Quality of life in systemic sclerosis: Psychometric properties of the World Health Organization Disability Assessment Schedule II

Authors

  • Marie Hudson,

    Corresponding author
    1. Sir Mortimer B. Davis Jewish General Hospital and McGill University, Montreal, Quebec, Canada
    • Sir Mortimer B. Davis Jewish General Hospital, Room A-216, 3755 Cote Saint Catherine Road, Montreal, Quebec, H3T 1E2, Canada
    Search for more papers by this author
  • Russell Steele,

    1. Sir Mortimer B. Davis Jewish General Hospital and McGill University, Montreal, Quebec, Canada
    Search for more papers by this author
  • Suzanne Taillefer,

    1. Sir Mortimer B. Davis Jewish General Hospital and McGill University, Montreal, Quebec, Canada
    Search for more papers by this author
  • Murray Baron,

    1. Sir Mortimer B. Davis Jewish General Hospital and McGill University, Montreal, Quebec, Canada
    Search for more papers by this author
  • Canadian Scleroderma Research Group

    Search for more papers by this author
    • Investigators of the Canadian Scleroderma Research Group are as follows: M. Abu-Hakima, M. Fritzler, S. LeClercq: Calgary, Alberta, Canada; P. Docherty: Moncton, New Brunswick, Canada; J. Dunne: Vancouver, British Columbia, Canada; N. Jones: Edmonton, Alberta, Canada; N. Khalidi: Hamilton, Ontario, Canada; P. Rahman: St. John's, Newfoundland, Canada; J. Markland: Saskatoon, Saskatchewan, Canada; J.-P. Mathieu: Montreal, Quebec, Canada; J. Pope: London, Ontario, Canada; D. Robinson: Winnipeg, Manitoba, Canada; D. Smith: Ottawa, Ontario, Canada; E. Sutton: Halifax, Nova Scotia, Canada


Abstract

Objective

To determine the validity of the World Health Organization Disability Assessment Schedule II (WHODAS II) in systemic sclerosis (SSc).

Methods

Patients enrolled in the Canadian Scleroderma Research Group registry participated in a standardized evaluation and completed the WHODAS II. Criterion validity was assessed by comparing the WHODAS II with the Medical Outcomes Study Short Form 36 (SF-36), construct validity was assessed by examining how it relates to common measures of outcome in SSc, and discriminative validity was assessed by examining how it distinguishes patients with more severe disease from those with less severe disease.

Results

A total of 402 patients with SSc were included (mean ± SD age 55 ± 13 years, 87% women, mean ± SD disease duration 11 ± 9 years). The mean ± SD WHODAS II score was 24.6 ± 17.4, and the greatest impairments were in life activities and mobility. There were moderate to good correlations between the WHODAS II and the SF-36 Physical Component Summary score (r = −0.44), the SF-36 Mental Component Summary score (r = −0.41), and measures of function (r = 0.54), depression (r = 0.44), pain (r = 0.40), and fatigue (r = −0.49, P < 0.0001 for all). The WHODAS II was able to consistently distinguish patients with milder disease from those with more severe disease.

Conclusion

The WHODAS II had good psychometric properties in patients with SSc and should be considered a valid measure of health-related quality of life in SSc.

INTRODUCTION

Systemic sclerosis (SSc) is a multisystem disorder characterized by a disturbance in fibroblast function, microvascular disease, and immune system activation, culminating in fibrosis of the skin and internal organs (1). SSc is associated with significant morbidity, including disfiguring skin thickening, finger ulcers, joint contractures, pulmonary hypertension, interstitial lung disease, chronic diarrhea, and renal failure. Functional disability is considerable (2) and depressive symptoms and major depression are common (3, 4). The disease thus encompasses broad multidimensional issues including biologic, psychological, and social processes. Not surprisingly, health-related quality of life (HRQOL), when measured using the Medical Outcomes Study Short Form 36 (SF-36) (5, 6), has been reported to be compromised in patients with SSc (7, 8). Nevertheless, there has been little work to date on HRQOL in patients with SSc and experts have recommended additional research in this area (9).

The International Classification of Functioning, Disability and Health (ICF) is the World Health Organization's (WHO's) revised classification of functioning and disabilities that is based on a biopsychosocial model (10). The World Health Organization Disability Assessment Schedule II (WHODAS II) is a generic HRQOL instrument developed to operationalize the core dimensions of the ICF (11). It is a multidimensional instrument that has 6 domains: understanding and communicating, getting around, self- care, getting along with others, household and work activities, and participation in society. The WHODAS II distinguishes itself from other HRQOL instruments in that it is based on an international classification system, it is applicable across cultures, and it treats all disorders at parity when determining the level of functioning. Furthermore, the WHODAS II assesses some functional and health-related issues that are not addressed in the SF-36, including participation in society and sexuality.

