Content uploaded by Jean-François Catanzariti
Author content
All content in this area was uploaded by Jean-François Catanzariti on May 04, 2022
Content may be subject to copyright.
Accelerat ing the world's research.
French translation and validation of 3
functional disability scales for neck
pain
Jean-François Catanzariti
Archives of Physical Medicine and Rehabilitation
Cite this paper
Get the citation in MLA, APA, or Chicago styles
Downloaded fromAcademia.edu
Related papers
The ability t o change of t hree questionnaires for neck pain
Jean-François Cat anz arit i
Sensibilité au changement de trois quest ionnaires d’évaluation des cervicalgies
Jean-François Cat anz arit i
Measurement properties of translat ed versions of neck- specific questionnaires: a syst emat ic review
Bart Koes
Download a PDF Pack of the best related papers
French Translation and Validation of 3 Functional
Disability Scales for Neck Pain
Samantha Wlodyka-Demaille, MD, Serge Poiraudeau, MD, PhD, Jean-Franc¸ois Catanzariti, MD,
Franc¸ois Rannou, MD, Jacques Fermanian, MD, PhD, Michel Revel, MD
ABSTRACT. Wlodyka-Demaille S, Poiraudeau S, Catanzariti
J-F, Rannou F, Fermanian J, Revel M. French translation and
validation of 3 functional disability scale for neck pain. Arch Phys
Med Rehabil 2002;83:376-82.
Objective: To translate and assess the reliability and the
construct validity of 3 functional disability scales for neck pain.
Design: Reliability and validity study.
Setting: Tertiary care teaching hospital and outpatient clinic.
Participants: One hundred one patients (mean age, 49y).
Intervention: French translations were obtained by using
the “translation-backward translation” method. Adaptations
were made after a pilot study.
Main Outcome Measures: Impairment outcome measures
(visual analog scale [VAS] pain, neck range of motion, morn-
ing stiffness, score of neck sensitivity, radiologic score of
Kellgren) and patients’ perceived handicap (VAS) were re-
corded at the baseline visit. Three functional disability scales
(Neck Disability Index [NDI], Neck Pain and Disability Scale
[NPDS], Northwick Park Neck Pain Questionnaire [NPQ])
were recorded twice, at baseline visit and 24 hours later.
Reliability was assessed by using the intraclass correlation
coefficient (ICC) and the Bland and Altman method. Construct
(convergent and divergent) validity was investigated by using
the Spearman rank correlation coefficient and a factor analysis
was performed.
Results: Test-retest was excellent for the NPDS and NDI
(ICC ⫽.91, .93, respectively) and good for the NPQ (ICC ⫽
.84). The Bland and Altman method showed no systematic
trend. Expected convergent and divergent validity were ob-
served only for the NPDS; 3 main factors were extracted by
factor analysis and explained 78% of the cumulative variance.
Conclusion: The 3 translated scales are valid, but the NPDS
seems to have the best construct validity.
Key Words: Disability; Neck pain; Outcome assessment;
Rehabilitation; Reproducibility of results; Translations.
©2002 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
CERVICAL SPINE DISORDERS are common complaint.
The point prevalence is nearly 50% in the United States,
1
and up to 41% of the working population reported neck pain
during the last 6 months.
2
In 30% of the cases, neck pain lasted
more than 6 months.
1,3
It is costly in terms of treatment and
lost work time. Total costs for posttraumatic neck disorders
are estimated to be $3.75 billion per year in the United
States.
4,5
Neck pain is usually defined as pain in the neck, shoulders,
and arms.
1,6
Symptoms arising in the neck are often poorly
localized, and there may be difficulty making a precise ana-
tomic diagnosis, particularly because the clinical signs of neck
disorders are neither sensitive nor specific.
1
Treatments for
neck pain are varied and not specific. A collar is used for acute
neck pain, but it encourages persisting neck stiffness.
7
Manual
therapies,
1
mobilization, or manipulative techniques performed
by physiotherapists have good results in chronic pain and
traction, transcutaneous electric nerve stimulation, and ultra-
sound are active but less popular.
1,8
Patient education to correct
postural abnormalities is usually necessary.
1
Drugs are regu-
larly associated with physical treatment, and surgery is the
last resort, except for radicular or spinal cord compression.
1
The results and perception of benefit of these treatments
vary.
1
To evaluate neck disorders and treatments, assessing
outcome measures of impairment and disability is necessary.
Impairment is usually evaluated by the pain intensity assessed
on a visual analog scale
9
(VAS), by determining cervical range
of motion (ROM), and by assessing cervical osteoarthritis on
radiograph.
10
Disability is evaluated by assessing daily living
activities.
Although measuring health status is an important component
of research by international health services and of clinical
practice, few validated neck pain and disability scales exist.
Disability outcome measures are needed, especially for non–
English-speaking patients.
11
Such scales can be developed by
creating a new scale, in the foreign population, or can be
developed by translating and adapting a preexisting scale. The
first possibility leads to multiplication of outcome measures
and lack of normalization.
