Content uploaded by Otavio T Ranzani
Author content
All content in this area was uploaded by Otavio T Ranzani on Aug 25, 2015
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
L E T T E R Open Access
Importance of a registered and structured
protocol when conducting systematic
reviews: comments about nebulized
antibiotics for ventilator-associated
pneumonia
Fernando G. Zampieri
1,2,3
, Antonio P. Nassar Jr
1,2,4
, Dimitri Gusmao-Flores
1,5
, Leandro U. Taniguchi
2,6
,AntoniTorres
7
and Otavio T. Ranzani
1,7,8,9*
See related Letter by Gu, http://www.ccforum.com/content/19/1/236, and related research by Zampieri et al., http://ccforum.com/content/19/1/150
We appreciate Gu’s [1] interest in our study. We
apologize and agree with his comment about attribut-
ing units to standardized mean difference (SMD).
Nevertheless, similar to the SMD, results in mean dif-
ference (control –nebulized) were unaffected by neb-
ulized antibiotics (2.67 days, 95 % confidence interval
(CI) –2.89, 8.23 for ICU length of stay (LOS); and
0.70 days, 95 % CI −3.40, 4.80 for mechanical ventilation).
However, we strongly disagree with other points raised by
the letter.
First, the study protocol was defined a priori [2]. We
disagree that combining observational studies with inter-
vention studies is reserved only for safety evaluation.
This topic has been discussed in the literature and com-
bining both types of studies was adequate for our aim
[3]. Furthermore, we presented the main results separat-
ing interventional studies from observational studies,
thereby allowing the reader to interpret both analyses
independently.
Second, both of the studies cited as “case–control
studies”[1] received this denomination in their title
and abstract. However, by reading their methods it
becomes clear that they are actually matched cohort
studies [4, 5]. Indeed, they matched exposed patients
(“nebulized group”) to unexposed patients (“no-nebu-
lized group”). A case–control design starts with the
outcome (case = “clinical success”) and matches them
with controls (“clinical failures”). Therefore, our measure
of effect was correct [5]. For exploration, we report the
analysis for clinical cure using the odds ratio (OR) (Fig. 1).
The results are unchanged.
Third, Kalin’sstudywasincludedbecauseitful-
filled our inclusion/exclusion criteria [2]. Gu’ssug-
gestion to exclude this study based solely on its
effects in heterogeneity could be considered selective
reporting [1].
Our study provided data for further trials aiming to
evaluate the effect of nebulized antibiotics in ventilator-
associated pneumonia (VAP) [2].
* Correspondence: otavioranzani@yahoo.com.br
1
Cooperative Network for Research—AMIB-Net, Associação de Medicina
Intensiva Brasileira, Rua Arminda, 93, 7 andar, São Paulo 04545-100, Brazil
7
Department of Pulmonology, Hospital Clinic of Barcelona, Institut
D’investigacions August Pi I Sunyer (IDIBAPS), University of Barcelona, Ciber
de Enfermedades Respiratorias (CIBERES), Carrer Villarroel, 170, Barcelona
08036, Spain
Full list of author information is available at the end of the article
© 2015 Zampieri et al.
Open Access
This article is distributed under the terms of the Creative Commons Attribution
4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Zampieri et al. Critical Care (2015) 19:298
DOI 10.1186/s13054-015-1020-8
Abbreviations
CI: Confidence interval; LOS: Length of stay; OR: Odds ratio;
SMD: Standardized mean difference; VAP: Ventilator-associated pneumonia.
Competing interests
The authors declare that they have no competing interests.
Author contributions
FGZ conceived the study, participated in data acquisition, data analysis,
and interpretation, and helped to revise the manuscript for important
intellectual content. APN participated in conception of the study, data
acquisition, data analysis, and interpretation, and helped to revise the
manuscript for important intellectual content. DG-F participated in conception
of the study and interpretation, and helped to revise the manuscript for
important intellectual content. LUT participated in conception and design of
the study and interpretation, and helped to revise the manuscript for important
intellectual content. AT participated in conception and design of the study and
interpretation, and helped to revise the manuscript for important intellectual
content. OTR conceived the design of the study, participated in data acquisition
and interpretation, and wrote the draft of the manuscript. All authors read and
approved the final manuscript.
Author details
1
Cooperative Network for Research—AMIB-Net, Associação de Medicina
Intensiva Brasileira, Rua Arminda, 93, 7 andar, São Paulo 04545-100, Brazil.