The WHODAS II has been evaluated and found to be valid in some rheumatologic conditions, in particular ankylosing spondylitis (12), back pain (13), and musculoskeletal conditions (defined as low back pain, rheumatoid arthritis, and osteoarthritis) (14). Given that the morbidity of SSc results from an interplay of biopsychosocial factors, the WHODAS II may be particularly well suited to assess HRQOL in this condition. We therefore undertook this study to determine the validity of the WHODAS II to measure HRQOL in patients with SSc.

PATIENTS AND METHODS

Design.

We conducted a cross-sectional study of a convenience sample of patients with SSc. Our objective was to assess criterion validity of the WHODAS II by comparing results obtained with this instrument with the SF-36; to assess construct validity by examining how the WHODAS II relates to common measures of outcome in SSc, including patient and physician global assessments, disease activity and severity, function, pain, fatigue, and depression; and to assess discriminative validity by examining how the WHODAS II distinguishes patients with more severe disease from those with less severe disease.

Study patients.

The study patients consisted of those enrolled in the Canadian Scleroderma Research Group registry. Patients in this registry are recruited from 15 centers across Canada. Patients must have a diagnosis of SSc made by the referring rheumatologist, be ≥18 years of age, and be fluent in English or French. The patients included in this study were those whose baseline visit occurred between September 2004 and August 2006. Patients recruited into the registry undergo an extensive standardized evaluation including a history, physical evaluation, patient and physician global assessments, and laboratory investigations. Patients also complete a series of self-administered questionnaires to measure HRQOL, function, depression, fatigue, and pain (see section on outcome measures below).

Study instrument.

The self-administered WHODAS II consists of 36 Likert formatted questions covering 6 domains: understanding and communicating (cognition and conversation), getting around (mobility), self-care (attending to one's hygiene, dressing, eating, and staying alone), getting along with others (interpersonal interactions and sexuality), life activities (divided between household responsibilities and work/school activities), and participation in society (joining in community activities) during the last 30 days. Using SPSS syntax (SPSS, Chicago, IL) available through the WHO, each domain is weighted and scored separately. An overall score can also be generated. The scores range from 0 (best) to 100 (worst). Permission to translate the English version of the WHODAS II into Canadian French and to use both versions was obtained from the WHO.

Translation of the WHODAS II into Canadian French.

The translation protocol was based on methodology that has been previously published (15, 16) and applied to generate Canadian French versions of evaluation instruments (17). The details of the translation process and validation of the translated instrument will be reported separately.

Outcome measures.

The SF-36 (5, 6) is widely used to measure quality of life. It consists of 8 domains: physical functioning, social functioning, role limitations related to physical problems, role limitations related to emotional problems, mental health, vitality, bodily pain, and general health perceptions. Each domain can be scored separately. Scores can also be summarized into a Physical Component Summary (PCS) score and a Mental Component Summary (MCS) score. Scores range from 0 (worst) to 100 (best). For the purposes of this study, fatigue was measured using the vitality domain of the SF-36.

Function was assessed using the Scleroderma Health Assessment Questionnaire (SHAQ). The SHAQ consists of the Health Assessment Questionnaire (HAQ) disability index (DI) and scales that measure symptoms specific to SSc (2). The HAQ DI is a self-administered measure intended to assess functional ability in patients with arthritis (18). It includes 20 questions in 8 categories (dressing, standing, eating, walking, toileting, reaching, gripping, and instrumental activities). The use of assistive aids and devices to help with function is recorded. Scores are derived as the average of the score for the most abnormal activity in each of the 8 categories, taking into account the use of assistive aids and devices. Scores range from 0 (no disability) to 3 (severe disability). The disease-specific questions in the SHAQ relate to the severity of Raynaud's phenomenon, digital ulcers, gastrointestinal symptoms, lung symptoms (shortness of breath), and overall disease severity in the past week. Each question is anchored by the descriptors “does not interfere” and “very severe limitation” and scored separately. Unlike the visual analog scales originally used for the SHAQ, the assessments in this study were made using 11-point numerical rating scales (NRS).