1,4,9,12
The second has the advantage
of allowing comparison of different populations in interna-
tional trials, even if cultural adaptations are needed. The psy-
chometric properties of the created or translated scales must be
tested to obtain a validated instrument.
9,13
Five functional scales measuring neck disorders have been
developed and published in English-speaking countries: the
Neck Disability Index
4,14
(NDI), the Neck Pain and Disability
Scale
15
(NPDS), the Northwick Park Neck Pain Question-
naire
16
(NPQ), the Disability Rating Index,
17
and the Copen-
hagen Neck Functional Disability Scale.
18
None of them has
been translated or validated in a language other than English.
Three of these scales are specific to neck disorders and have
been partly validated; the NDI is based on the Oswestry Low
Back Pain Disability Questionnaire
19
(ODQ), a low back re-
gion-specific questionnaire. The NPDS is a recently published
scale developed by using the Million Visual Analog Scale. The
NPQ is a self-questionnaire adapted from the ODQ and has
nine 5-part questions to assess symptoms and disability.
We sought to translate 3 neck disability scales into French
and to assess their construct validity and reliability. If 1 or
From the Department of Physical and Rehabilitation Medicine, Hoˆpital Cochin
(Wlodyka-Demaille, Poiraudeau, Rannou, Revel) and Department of Biostatistics,
Hoˆpital Necker (Fermanian), Universite´ Rene´ Descartes, Paris, and Outpatient Clinic
in Physical and Rehabilitation Medicine (Catanzariti), Lille, France.
Accepted April 9, 2001.
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the authors(s) or upon any
organization with which the author(s) is/are associated.
Reprint requests to Serge Poiraudeau, MD, Service de re´e´ducation et de re´adapta-
tion de l’appareil locomoteur et des pathologies du rachis, 27 rue du Fbg St Jacques,
75014 Paris, France, e-mail: serge.poiraudeau@cch.ap-hop-paris.fr.
0003-9993/02/8303-6617$35.00/0
doi:10.1053/apmr.2002.30623
376
Arch Phys Med Rehabil Vol 83, March 2002
more of these scales has sufficient psychometric properties, it
will not be necessary to create a new French scale for neck
disorders.
METHODS
Participants
Outpatients and inpatients, ages 18 to 70 years and having
neck disorders for at least 15 days, were recruited from 1
rehabilitation department, 2 rheumatology departments, and an
outpatient clinic. Patients were excluded for (1) arthritis (in-
flammatory diseases), (2) loco-regional tumor or metastasis, (3)
myopathy, (4) inability to speak and read French fluently, (5)
severe psychiatric disorders, (6) absence of cervical spine ra-
diograph in the last 2 years, or (7) absence of oral consent.
The Scales
Neck Disability Index. The NDI
4,14
consists of 5 items
derived from the ODQ
19
and 5 items identified from feedback
from the practitioners, the patients, and a review of the litera-
ture. The items explore pain intensity, personal care, lifting,
reading, headaches, concentration, work, driving, sleeping, and
recreation. In ten 6-part questions, the NDI is scored from 0 (no
disability) to 5 (total disability), and the total score varies from
0 to 50 (total disability).
Neck Pain and Disability Scale. The NPDS
15
was devel-
oped by using the Million Visual Analog Scale
20
and assesses
neck pain in 20 items. The items explore pain intensity; its
interference with vocational, recreational, social, and func-
tional aspects of living; and the presence and extent of associ-
ated emotional factors. Patients answer by marking a 10-cm
VAS; it has 6 major divisions marked in equal intervals by
vertical bars (whole number). Midpoints for each interval are
also marked with 2 dots (half a point on a vertical slash).
Scoring of each item varies along a continuous scale from 0 to
5, with increments as small as a quarter point (a quarter on the
2 dots). The score for each item ranges from 0 to 5, and the
total score, the sum of the item scores, ranges from 0 (no
disability) to 100 (total disability).
Northwick Park Neck Pain Questionnaire. The NPQ
16
comprises nine 5-part questions also derived from the ODQ.
Each item ranges from 0 to 4, and the total score is obtained
by summing the items (range, 0 ⫽no disability, 36 ⫽total
disability). The items explore neck pain intensity, sleeping,
pins and needles in the arms at night, duration of symptoms,
carrying, reading and watching television, working and house-
work, social activities, and driving. A tenth question is added at
the second presentation to compare the patient’s clinical status
with the last time the NPQ was completed. The score is
presented as a percentage.
Translations
The 3 scales were translated by the forward and backward
translation procedure. A team of several professional bilingual
translators each translated the original scales once. As recom-
mended,
21-23
they were encouraged to strive for idiomatic
rather than word-for-word translation. None of them was fa-
miliar with this type of instrument or the questionnaires. Then
the investigators (physiatrists involved in spine diseases) re-
viewed the translations to make cultural and vocabulary adap-
tations. A backward translation of the reviewed version was
then made into English, to verify that the means of the scales
were preserved. Translation difficulties, cultural diversity, con-
ceptual equivalence, and vocabulary differences were high-
lighted by this translation technique.