2
Emergency Medicine Discipline, Faculty of Medicine, University of São Paulo,
Rua Dr. Enéas de Carvalho Aguiar, 255, 5th floor, room 5023, São Paulo
05403-010, Brazil.
3
Intensive Care Unit, Hospital Alemão Oswaldo Cruz, Rua
João Julião, 331, São Paulo 01323-903, Brazil.
4
Adult Intensive Care Unit, A.C.
Camargo Cancer Center, Rua Professor Antônio Prudente, 211, São Paulo
01509-010, Brazil.
5
Intensive Care Unit, University Hospital Prof. Edgar Santos,
Universidade Federal da Bahia, Rua Augusto Viana, Salvador 40110-910, Brazil.
6
Research and Education Institute (IEP), Hospital Sirio-Libanes, Rua Prof.
Daher Cutait, 69, São Paulo 01308-060, Brazil.
7
Department of Pulmonology,
Hospital Clinic of Barcelona, Institut D’investigacions August Pi I Sunyer
(IDIBAPS), University of Barcelona, Ciber de Enfermedades Respiratorias
(CIBERES), Carrer Villarroel, 170, Barcelona 08036, Spain.
8
Amil Critical Care
Group, Hospital Paulistano, Rua Martiniano de Carvalho, 741, São Paulo
01321-001, Brazil.
9
Respiratory Intensive Care Unit, Pulmonary Division, Heart
Institute, Hospital das Clínicas, University of São Paulo, Av. Dr. Arnaldo, 455
Laboratório de Pneumologia, 2° andar, sala 2144, Cerqueira César, 01246903
Sao Paulo, Brazil.
References
1. Gu WJ. Nebulized antibiotics for ventilator-associated pneumonia:
misleading analysis and interpretation of the data. Crit Care.
2015;19:236.
2. Zampieri FG, Nassar AP, Gusmao-Flores D, Taniguchi LU, Torres A,
Ranzani OT. Nebulized antibiotics for ventilator-associated pneumonia: a
systematic review and meta-analysis. Crit Care. 2015;19:150.
3. Shrier I, Boivin JF, Steele RJ, Platt RW, Furlan A, Kakuma R, et al. Should
meta-analyses of interventions include observational studies in addition to
randomized controlled trials? A critical examination of underlying principles.
Am J Epidemiol. 2007;166:1203–9.
4. Cummings P, McKnight B, Greenland S. Matched cohort methods for injury
research. Epidemiol Rev. 2003;25:43–50.
5. Rothman KJ, Greenland S, Lash TL. Modern epidemiology. 3rd ed.
Philadelphia, PA: Lippincott Williams & Wilkins; 2008.
Study
Random effects model
Heterogeneity: I−squared=34%, tau−squared=0.145, p=0.1267
Type = Observational
Type = Randomized
Random effects model
Random effects model
Heterogeneity: I−squared=53.3%, tau−squared=0.2501, p=0.0576
Heterogeneity: I−squared=0%, tau−squared=0, p=0.5424
Doshi
Ghannam
Kalin
Kofteridis
Korbila
Tumbarello
Hallal
Le Conte
Lu
Niederman
Rattanaupawan
Events
24
13
4
23
62
72
5
7
14
27
26
Total
440
311
129
44
13
29
43
78
104
5
21
20
32
51
Nebulized Antibiotics
Events
20
5
6
14
26
57
3
3
11
14
26
Total
372
265
107
51
9
15
43
43
104
5
17
20
16
49
Control
0.01 0.1 1 10 100
Odds Ratio
Favours Control Favours Nebulized Antibiotics
OR
1.66
1.81
1.26
1.86
22.09
0.24
2.38
2.53
1.86
7.86
2.33
1.91
0.77
0.92
95%−CI
[1.10; 2.49]
[1.01; 3.22]
[0.71; 2.24]
[0.82; 4.21]
[1.01; 483.26]
[0.05; 1.05]
[0.99; 5.72]
[1.11; 5.76]
[1.05; 3.27]
[0.28; 217.11]
[0.50; 10.91]
[0.52; 7.01]
[0.13; 4.49]
[0.42; 2.02]
W(random)
100%
66.6%
33.4%
13.6%
1.7%
6.1%
12.6%
13.6%
19.0%
1.4%
5.7%
7.4%
4.6%
14.2%
Fig. 1 Forest plot for clinical cure using odds ratios (OR). Pfor overall effect = 0.015. CI confidence interval
Zampieri et al. Critical Care (2015) 19:298 Page 2 of 2