The Center for Epidemiologic Studies Depression Scale (CES-D) is a 20-item scale that was designed to measure depression in the general population (19, 20). It is also useful in clinical and psychiatric settings. The CES-D asks an individual to report the frequency with which each of 20 events was experienced during the previous week. The items are graded on a 4-point scale ranging from 0 to 3 and corresponding to the frequency of each symptom in the past week. The CES-D yields a summary score that ranges from 0 to 60, with higher scores indicating greater depression. The scale is used as a simple indicator of the degree of depression. If the total score is ≥16, the patient may have experienced some depression in the past week. Some recommend a score of ≥19 as a cutoff when evaluating patients with chronic pain (21).

The Short Form McGill Pain Questionnaire (MPQ) (22–25) includes a pain rating index that contains 11 questions referring to the sensory dimension of the pain experience and 4 questions related to the affective dimension. Each descriptor is ranked on a 4-point intensity scale (0 = none, 1 = mild, 2 = moderate, 3 = severe). The score ranges from 0 to 45.

Disease activity was measured using the Valentini Scleroderma Disease Activity Index (25, 26), which consists of 10 variables: total skin score >14; scleredema; digital necrosis; arthritis; total lung capacity <80%; erythrocyte sedimentation rate >30 mm/hour; hypocomplementemia; and change in cardiopulmonary, skin, and vascular symptoms in the past month. Each variable is weighted and the final score ranges from 0 (no activity) to 10 (very active). Disease severity was measured using the Medsger Disease Severity Scale (27, 28), which assesses general health and 8 organ systems: peripheral vascular, skin, joint/tendon, muscle, gastrointestinal tract, lungs, heart, and kidneys. Each system is scored separately from 0 to 4 depending on whether there is no, mild, moderate, severe, or end-stage involvement.

The physician global assessments of activity, severity, and damage were assessed using 11-point NRS asking the physician to rate the patient's scleroderma in the past week. The patient global assessment of disease severity was assessed as part of the SHAQ, described above.

Statistical analysis.

Descriptive statistics were used to summarize the baseline characteristics of the patients. To assess the criterion and construct validity of the WHODAS II, correlations between the WHODAS II, the SF-36, and various outcome measures were assessed using nonparametric Kendall's tau tests. In general, correlation coefficients <0.3, 0.3–0.49, and >0.5 represent small, medium, and large correlations, respectively (29).

Discriminative validity assesses the ability of a scale to distinguish between groups that differ according to some factor that is expected to be associated with the construct being measured. Therefore, dichotomous subsets were constructed to identify patients with less severe and more severe disease according to various outcome variables. Patients were placed in subsets based on total body skin score according to the same cutoffs of modified Rodnan skin scores (<20 and ≥20) (30, 31) used to validate the SHAQ (2). For forced vital capacity (FVC) and lung diffusing capacity for carbon monoxide (DLCO), we also applied the same cutoffs as those used by Khanna et al (32), namely, percent predicted FVC >70 and ≤70, and percent predicted DLCO ≥50 and <50. Patients were placed in subsets by HAQ DI scores indicating mild (<1) or moderate to severe (≥1) functional impairment and by CES-D scores indicating presence (≥19) or absence (<19) of depression. Because no a priori cut points were available for the outcomes of patient and physician global assessments of severity, disease activity and severity, SF-36 MCS and PCS, fatigue, and pain, median values were used. The mean WHODAS II score of each subset for each variable was computed and the discriminative validity of the WHODAS II was tested by comparing the means of the subsets using the nonparametric Wilcoxon's signed rank test.

A priori, P values less than 0.05 were considered statistically significant despite the fact that multiple comparisons were to be performed. No adjustment for multiple testing was used because our objectives were exploratory in nature. Thus, the P value served as a relative measure of potential clinical interest rather than as a mechanism for performing formal tests of hypotheses. Similar reasons not to correct for multiple comparisons have previously been invoked (32). Nevertheless, post hoc, we summed the total number of comparisons and applied a Bonferroni correction. We had a total of 106 comparisons (63 for criterion validity, 26 for construct validity, and 17 for discriminative validity). Thus, using this conservative method of correction for multiple testing, a P value less than 0.0005 was considered statistically significant.