24-35
A bilingual (native
English-speaking person), involved in medical adaptation and
translation, made the comparison between the originals and the
last French versions to detect discrepancies and difficulties in
adaptation.
Pretesting
For pretesting, a sample population completed the translated
questionnaires to check for any misunderstandings and devia-
tions in the translation. The quality and acceptability of the
translations were tested item by item. The pilot test served to
make final adaptations when necessary.
36,37
Testing the Scales
For each scale and each item, a “never done” choice was
added. Questions answered as “never done” by more than 5%
of patients were adapted or eliminated. The distribution of the
answers was examined, and if the distribution was badly
skewed, the item was eliminated or adapted.
Acceptability. The acceptability of each scale was studied
item by item concerning missing responses and multiple re-
sponses for a single-choice item. The answers were checked
before data recording and statistical analysis. The time needed
to complete the questionnaire was noted.
Test-retest. Self-administered questionnaires were com-
pleted twice at a 24-hour interval. We chose this interval to
avoid variations in clinical status and to avoid the patients
remembering previous answers. Delusive items were randomly
added to the 3 scales; these items had no value in the statistical
study. The patient’s overall assessment of neck pain was re-
corded on a 5-level adjectival ordinal scale to distinguish
clinically stable from improved or worsened patients.
Validity. Construct validity was investigated in 3 ways,
convergent validity, divergent validity, and factor analysis.
Convergent validity was assessed by correlating global scale
scores with variables that could be expected to have a converg-
ing relationship. These variables were (1) a pain VAS (VAS-
P), (2) a handicap VAS (VAS-Hd), and (3) a functional dis-
ability VAS (VAS-Fd). All the VASs ranged from 0mm (no
pain, no handicap, no disability, respectively) to 100mm (max-
imum pain, handicap, disability, respectively).
9
The VAS-Hd
was accompanied by the question, “What is your handicap
level due to your neck pain?” The VAS-Fd was evaluated by
the question, “How much does your neck pain bother you in
your daily activities?” These 2 VASs were explained to each
patient and were completed after the investigator was sure that
the concept was understood.
Divergent validity was assessed by correlating the global
scale scores with variables known to have a moderate relation
or no relation with functional disability and pain including (1)
the Hospital Anxiety and Depression Scale (HADS; range, 0 ⫽
depression, 21 ⫽anxiety)
38,39
; (2) the score of neck sensitivity
(cervico-occipital junction, upper cervical spine, middle cervi-
cal spine, lower cervical spine, trapezius; 0 ⫽no pain during
deep palpation, 1 ⫽pain during palpation; 2 ⫽pain during
light palpation; range, 0–20); (3) the radiologic score of osteo-
arthritis adapted from the Kellgren radiologic score
10
(0 ⫽no
osteoarthritis, 1 ⫽uncertain, 2 ⫽minimal osteoarthritis, 3 ⫽
moderate osteoarthritis, 4 ⫽severe osteoarthritis; range, 0–4);
and (4) cervical ROM (by measuring neck flexion and exten-
sion [in cm] rotation, by adding the left and right acromiocla-
vicular and mandibular distances [in cm]). Factor analysis was
performed by using principal component analysis to extract
factors. The retained factors in each scale had eigenvalues
greater than 1. Independent factors were obtained by using the
varimax rotation method.
377
FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
Study Design
Demographic and clinical characteristics of patients were
collected at baseline visit (day 1), and the 3 scales were
completed. At day 2, the 3 scales and the 20 delusive questions
were completed by the patients. Strict instructions were given
to the patients that questionnaires had to be performed alone
without help, that the instructions had to be read before an-
swering items, and that each item had 1 and only 1 response.
The questionnaires were returned by mail. If the questionnaire
was not returned 10 days after day 2, a postal reminder was
sent, and if this was not sufficient, the patient received a
telephone call.
Statistical Analysis
Statgraphics SYSTAT 9 Delta Soft威software
a
for Windows
was used for all statistical analyses. Quantitative variables were
described by using means, standard deviations (SDs), and
minimum and maximum values.
Test-retest was assessed by using simultaneously the intra-
class correlation coefficient
40
(ICC) and the Bland and Altman
method.
41
These 2 methods give complementary information as
shown by Atkinson and Nevill
42
and I-Kuei Lin and Chin-
chilli.
43
Construct validity was assessed with the nonparametric
Spearman rank correlation coefficient (r) for correlation be-
tween 2 quantitative variables because a normal distribution
could not be shown for all the parameters studied. Spearman
coefficient values were interpreted as excellent relationship
(ⱖ.9), good (.90–.71), moderate (.70–.5), fair (.50–.3), and
little or none (ⱕ0.3).
44
Factor analysis was performed as described earlier to explore
the factorial structure of the 3 translated scales.
