-negative fecal immunochemical test results for
colorectal cancer screening. Gastrointest Endosc 2014 Oct 4. pii:
S0016-5107(14)02075-6. doi: 10.1016/j.gie.2014.08.006.
31. Brenner H, Altenhofen L, Stock C, Hoffmeister M. Prevention, early
detection, and overdiagnosis of colorectal cancer within 10 years
of screening colonoscopy in Germany. Clin Gastroenterol Hepatol
2015;13(4):717-23. pii: S1542-3565(14)01313-5. doi: 10.1016/
j.cgh.2014.08.036. [Epub 2014 Sept 15]
32. Esserman LJ, Thompson IM Jr, Reid B. Overdiagnosis and overtreat-
ment in cancer: an opportunity for improvement. JAMA 2018;
310(8):797-98.
33. Holme Ø, Bretthauer M, Fretheim A et al. Flexible sigmoidoscopy
versus faecal occult blood testing for colorectal cancer screening
in asymptomatic individuals. Cochrane Database Syst Rev 2013 Oct
1;9:CD009259. doi: 10.1002/14651858.CD009259.pub2.
34. Yang DX1, Gross CP, Soulos PR, Yu JB. Estimating the magnitude
of colorectal cancers prevented during the era of screening: 1976
to 2009. Cancer 2014;120(18):2893-901. doi:10.1002/cncr.28794
35. Atkin WS, Edwards R, Kralj-Hans I et al. Once-only flexible
sigmoidoscopy screening in prevention of colorectal cancer: a
multicentre randomised controlled trial. Lancet 2010;375:1624-33.
36. Segnan N, Armaroli P, Bonelli L et al. Once-only sigmoidoscopy in
colorectal cancer screening: follow-up findings of the Italian Ran-
domized Controlled Trial–SCORE. J Natl Cancer Inst 2011;103:
1310-22.
37. Haug U, Knudsen AB, Brenner H, Kuntz KM. Is fecal occult blood
testing more sensitive for left- versus right-sided colorectal neo-
plasia? A systematic literature review. Expert Rev Mol Diagn
2011;11(6):605-16.
38. Singh S, Singh PP, Murad MH et al. Prevalence, risk factors, and
outcomes of interval colorectal cancers: a systematic review and
meta-analysis. Am J Gastroenterol 2014 Sep;109(9):1375-89. doi:
10.1038/ajg.2014.171. Epub 2014 Jun 24. Review. PubMed PMID:
24957158.
39. Astler VB, Coller FA. The prognostic significance of direct exten-
sion of carcinoma of the colon and rectum. Ann Surg 2003;139(6):
846-52.
A CIASCUNO IL SUO
ABBONAMENTI 2015
E&P on-line + E&P on-line + E&P on-line +
Suppl on-line E&P di carta + E&P di carta +
Suppl on-line supplementi di carta
PRIVATI ITALIA
1 anno 72 euro 80 euro 95 euro
2 anni 135 euro 150 euro 180 euro
3 anni 190 euro 210 euro 250 euro
ENTI ITALIA AD ACCESSO UNICO ENTI ITALIA AD ACCESSO MULTIPLO: ABBONAMENTI DA CONCORDARE CON L’EDITORE
1 anno 148 euro 155 euro 170 euro
2 anni 275 euro 290 euro 320 euro
3 anni 390 euro 410 euro 450 euro
ENTI ESTERO
1 anno 165 euro 180 euro 210 euro
2 anni 310 euro 335 euro 395 euro
3 anni 425 euro 475 euro 555 euro
PRIVATI ESTERO
1 anno 85 euro 100 euro 130 euro
2 anni 160 euro 190 euro 245 euro
3 anni 225 euro 265 euro 350 euro
PROMOZIONI 2015
■Per giovani epidemiologi: abbonamento on line a 45 euro per gli under 30.
■Per generosi epidemiologi già abbonati a E&P: regala un abbonamento a E&P per il 2015.
Costa solo 50 euro per l’edizione on line e 60 euro per avere anche il cartaceo.
Ovviamente, l’abbonamento sarà accompagnato da un biglietto che svelerà l’identità del donatore
per fare una gran bella figura e nello stesso tempo aiutare E&P.
■Per epidemiologi “contagiosi”: se ti piace E&P e fai sottoscrivere due nuovi abbonamenti a chi
non conosce la rivista o non è più abbonato da almeno due anni, il tuo abbonamento
o il tuo rinnovo è gratuito.
EPIDEMIOLOGIA & PREVENZIONE
Modalità di abbonamento per il 2015
data ................................................. Abbonamento annuo a partire dal primo numero raggiungibile:
❐Tipo di abbonamento .............................................................. ❐euro ..............................................................
Modalità di pagamento:
❐Versamento: a mezzo conto corrente postale n. 55195440 ❐PayPal: sul sito www.epiprev.it
intestato a Inferenze scarl, via Ricciarelli 29, 20148 Milano
(allegare la ricevuta di versamento alla richiesta di abbonamento) ❐Bonifico bancario: UNIPOL BANCA, P.za Wagner 8, 20145 Milano
IBAN IT 53 P 03127 01600 0000 0000 3681 intestato a Inferenze scarl,
via Ricciarelli 29, 20148 Milano (allegare la contabile alla richiesta di abbonamento)
❐Carta di credito: ❐American Express ❐Carta Sì ❐Master Card ❐Eurocard ❐VISA
cognome e nome .............................................................................................................................................................................................................................................
azienda ...........................................................................................................................................................................................................................................................
indirizzo ...........................................................................................................................................................................................................................................................
cap ....................................................... località ............................................................................................................................................................. prov. .....................
tel. ................................................................... fax. ................................................................... e-mail ........................................................................................................
numero _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ scadenza _ _ / _ _ / _ _ firma .......................................................................................................
cod. CV2 _ _ _ (ultime tre cifre stampate sul retro della carta, per una garanzia di sicurezza in più)
Compilare e inviare a Inferenze - via Ricciarelli 29, 20148 Milano; e-mail abbonamenti@inferenze.it o per fax allo 02 48706089
EPIDEMIOLOGIA
& PREVENZIONE