Finally, a more advanced technique, partial least squares regression (33), was also used to assess the criterion and construct validity of the WHODAS II. Good correlations between the overall WHODAS II scores and the criterion and/or constructs may still mask poor correlations between subscales of the WHODAS II and the criterion and/or construct validity. Partial least squares regression makes it possible to assess the appropriateness of the weighting of the individual WHODAS II subscales beyond simple correlations of overall scores. Therefore, for criterion validity, we used partial least squares to estimate weights for the WHODAS II subscales that would be optimal for patients with SSc with respect to the 7 domains of the SF-36 as well as its 2 main subscales (the PCS and MCS) and compared these with the original weights provided in the current WHODAS II syntax. For construct validity, we repeated the same process using indicators of function, pain, and depression (namely, the HAQ DI, MPQ, and CES-D, respectively) to estimate weights for the WHODAS II subscales that would be optimal for patients with SSc and again compared these with the weights in the current syntax. For both criterion and construct validity, we used 10-fold cross-validated root mean-squared error of prediction of individual response measures as our criteria for assessing the quality of the weighting. All statistical analyses were performed with SPSS software, version 13 (SPSS, Chicago, IL) and the R statistical package (R Foundation for Statistical Computing, Vienna, Austria).

Ethical considerations.

Ethics committee approval for this study was obtained at each site and each patient provided informed written consent to participate in this study.

RESULTS

At the time this study was undertaken, 438 patients had been recruited into the Canadian Scleroderma Research Group registry. Thirty-six were excluded because their demographic information had not yet been entered into the database (n = 25) or their WHODAS II data were missing (n = 11). The baseline characteristics of the 402 patients included in this study are presented in Table 1. Mean ± SD age was 55 ± 13 years and 87% were women. The mean ± SD duration of disease (measured from the time of onset of the first non-Raynaud's manifestation of SSc) was 11 ± 9 years. On average, disease activity and severity were low. Functioning measured by the HAQ DI was mildly impaired but depression, pain, and fatigue were common. The mean ± SD WHODAS II score was 24.6 ± 17.4 (Table 2). The greatest impairments appeared in life activities and mobility (getting around). The mean ± SD SF-36 PCS score was 36.4 ± 11.0 and the mean SF-36 MCS score was 49.1 ± 11.1. The physical functioning, general health, and role physical subscales of the SF-36 had the lowest mean scores.

Table 1. Baseline characteristics of a cohort of patients with systemic sclerosis (SSc) enrolled in the Canadian Scleroderma Research Group registry (n = 402)*
 Mean ± SD or %Median (IQR)
  • *

    IQR = interquartile range; HAQ DI = Health Assessment Questionnaire disability index.

  • Low values for global assessments, Scleroderma Disease Activity Score, disease severity score, HAQ DI, depression, and pain represent better outcomes and high values represent worst outcomes. In contrast, low values for fatigue represent worst outcomes and high values represent better outcomes.

Demographics  
 Women86.6 
 Age, years55.4 ± 12.756.0 (47.0–65.0)
 Married or living with someone72.1 
 Living alone14.2 
 Education (more than high school)44.9 
 Currently employed or in school37.6 
 Yearly family income ≥$50,00045.4 
Disease characteristics  
 Disease duration in years  
  Since onset of Raynaud's14.0 ± 11.411.4 (4.9–19.6)
  Since onset of first non-Raynaud's manifestation   of SSc10.6 ± 8.58.0 (3.6–14.8)
 Physician global assessments (range 0–10)  
  Disease severity2.7 ± 2.32 (1–4)
  Disease activity2.3 ± 2.02 (1–3)
  Disease damage3.3 ± 2.43 (1–5)
 Patient global assessments (range 0–10)  
  Health3.7 ± 2.53 (2–5)
  Pain3.7 ± 2.83 (1–6)
  Raynaud's phenomenon3.0 ± 3.02 (0–5)
  Finger ulcers2.1 ± 3.00 (0–4)
  Gastrointestinal problems2.1 ± 2.90 (0–4)
  Breathing2.2 ± 2.61 (0–4)
  Disease severity3.6 ± 2.53 (1–5)
 Scleroderma Disease Activity Score (range 0–10)2.8 ± 2.11.9 (1.2–3.7)
 Disease severity score (range 0–4)  
  General1.0 ± 1.31 (0–2)
  Gastrointestinal tract2.1 ± 0.772 (2–2)
  Muscle0.3 ± 0.80 (0–0)
  Skin1.3 ± 0.71 (1–2)
  Peripheral vascular system1.6 ± 1.21 (1–3)
  Heart0.5 ± 1.10 (0–0)
  Lung1.2 ± 1.21 (0–2)
  Joint/tendon0.94 ± 1.30 (0–0)
  Kidney0.2 ± 0.70 (0–0)
 HAQ DI (range 0–3)0.8 ± 0.70.75 (0.25–1.25)
 Depression (range 0–60)13.9 ± 10.212 (6–19)
 Pain (range 0–45)6.9 ± 7.84 (1–9.5)
 Fatigue (range 0–100)45.7 ± 10.845.9 (36.5–52.1)
Table 2. Baseline values of WHODAS II and SF-36 in a cohort of patients with systemic sclerosis enrolled in the Canadian Scleroderma Research Group registry*
 Mean ± SDMedian (IQR)
  • *