RESULTS
Participants
The results are described by using mean ⫾SD (minimum
and maximum value). For the pilot study, 29 patients (19
women, 10 men) tested the 3 translated scales. Their mean age
was 51.79 ⫾10.39 years (range, 33–69y). Sixteen were em-
ployed, 6 were on sick leave, 6 were retired, and 1 was
unemployed.
For the study, 101 patients (69 women, 32 men) completed
the 3 scales from December 1999 to July 2000. Seventy pa-
tients (69.31%) were inpatients and outpatients in hospital
departments, and 31 (30.69%) were recruited in private prac-
tice. Table 1 shows demographic and clinical characteristics of
the patients. Exactly 92.08% were right-handed, 49.50% were
employed, 24.75% on sick leave, 14.85% retired, and 10.89%
unemployed.
Translation
Forward-backward translation was used to translate the 3
scales into French. Three translator teams and 3 investigators
were involved in this work. Three steps of forward-backward
translation were needed to obtain a consensus for the last
French version of the 3 scales. The final step was to retranslate
into English the last French version obtained to compare it with
the original English version. This step allowed confirmation
that the translated version and the original English version
explored the same dimensions. Instructions to patients were
repeated twice in different forms, and some of them were
highlighted to reduce misunderstanding. This step of forward-
backward translation and adaptation took 5 months.
Pilot Study
Some adaptations to the questions were made based on
observation of the pilot study. They concerned work (extended
to professional and domestic activities), driving (for the NDI
and the NPQ), and the use of medication to aid sleeping. In
French, concepts of leisure and social activities have slightly
different meanings than they do in English, and details and
examples were given to help the patients make their choices.
The patients did not understand how to mark the subdivided
VAS in the NPDS. The developers of the scales were contacted
and, after discussion and with their approval, we replaced the
subdivided VAS with the “classical” undivided VAS.
9
The
total score of this adapted scale ranges from 0 to 2000 by
summing the score of each item (each ranging from 0 to 100).
Testing the Scales
Item analysis. None of the items was excluded, even those
concerning driving, for which more than 5% of patients an-
swered “never done.” There was no ceiling or floor effect; none
of the distribution of answers distributions was badly skewed.
Acceptability. It took 7.37 ⫾6.77 minutes (range, 1–60min)
to answer the NDI, 6.41 ⫾3.77 minutes (range, 1.3–20min) for
the NPDS, and 7.24 ⫾7.44 minutes (range, 1–60min) for the
NPQ. The questions were well accepted by the patients.
Few items were omitted (1.98%, 1.98%, 2.97% for the NDI,
NPDS, NPQ, respectively). There were also few multiple an-
Table 1. Demographic and Clinical Characteristics of the Patients
Mean SD Minimum Maximum
Age (y) 49.178 11.251 21 69
Neck pain duration (mo) 112.944 114.424 0.5 531
Present episode duration (mo) 10.853 37.604 0.5 363
Morning stiffness (min) 53.990 153.933 0 1440
ROM flexion-extension (cm) 13.144 3.749 3 21
Rotation right and left (cm) 24.109 5.607 5 45
Score of neck sensitivity (range, 0–20) 1.901 2.777 0 15
Score of Kellgren (range, 0–4) 1.525 1.043 0 4
VAS-P (range, 0–100) 44.683 24.767 0 100
VAS-Fd (range, 0–100) 58.752 26.191 0 100
VAS-Hd (range, 0–100) 55.792 30.360 0 100
Score of depression (HADS) (range, 0–21) 6.881 4.346 0 18
Score of anxiety (HADS) (range, 0–21) 10.762 3.884 2 20
378 FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
swers (6.93%, 4.95%, 5.94% for the NDI, NPDS, NPQ, re-
spectively).
Test-retest. One patient who did not return the second
questionnaire was excluded from the study, and 8 patients were
excluded for test-retest because their clinical status changed
between the first and the second evaluation. (They considered
their neck pain as “much better” or “much worse” in the NPQ.)
The questionnaires were administered twice at an interval of
26.68 ⫾12.1 hours (range, 16.45–104.45h). The mean score of
NDI at the baseline visit was 20.02 ⫾8.97 (range, 2–41; scale
range, 0–50); for the NPDS, it was 952.03 ⫾388.1 (range,
135–1931; scale range, 0–2000); and for the NPQ, it was
15.49 ⫾5.99 (range, 1–28; scale range, 0–36).
The mean score at the second visit (day 2) was 19.91 ⫾9.02
(range, 3–39) for the NDI; 939.37 ⫾415.76 (range, 112–1936)
for the NPDS, and 17.4 ⫾5.88 (range, 4–31) for the NPQ.
The repeatability assessed by the ICC was excellent for the
NDI (.93) and the NPDS (.91). The NPQ had good repeata-
bility (.84). For the 3 scales, Bland and Altman
41
analysis (figs
1A–C) showed that the means of the differences were .310 ⫾
3.367, 16.46 ⫾179.56, and ⫺1.77 ⫹3.22 for the NDI, NPDS,
and NPQ, respectively. The distribution of the differences was
homogeneous with no systematic trend (r⫽⫺.048 for the
NDI, r⫽⫺.184 for the NPDS, r⫽.063 for the NPQ).