    The WHODAS II ranges from 0 (best) to 100 (worst) and the SF-36 ranges from 0 (worst) to 100 (best). WHODAS II = World Health Organization Disability Assessment Schedule II; SF-36 = Short Form 36; IQR = interquartile range.

WHODAS II  
 Understanding and communicating13.5 ± 16.45 (0–25)
 Getting around30.5 ± 23.931.3 (6.3–50)
 Self-care15.3 ± 20.210 (0–30)
 Getting along with people16.0 ± 17.68.4 (0–25)
 Life activities (not at work or in school)36.3 ± 27.240 (10–50)
 Life activities (at work or at school)47.3 ± 42.235.7 (7.1–100)
 Participation in society26.4 ± 18.825 (12.5–41.7)
 Total if working or in school22.7 ± 15.820.7 (8.7–32.6)
 Total if not working or in school23.6 ± 17.222.6 (9.2–33.3)
 Overall total24.6 ± 17.423.6 (9.7–35.9)
SF-36  
 Physical function36.0 ± 11.636 (25.5–46.5)
 Social function43.2 ± 11.745.9 (35–56.9)
 Role physical39.8 ± 12.039.7 (29.9–49.5)
 Role emotional44.5 ± 12.548.1 (32.6–55.9)
 Mental health48.4 ± 10.550 (41.6–57.8)
 Vitality45.7 ± 10.845.9 (36.5–52.1)
 Bodily pain42.5 ± 9.441.4 (37.2–51.1)
 General health37.9 ± 11.036.3 (30.5–45.8)
 Physical Component Summary36.4 ± 11.037.8 (28.3–44.2)
 Mental Component Summary49.1 ± 11.151.4 (41.4–58.6)

Criterion validity.

There was moderate correlation between the WHODAS II and the SF-36 PCS (r = −0.44, P < 0.0001) and between the WHODAS II and the SF-36 MCS (r = −0.41, P < 0.0001) (Table 3). The negative correlations are due to the fact that higher scores on the WHODAS II represent worse HRQOL, whereas higher scores represent better HRQOL on the SF-36. Similarly, there were overall moderate to good correlations between the corresponding domains of the WHODAS II and the SF-36. For example, mobility on the WHODAS II correlated strongly with physical function on the SF-36 (r = −0.65, P < 0.0001) and participation in society measured on the WHODAS II also correlated well with social function on the SF-36 (r = −0.49, P < 0.0001).

Table 3. Criterion validity: correlation coefficients (r) between the WHODAS II and the SF-36*
 WHODAS II activity limitationsWHODAS II participation
CommunicationMobilitySelf-careInterpersonalWorkParticipationWHODAS II overall
  • *

    All correlations were significant at P ≤ 0.0001. WHODAS II = World Health Organization Disability Assessment Schedule II; SF-36 = Short Form 36; PCS = Physical Component Summary; MCS = Mental Component Summary.

SF-36       
 Impairment       
  Mental health−0.39−0.26−0.22−0.34−0.33−0.29−0.42
  Vitality−0.38−0.43−0.30−0.36−0.44−0.40−0.44
  Bodily pain−0.27−0.47−0.37−0.27−0.40−0.44−0.43
 Activity limitations       
  Physical function−0.26−0.65−0.43−0.32−0.50−0.49−0.42
 Participation       
  Role physical−0.34−0.55−0.45−0.35−0.58−0.59−0.56
  Role emotional−0.43−0.36−0.33−0.36−0.39−0.45−0.49
  Social function−0.37−0.50−0.40−0.42−0.50−0.49−0.56
 Composite       
  PCS−0.23−0.61−0.44−0.29−0.52−0.52−0.44
  MCS−0.40−0.24−0.22−0.35−0.30−0.31−0.41

Using partial least squares regression, we found the subscales of the WHODAS II to be properly weighted with respect to 1) the 7 domains of the SF-36 (Kendall's tau of 0.90 for measuring the relationship between WHODAS II scores obtained using the estimated weights and scores obtained using the original weights proposed by the WHO) and 2) the 2 primary SF-36 subscales (the PCS and MCS; Kendall's tau of 0.89 and 0.90, respectively). Additionally, there was little to no difference in the reduction in prediction error of the various SF-36 subscales when using the estimated weights for the subscales in place of the original weights (Table 4).

Table 4. Comparison of percent reduction in root mean-squared error prediction (% RMSEP) between WHODAS II scores obtained with the original and estimated weights for the WHODAS II subscales*
 % RMSEP using original WHODAS II subscale weights% RMSEP using estimated WHODAS II subscale weights
  • *

    WHODAS II = World Health Organization Disability Assessment Schedule II; SF-36 = Short Form 36; PCS = Physical Component Summary; MCS = Mental Component Summary; HAQ DI = Health Assessment Questionnaire disability index; CES-D = Center for Epidemiologic Studies Depression Scale; MPQ = McGill Pain Questionnaire.

SF-36 PCS37.139.8
SF-36 MCS18.513.6
SF-36 subscales  
 Mental health14.210.4
 Vitality20.918.7
 Bodily pain25.724.8
 Physical function37.237.9
 Role physical44.544.4
 Role emotional28.323.5
 Social function34.530.8
HAQ DI34.835.2
CES-D20.715.5
MPQ15.812.9

Construct validity.

The correlations were good between the WHODAS II and function (r = 0.54) and moderate between the WHODAS II and measures of depression (r = 0.44), pain (r = 0.40), and fatigue (r = −0.49; P < 0.0001 for all) (Table 5). The correlation between the WHODAS II and disease activity was weak but significant and the correlations between the WHODAS II and disease severity in 7 organ systems (general, gastrointestinal, muscle, heart, lung, joint/tendon) were also weak but significant. The WHODAS II did not correlate with disease duration and disease severity in the peripheral vascular or kidney systems.

Table 5. Construct validity: correlation coefficients (r) between the WHODAS II and various outcome measures*
 rP
  • *

    WHODAS II = World Health Organization Disability Assessment Schedule II; NS = not significant; SSc = systemic sclerosis; HAQ DI = Health Assessment Questionnaire disability index.

Mean disease duration in years  
 Since onset of Raynaud's−0.06NS
 Since onset of first non-Raynaud's  manifestation of SSc0.01NS
Physician global assessments  
 Disease severity0.30≤ 0.0001
 Disease activity0.22≤ 0.0001
 Disease damage0.25≤ 0.0001
Patient global assessments  
 Health0.45≤ 0.0001
 Pain0.37≤ 0.0001
 Raynaud's phenomenon0.29≤ 0.0001
 Finger ulcers0.21≤ 0.0001
 Gastrointestinal problems0.28≤ 0.0001
 Breathing0.35≤ 0.0001
 Disease severity0.46≤ 0.0001
Scleroderma Disease Activity  Score0.18≤ 0.0001
Disease severity score  
 General0.18≤ 0.0001
 Gastrointestinal tract0.19≤ 0.0001
 Muscle0.20≤ 0.0001
 Skin0.15≤ 0.0001
 Peripheral vascular system0.06NS
 Heart0.11≤ 0.0001
 Lung0.08≤ 0.04
 Joint/tendon0.13≤ 0.001
 Kidney0.04NS
HAQ DI0.54≤ 0.0001
Depression0.44≤ 0.0001
Pain0.40≤ 0.0001
Fatigue−0.49≤ 0.0001

Again, using partial least squares regression, we derived a measure that best summarized the correlation among 3 quality of life indicators (depression, pain, and function) and used this to estimate optimal weights for the WHODAS II subscales. The correlation between the WHODAS II scores obtained using the estimated and original weights for the subscales was very high (Kendall's tau 0.93). Also, there was little to no difference in the reduction in prediction error of the 3 quality of life indicators when using the estimated weights for the subscales in place of the original weights suggested by the WHO (Table 4).

Discriminative validity.

Patients were divided into dichotomous groups based on a number of disease variables (Table 6). The WHODAS II was able to consistently discriminate between patients with milder disease and those with worse disease. All comparisons were significant with P values less than 0.0001.