Construct validity. Table 2 shows the results of convergent
and divergent validity for the NDI, NPDS, and NPQ. For the
NPDS, the expected convergent and divergent validities were
observed. The correlation with the other algofunctional scales
were good for the NDI (r⫽.793) and the NPQ (r⫽.882). For
the NDI, VAS-P and VAS-Fd were expected to have better
Fig 1. Bland and Altman
41
graphic representation of (A) the reliabil-
ity of the NDI (mean of the difference, .310 ⴞ3.37), (B) the reliability
of the NPDS (mean of the difference, 16.46 ⴞ179.56), and (C) of the
reliability of the NPQ (mean of the difference, ⴚ1.77 ⴞ3.22).
Table 2. Construct Validity of the 3 Scales*
NDI NPDS NPQ
Convergent Validity
NDI 1 .793 .882
NPDS .793 1 .733
NPQ .882 .733 1
VAS-Fd .498
†
.627 .527
VAS-P .480
†
.515 .432
†
VAS-Hd .602 .667 .535
Divergent Validity
Score of anxiety (HADS) .426 .399 .370
Score of depression (HADS) .545
†
.493 .474
Score of neck sensitivity .302 .312 .275
Score of Kellgren .166 .039 .124
ROM (flexion-extension) ⫺.406 ⫺.455 ⫺.433
ROM (rotation) .247 .278 .246
* Correlation coefficient with other variables.
†
Unexpected correlation.
379FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
correlation with the scale score. The depression score of the
HADS, which was expected to be divergent, was convergent.
For the NPQ, the expected divergent validity was observed, but
the VAS-P was expected to have a better correlation with the
scale score.
Factor analysis. Table 3 shows the results of factorial
analysis of the 3 scales. For the NPDS, 3 factors were extracted
with eigenvalues greater than 1, which accounted for 78% of
the total variance. Table 4 shows the loading of each question
after varimax rotation on the 3 factors. The first factor repre-
sents function and disability (items 8–15), the second factor
represents neck pain intensity during movement (items 1, 3, 7,
16–20), and the third factor represents static neck pain inten-
sity (items 2, 4–6).
For the NDI, factor analysis extracted 2 factors with eigen-
values greater than 1, which accounted for 55% of the total
variance. Table 5 shows the loading of each question after
varimax rotation on the 2 factors. The first factor represents
function and disability (items 2, 6–8, 10); the second factor
represents neck pain (items 1, 4, 5, 9).
For the NPQ, 2 factors were extracted with eigenvalues
greater than 1, which accounted for 54% of the total variance.
Table 5 shows the loading of each question after varimax
rotation on the 2 factors. Neither of the factors could be
characterized.
DISCUSSION
Our study describes successive steps in translating and
adapting 3 algofunctional neck pain and disability scales into
French and provides analysis of the psychometric properties of
the translated scales. The 3 translated scales have acceptable
construct validity and good repeatability. The NPDS seems to
have a better factorial structure than the NPQ and NDI.
Clinicians need to assess regional disability to measure a
patient’s limitations and clinical course before they can pro-
pose or evaluate local treatment.
3
Disability caused by neck
disorders is not well known.
1-3,8
There are few specific instru-
ments assessing such disability, and they are not usually in-
cluded among variables used for evaluation of clinical disabil-
ity in neck disorders.
2,8
No specific neck disability instrument
exists in French; thus, development and validation of measure-
ment instruments for reflecting patient disability and enabling
international comparison of health outcomes are needed.
13,22
Translation and adaptation of preexisting scales has 2 ad-
vantages: translations provide a very cost-effective access to
disability and pain measures in the target language, and if the
translation shows good construct validity and good metric
properties, such translated instruments can be used in interna-
tional comparative studies.
35
On the other hand, developing a
new instrument is a time-consuming process in which the
efforts are devoted to conceptualization of the measure and
selection and reduction of its items.
21
Simple translation is
usually successful if the culture of the target population is
similar to that of the original population.
21
Because French
culture is mostly occidental, cultural adaptations during trans-
lation of scales from English should be minor. Another advan-
tage of translation is that original and translated versions of the
scale can be compared as a single instrument, without increas-
ing the number of outcome measure.
29
In the absence of a published consensus on translating
methods,
21
forward-backward translation was used. Conceptual
Table 3. Factors in Factor Analysis of the 3 Translated Scales
Factor Eigenvalue % Variance Cumulative %
NPDS
Factor 1 13.035 65.174 65.174
Factor 2 1.447 7.233 72.407
Factor 3 1.120 5.601 78.008
NDI
Factor 1 4.216 42.159 42.159
Factor 2 1.277 12.773 54.932
NPQ
Factor 1 3.866 42.958 42.958
Factor 2 1.033 11.483 54.441
Table 4. Varimax-Rotated Factor Matrix of the NPDS
Question Factor 1 Factor 2 Factor 3
1 .268 .592* .529
2 .361 .484 .710*
3 .418 .614* .774
4 .345 .291 .443*
5 .272 .320 .8038
6 .399 .209 .809*
7 .519 .543* .473
8 .771* .389 .203
9 .704* .451 .265
10 .635* .524 .116
11 .630* .245 .470
12 .814* .245 .392
13 .778* .278 .354
14 .834* .228 .283
15 .729* .268 .400
16 .377 .761* .344
17 .266 .863* .228
18 .267 .846* .195
19 .276 .778* .267
20 .258 .542* .409
* The highest loading of each item.
Table 5. Varimax-Rotated Factor Matrix of the NDI and NPQ
Question Factor 1 Factor 2
NDI
1 .158 .821*
2 .590* .412
3 .483* .431
4 .412 .662*
5⫺.03 .789*
6 .560* .290
7 .757* .040
8 .575* .231
9 .486 .605*
10 .817* .015
NPQ
1 .203 .673*
2 .663* .488
3 .522* .158
4 .035 .841*
5 .734* .064
6 .650* .122
7 .792* .201
8 .730* .296
9 .399 .543*
* The highest loading of each item.
380 FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
rather than literal translation must be made to preserve the
meaning of each item,
34
and these guidelines were provided to
the translation teams before they started translating. Concern-
ing patient instructions, the translation should not be literal but
should be adapted to the target population.
21
The aim was to
reduce misunderstanding as much as possible and thus the
amount of missing data.
21
In this study, we attempted to max-
imize patient comprehension by repeating and highlighting the
instructions.
Pilot studies are important to check the acceptability and
feasibility of translated questionnaires.
13,21,25,45
Confirmation
by a sample of patients that questions are acceptable without
arousing reluctance or hesitation to reply is necessary before
their use.
21
In this work, the main adaptations to the scales were
made following the results of the pilot study. For example, the
concepts of “social activities” and “leisure” have different
meanings in French and in American culture.
25
These terms are
not commonly used in French and had to be more precisely
defined and illustrated by examples. “Symptoms” was also a
term that was not precise enough for French patients (NPQ
item 4), and examples were added in the final versions. After
the adaptations were made, the items were acceptable and were
correctly understood in the final testing.
Testing the scales revealed no problems of skewness of the
answers, and no items were eliminated for this reason. How-
ever, more than 5% of patients answered “never done” for the
driving item in the NDI and NPQ. These items were not
eliminated because they are clinically relevant for disability
assessment of neck pain. Moreover, patients tested in this study
mostly live in urban zones where public transportation is
readily available and easier to use than a private car. The fact
that these scales are designed to be used by the general French
population, where more than 95% of adults drive, was another
reason these items were kept in the final version.
The repeatability of the scales was excellent or good. It is
unlikely that these results could be because of the short interval
between the 2 tests. The number of items was high (40), and the
addition of delusive items would have made their memoriza-
tion difficult. Patients might remember some questions but
would be unlikely to remember their previous answers. More-
over, for the NDI and NPQ, ICC values were the same as those
observed during development of the original scales when pa-
tients were interviewed at an interval of 2 to 5 days.
4,11,16
The
ICC value of the NPDS is reported here for the first time.
Construct validity is a major criterion of validity of a ques-
tionnaire. It is usually assessed by convergent and divergent
validities and by factorial analysis.
44
Observed and expected
convergent and divergent validity of the NPDS suggests good
construct validity, which is confirmed by the factorial structure
of the scale. Three factors were extracted and were easily
characterized. They explained more than 78% of the variance.
Four factors were characterized in the original version of the
NPDS,
15
and they explained 66.6% of the variance. The first
factor in the French version of the scale corresponds to the first
and fourth factors extracted in the original scale, whereas the
second and the third factors remained the same. The factorial
structure of the English and French scales were thus very close.
For the NDI, 2 measures were expected to be more conver-
gent (VAS-P, VAS-Fd). The fair correlation with pain is prob-
ably because of the few items directly focused on it (only 1
item in the questionnaire). On the other hand, the depression
score of HADS, which was expected to be divergent, was
convergent. This observation highlights the link between pa-
tient’s perceived pain and psychologic status.
19,46,47
After or-
thogonal rotation, 2 factors were extracted and easily charac-
terized, suggesting good factorial structure. However, only
55% of the cumulative variance was explained by these 2
factors. The factorial structure of the NDI was not studied in
the original English version.
4,14
For the NPQ, 1 criterion was expected to be more conver-
gent (VAS-P). Like the NDI, this questionnaire had only 1
question directly focused on pain. The factorial structure of this
scale seems to be uncertain because it was not possible to
characterize the 2 factors after orthogonal rotation. Moreover,
these 2 factors explain only 54% of the cumulative variance.