Table 6. Discriminative validity: comparison of WHODAS II scores for various disease subsets*
 NWHODAS II score, mean ± SDP
  • *

    WHODAS II = World Health Organization Disability Assessment Schedule II; DLCO = diffusing capacity for carbon monoxide; HAQ DI = Health Assessment Questionnaire disability index; CES-D = Center for Epidemiologic Studies Depression Scale; SF-36 = Short Form 36.

  • P values are for comparisons between subsets.

  • Median values were used as cutoffs.

Total body skin score  ≤ 0.0001
 ≥207531.1 ± 18.1 
 <2031722.7 ± 16.7 
Forced vital capacity, % predicted  ≤ 0.0001
 ≤704731.1 ± 18.3 
 >7024021.5 ± 16.5 
Lung DLCO, % predicted  ≤ 0.0001
 <504625.5 ± 17.6 
 ≥5023122.4 ± 16.8 
HAQ DI  ≤ 0.0001
 ≥116536.1 ± 15.2 
 <123416.5 ± 13.4 
CES-D  ≤ 0.0001
 ≥1910538.8 ± 15.3 
 <1928919.4 ± 14.6 
SF-36 Mental Component  Summary score  ≤ 0.0001
 <5119433.4 ± 16.1 
 ≥5120216.1 ± 13.5 
SF-36 Physical Component  Summary score  ≤ 0.0001
 <3619834.8 ± 15.0 
 ≥3619814.3 ± 12.3 
Patient global assessment of  disease severity  ≤ 0.0001
 ≥324432.1 ± 15.7 
 <315412.5 ± 11.6 
Physician global assessment of  disease severity  ≤ 0.0001
 ≥224728.6 ± 17.6 
 <214717.5 ± 14.4 
Fatigue  ≤ 0.0001
 <4620434.1 ± 15.7 
 ≥4619615.0 ± 13.1 
Pain  ≤ 0.0001
 ≥421430.9 ± 16.2 
 <417516.4 ± 14.3 
Scleroderma Disease Activity Score  ≤ 0.0001
 ≥324030.3 ± 16.1 
 >314914.9 ± 13.6 
Disease severity score   
 General  ≤ 0.0001
  ≥118827.8 ± 17.8 
  <117721.0 ± 16.6 
 Gastrointestinal tract  ≤ 0.0001
  ≥236725.8 ± 17.3 
  <22812.8 ± 14.1 
 Skin  ≤ 0.0001
  ≥136824.5 ± 17.3 
  <12422.1 ± 16.0 
 Peripheral vascular system  ≤ 0.0001
  ≥128225.3 ± 18.0 
  <111022.7 ± 15.5 
 Lung  ≤ 0.0001
  ≥125525.7 ± 17.5 
  <114022.2 ± 16.9 

DISCUSSION

In this large cohort of more than 400 patients with SSc, we have demonstrated that the WHODAS II, an instrument designed to measure HRQOL that is compatible with the new WHO classification of functioning and disabilities, correlates well with the SF-36 and common outcomes of disease including function, depression, pain, and fatigue. The WHODAS II also has good ability to discriminate between subsets of patients with less severe disease compared with those with more severe disease. Thus, the WHODAS II has good criterion, construct, and discriminative validity in SSc.

In addition, using partial least squares regression, we found that the subscales of the WHODAS II are appropriately weighted for patients with SSc. Otherwise, if the subscales of the WHODAS II should have been weighted differently for patients with SSc, we would have seen much lower correlations between the estimated scores and those obtained using the original weights suggested by the WHO. As it stands, there is no advantage to using a different weighting system for the WHODAS II subscales in SSc over those suggested by the WHO. The use of partial least squares regression for assessing construct and criterion validity for an existing instrument is novel and adds significantly to the validation process in 2 ways. First, it allows an assessment of the appropriateness of the weighting of the individual subscales beyond simple correlations of overall scores with criterion and/or construct measures, which could mask an improper weighting of one or more subscales. Second, it allows confirmation that the weights are appropriate in the disease under consideration, in this case SSc. The WHODAS II has been proposed as an HRQOL measure that can be used across diseases. Our findings of the appropriateness of the weights in our cohort confirm that the WHODAS II is indeed appropriate in SSc.