Factorial analysis was not performed for the original version of
the NPQ. Internal consistency was insufficiently studied but
showed that when pain intensity increased, the questionnaire
score also increased.
16
CONCLUSION
The 3 scales translated into French are reliable and valid
instruments for the assessment of pain and disability in neck
disorders. The NPDS appears to have the best construct valid-
ity with a good factorial structure. These results have to be
confirmed by sensitivity to change analysis. If they are, it will
not be necessary to create a new French neck pain and disabil-
ity scale. The translated French scales should be valuable for
assessing the effectiveness of physical therapy, local treatment,
surgery, and manipulative therapeutics in neck disorders, as
well as for use in the field of forensic science to evaluate
disability in daily living activities after motor vehicle crashes.
4
International clinical trials including English-speaking and
French-speaking populations could be considered by using the
neck algofunctional scales.
References
1. Barry M, Jenner JR. Pain in neck, shoulder and arm. Br Med J
1995;310:183-6.
2. Leclerc A, Niedhammer I, Landre MF, Ozguler A, Etore P,
Pietri-Taleb F. One-year predictive factors for various aspects of
neck disorders. Spine 1999;24:1455-62.
3. Gross AR, Hondras MA, Aker PD, Peloso P, Goldsmith CH.
Manual therapy for mechanical neck disorders. Cochrane Library
[CD-ROM] 2000;2.
4. Vernon H, Mior S. The Neck Disability Index: a study of reliabil-
ity and validity. J Manipulative Physiol Ther 1991;14:409-15.
5. Clark W, Haldeman S. The development of guideline factors for
the evaluation of disability in neck and back injuries. Spine
1993;18:1736-40.
6. Bovim G, Schrader H, Sand T. Neck pain in the general popula-
tion. Spine 1994;19:1307-9.
7. Revel M, Poiraudeau S, Mayoux-Benhamou A. Algies rachidi-
ennes et radiculaires. In: Held JP, Dizien O, editors. Traite´de
me´decine physique et de re´adaptation. Paris: Me´decine-Sciences
Flammarion; 1998. p 362-79.
8. Aker PD, Gross AR, Goldsmith CH, Peloso P. Conservative
management of mechanical neck pain: systematic overview and
meta-analysis. Br Med J 1996;313:1291-6.
9. Huskisson EC. Measurement of pain. Lancet 1974;9:1127-31.
10. Kellgren JH, Lawrence DM. Radiological assessment of osteoar-
thritis. Ann Rheum Dis 1957;16:494-502.
11. Andre´ JM, Salle JY, Echelles d’e´valuation fonctionnelle. In: Held
JP, Dizien O, editors. Traite´deme´decine physique et de re´adap-
tation. Paris: Me´decine-Sciences Flammarion; 1998. p 163-8.
12. Bombardier C, Tugwell P. Methodological considerations in func-
tional assessment. J Rheum 1987;14(15):6-10.
13. Bullinger M, Alonso J, Apolone G, et al. Translating health status
questionnaires and evaluating their quality: the IQOLA project
approach. J Clin Epidemiol 1998;51:913-23.
14. Hains F, Waalen J, Mior S. Psychometric properties of the Neck
Disability Index. J Manipulative Physiol Ther 1998;21:75-80.
15. Wheeler AH, Goolkasian P, Baird AC, Darden BV. Development
of the Neck Pain and Disability Scale: item analysis, face and
criterion-related validity. Spine 1999;24:1290-4.
381FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
16. Leak AM, Cooper J, Dyer S, Williams KA, Turner-Stokes L,
Frank AO. The Northwick Park Neck Pain Questionnaire, devised
to measure neck pain and disability. Br J Rheum 1994;33:469-74.
17. Salen BA, Spangfort EV, Nyren AL, Nordemar R. The Disability
Rating Index: an instrument for the assessment of disability in
clinical settings. J Clin Epidemiol 1994;47:1423-34.
18. Jordan A, Manniche C, Mosdal C, Hindsberger C. The Copenha-
gen Neck Functional Disability Index Scale: a study of reliability
and validity. J Manipulative Physiol Ther 1998;21:520-7.
19. Fairbank JC, Couper J, Davies JB, O’Brien JP. The Owestry
Low Back Pain Disability Questionnaire. Physiotherapy 1980;
66:271-3.
20. Million R, Haavik K, Jayson MI, Baker RD. Evaluation of low
back pain in the assessment of lumbars corsets with and without
back support. Ann Rheum Dis 1981;40:449-54.
21. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation
of health-related quality of life measures: literature review and
proposed guidelines. J Clin Epidemiol 1993;46:1417-32.
22. Perneger T, Leple`ge A, Guillain H, Ecosse E, Etter JF. COOP
charts in French: translation and preliminary data on instrument
properties. Qual Life Res 1998;7:683-92.
23. Perneger T, Leple`ge A, Etter JF. Cross-cultural adaptation of a
psychometric instrument: two methods compared. J Clin Epide-
miol 1999;52:1037-46.
24. Ren XS, Amick B, Zhou L, Gandek B. Translation and psycho-
metric evaluation of a Chinese version of the SF-36 Health Survey
in the United States. J Clin Epidemiol 1998;51:1129-38.
25. Bjorner JB, Kreiner S, Ware JE, Damsgaard MT, Bech P. Differ-
ential item functioning in the Danish translation of the SF-36.
J Clin Epidemiol 1998;51:1189-102.
26. Wagner A, Gandek B, Aaronson NK, et al. Cross-cultural com-
parisons of the content of SF-36 translations across 10 countries:
results from the IQOLA project. J Clin Epidemiol 1998;51:925-
32.
27. Brislin RW. Back-translation for cross-cultural research. J Cross
Cultural Psychol 1970;1:185-216.
28. Slim K, Bousquet J, Kwiatkowski F, Lescure G, Pezet D, Chip-
poni J. [First validation of the French version of the Gastrointes-
tinal Quality of Life Index (GIQLI)] [French]. Gastroenterol Clin
Biol 1999;23:25-31.
29. Gonzalez V, Stewart A, Ritter P, Lorig K. Translation and vali-
dation of arthritis outcome measures into Spanish. Arthritis
Rheum 1995;38:1429-46.
30. Fukuhara S, Bito S, Green J, Hsiao A, Kurokawa K. Translation,
adaptation and validation of the SF-36 Health Survey for use in
Japan. J Clin Epidemiol 1998;51:1037-44.
31. Apolone G, Mosconi P. The Italian SF-36 Health Survey: trans-
lation, validation and norming. J Clin Epidemiol 1998;51:1025-
36.
32. Loge J, Kaasa S, Hjermstad M, Kvien T. Translation and perfor-
mance of the Norwegian SF-36 Health Survey in patients with
rheumatoid arthritis. Data quality, scaling assumptions, reliability,
and construct validity. J Clin Epidemiol 1998;51:1129-38.
33. Aaronson N, Muller M, Cohen PD, et al. Translation, validation,
and norming of the Dutch language version of the SF-36 Health
Survey in community and chronic disease populations. J Clin
Epidemiol 1998;51:1055-68.
34. Hendrieson W, Russel I, Prihoda T, et al. Development and initial
validation of a dual-language English-Spanish format for the
arthritis impact measurement scales. Arthritis Rheum 1989;32:
1153-9.
35. Leple`ge A, Ecosse E, Verdier A, Perneger T. The French SF-36
Health Survey: translation, cultural adaptation and preliminary
psychometric evaluation. J Clin Epidemiol 1998;51:1013-23.
36. Dougados M. La mesure: initiation du clinicien au langage statis-
tique et me´thodologique. Paris: Expansion Scientifique Franc¸aise;
1996. p 95-137.
37. Didier JP. La me´trologie en re´e´ducation et re´adaptation fonction-
nelle. In: Held JP, Dizien O, editors. Traite´deme´decine physique
et de re´adaptation. Paris: Me´decine-Sciences Flammarion; 1998.
p 14-22.
38. Zigmond AS, Snaith RP. The hospital anxiety and depression
scale. Acta Psychiatr Scand 1983;67:367-70.
39. Razavi D, Delvaux N, Farvacques C, Robaye E. Screening for
adjustment disorders and major depressive disorders in cancer
in-patients. Br J Psychiatry 1990;156:79-83.
40. Shout PE, Fleiss JL. The effects of measurement errors on some
multivariables procedures. Am J Public Health 1977;67:1188-91.
41. Bland JM, Altman DG. Statistical methods for assessing agree-
ment between two methods of clinical measurement. Lancet 1986;
1:307-10.
42. Atkinson G, Nevill A. Comment on the use of concordance
correlation to assess the agreement between two variables. Bio-
metrics 1997;52:775-7.
43. I-Kuei Lin L, Chinchilli V. Rejoinder to the letter to the editor
from Atkinson and Nevill [letter]. Biometrics 1997;52:777-8.
44. Fermanian J. [Measuring agreement between 2 observers: a quan-
titative case] [French]. Rev Epidemiol Sante´ Publique 1984;32:
408-13.
45. Bae SC, Cook EF, Kim SY. Psychometric evaluation of a Korean
health assessment questionnaire for clinical research. J Rheum
1998;25:1975-9.
46. Jordan A, Bendix T, Nielsen H, Hansen FR, Host D, Winkel A.
Intensive training, physiotherapy, or manipulation for patients
with chronic neck pain: a prospective, single-blinded, randomised
clinical trial. Spine 1998;23:311-9.
47. Waddell G, Main CJ. Assessment of severity in low back disor-
ders. Spine 1984;9:204-8.
Supplier
a. SPSS Inc, 233 S Wacker Dr, 11th F1, Chicago, IL 60606.
382 FRENCH VALIDATION OF NECK DISABILITY SCALES, Wlodyka-Demaille
Arch Phys Med Rehabil Vol 83, March 2002