The Outcome Measures in Rheumatology Clinical Trials (OMERACT) filter is used to evaluate the measurement properties of outcome measures in rheumatology and has 3 components: truth, discrimination, and feasibility (34). The truth component addresses the question of whether the measure assesses what it is meant to assess in an unbiased and relevant manner and encompasses face, content, criterion, and construct validity. The discrimination component is meant to determine whether the measure discriminates between situations of interest, either at one time or over time and encompasses the concepts of reliability and sensitivity to change. The feasibility component addresses the question of whether the measure can be used easily, given constraints of time, money, and interpretation. This analysis allowed us to demonstrate that the WHODAS II has good characteristics in some, but not all components of the OMERACT filter. Criterion, construct, and discriminative validity were all specifically assessed. Although face and content validity and reliability were not specifically assessed in this study, they were assessed when the WHODAS II was developed (11). In addition, the WHODAS II takes only a few minutes to complete, does not cost anything, and the French Canadian version has been translated and validated (manuscript in preparation). Also, further studies to assess sensitivity to change in SSc are ongoing.

The WHODAS II is a new instrument. Therefore, it is currently difficult to appreciate the meaning of scores. For purposes of comparison, the mean ± SD WHODAS II in this cohort was 24.6 ± 17.4, whereas it was reported to be 23.9 ± 15.5 in ankylosing spondylitis (12), 22.8 ± 15.4 in patients with back pain (13), and 22.0 ± 14.3 in patients with various musculoskeletal conditions (14). The mean ± SD SF-36 PCS and MCS scores were 36.4 ± 11.0 and 49.1 ± 11.1, respectively, in the present study, whereas they have been reported to range from 31.8 ± 13.2 to 37 ± 12 and from 42 ± 11 to 49.8 ± 9.6, respectively, in other SSc cohorts (32, 35–37). Canadian normative values for the SF-36 PCS and MCS in patients ages 55–64 years are 49.0 (95% confidence interval 48.6–49.3) and 53.7 (95% confidence interval 53.4–54.0), respectively (38). Thus, from a physical point of view, HRQOL in our cohort appears comparable with that of patients with other rheumatic diseases and of other SSc cohorts but is significantly below that of individuals in the general population. However, from a mental health point of view, HRQOL in our patients is comparable with the normal population (6, 38).

There are some limitations to our study. First, this was a cross-sectional study of patients enrolled in a large SSc registry using data from their baseline visits. In the absence of data over time, we could not assess the responsiveness of the WHODAS II over time. Others have found that the WHODAS II is responsive to change in patients with certain rheumatic conditions (12). In addition, the ability of an instrument to detect clinically meaningful differences can be assessed cross-sectionally by comparing groups that are different in terms of some disease-related variable (39). Our discriminative analysis, which clearly showed the ability of the WHODAS II to distinguish between subsets of patients with less and more severe disease, is such an example. Thus, we believe that the WHODAS II will be found to be responsive over time. Ongoing longitudinal studies in our patients will allow us to confirm this. Second, a few studies have now been published demonstrating that the SF-36 is a valid measure of HRQOL in patients with SSc. Our study was designed to assess the validity of the WHODAS II. Whether it captures HRQOL better than other instruments in current use, in particular the SF-36, remains unknown. However, the fact that the WHODAS II contains items that are not contained in the SF-36 and that may be of particular importance in SSc (e.g., participation in society, sexuality) suggests that it may assess certain aspects of HRQOL better that the SF-36. Further studies to explore this will be needed.

The strength of our study lies in the large sample size. This is the largest study to date to investigate HRQOL in patients with SSc. In addition, our cohort spans Canada. Therefore, our patients are geographically, culturally, and linguistically diverse. This adds to the generalizability of our findings. It also makes the WHODAS II an especially appropriate instrument to measure HRQOL in our cohort because it was specifically designed to be applied across cultures (11).

Although SSc is known to be associated with significant impairment and disablement, HRQOL has not been well studied in these patients. The low prevalence of this disease may in part explain this, making it difficult for single investigators or centers to pursue meaningful research in SSc. The Canadian Scleroderma Research Group is a unique multicenter consortium of investigators whose overall objective is to improve the HRQOL of patients with SSc. The WHODAS II appears to be a valid tool to allow us to measure the impact of our research.

AUTHOR CONTRIBUTIONS

Dr. Hudson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Hudson, Steele, Baron.

Acquisition of data. Hudson, Taillefer, Baron.

Analysis and interpretation of data. Hudson, Steele, Taillefer, Baron.

Manuscript preparation. Hudson, Steele, Taillefer, Baron.

Statistical analysis. Hudson, Steele, Taillefer, Baron.

ROLE OF THE STUDY SPONSOR

The funding sources had no role in the design of the study, analysis of the data, preparation of the manuscript, or decision to submit for publication.

Ancillary