Rasch analysis suggests that health assessment questionnaire II is a generic measure of physical functioning for rheumatic diseases: a cross-sectional study

Rasch analysis suggests that health assessment questionnaire II is a generic measure of physical... Background: Versions of the Health Assessment Questionnaire (HAQ) are commonly used to measure physical functioning across multiple rheumatic diseases but there has been no clear demonstration that any HAQ version is actually generic. This study aimed to show that the HAQ-II instrument is invariant across different rheumatic disease categories using the Rasch measurement model, which would confirm that the instrument is generic. Methods: HAQ-II responses from 882 consecutive rheumatology clinic attendees were fitted to a Rasch model. Invariance across disease was assessed by analysis of variance of residuals implemented in RUMM2030. Rasch modeled HAQ-II scores across disease categories were compared and the mathematical relationship between raw HAQ-II scores and Rasch modeled scores was also determined. Results: The HAQ-II responses fitted the Rasch model. There was no substantive evidence for lack of invariance by disease category except for a single item (“opening car doors”). Rasch modeled scores could be accurately obtained from raw scores with a cubic formula (R 0.99). Patients with rheumatoid arthritis had more disability than patients with other kinds of inflammatory arthritis or autoimmune connective tissue disease. Conclusions: The HAQ-II can be used across different rheumatic diseases and scores can be similarly interpreted from patients with different diseases. Transforming raw scores to Rasch modeled scores enable a strictly linear, interval scale to be used. It remains to be seen how that would affect interpretation of change scores. Trial registration: ANZCTR ACTRN12617001500347. Registered 24th October 2017 (retrospectively registered). Keywords: Health assessment questionnaire, Psychometric properties, Disability Background disease is ‘activity limitations’. Activity limitations refer to According to the World Health Organisation (WHO) difficulties with day to day activities such as walking, talk- International Classification of Functioning, Health and ing, housework or self-care (for example). Activity limita- Disability (ICF), the effects of disease or injury are prin- tions are typically considered at the individual-level of cipally manifest as deficits of functioning [1]. Different functioning (that is, without reference to social context). aspects of functioning have been conceptualized within The WHO defines ‘Activity’ as ‘the execution of a task or the ICF model [2]. One aspect of functioning which is action by an individual’, which may interact with other intrinsically important to most people with rheumatic components of the ICF model including Environmental Factors that ‘make up the physical, social and attitudinal environment in which people live and conduct their lives’. * Correspondence: will.taylor@otago.ac.nz While activity limitations may be both influenced by and Department of Medicine, University of Otago Wellington, PO Box 7343, influence social context, for conceptual clarity and meas- Wellington, New Zealand Wellington Regional Rheumatology Unit, Hutt Valley District Health Board, urement, activity limitations are considered separate con- Wellington, New Zealand cepts from social context [3]. One important category of Full list of author information is available at the end of the article © The Author(s). 2018 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. Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 2 of 7 activity limitations concerns physical activities, which is The information is mainly used to inform point of care the typical concern of measures of so-called ‘physical clinical decision making. Data were obtained from 1000 functioning’ in the rheumatology literature. consecutive patient visits over 24 months and were ‘Physical functioning’ or ‘disability’ or a similar con- previously reported in an analysis of the PAS-II instru- cept has been endorsed by the Outcome Measures in ment [15]. Rheumatology Clinical Trials (OMERACT) group as a The HAQ-II is a 10-item version of the original core domain for outcome studies in every rheumatic HAQ-DI, with some new items to extend the range of disease it has considered [4–8]. While there are some assessed disability and was derived by fit to a Rasch disease-specific measures of physical functioning in measurement model [17]. Each item is rated on a rheumatology, the most commonly used instrument is 4-point scale (no difficulty, some difficulty, much diffi- the Health Assessment Questionnaire (HAQ) Disability culty, unable to do) and averaged over the number of Index and variants [9]. There are several advantages in answered items (must be at least 7) to obtain a total raw using the same instrument across different diseases [10]. score that can range from 0 to 3 (least to most disabled). In particular, direct comparisons can be made with The disease diagnoses were divided into 5 diagnostic regard to the severity of the functional deficit, which is categories (rheumatoid arthritis (RA), other inflamma- more difficult when disease-specific instruments are tory arthritis, auto-immune connective tissue diseases, used. It is likely that computer-adaptive testing (CAT) non-inflammatory disorders and others). “Other inflam- will be even better [11], but in most clinical situations matory arthritis” consisted of ankylosing spondylitis, that technology is not easily available [12]. psoriatic arthritis, gout and undifferentiated inflamma- In addition, versions of HAQ scores are one of the 3 tory arthritis. “Non-inflammatory disorders” consisted of components of the Routine Assessment of Patient Index regional pain syndromes, osteoarthritis and fibromyalgia Data 3 (RAPID3) [13] or Patient Activity Scale (PAS, syndrome. “Autoimmune connective tissue” diseases in- PAS-II) [14] which can be a useful monitor of health sta- cluded SLE, systemic sclerosis and undifferentiated con- tus in the clinic situation. The other two components nective tissue diseases. “Others” included conditions such are pain and global assessment of health status/disease polymyalgic rheumatica, inflammatory myositis, Sjogren’s activity. Treatment targets and thresholds for low dis- syndrome, Behcet’s disease, and plantar fasciitis. ease activity or remission have been identified for these Data were fitted to a polytomous unrestricted indices in rheumatoid arthritis. Since the three compo- partial-credit Rasch model using RUMM2030 software nents of these indices are potentially applicable to any [18]. The Rasch model is mathematically expressed disease where pain and functional deficit are key mani- below, and essentially means that the probability of any festations, it is possible that they may be generic [15]. particular response (x where X = x {0,1, …, m } associ- ni i For this to be the case, it would be helpful to confirm ated with the m + 1 successive category of item i)on that the HAQ instrument is also generic. We chose to any item i, is a function of the ‘ability’ (amount of trait, evaluate the HAQ-II variant of HAQ since it is shorter β ) of the person n and the ‘difficulty’ (amount of trait, than the original HAQ-DI (10 items versus 20 items) δ ) of the item. The thresholds between each of the m + i i and was developed using Rasch methodology, which 1 categories of each item are denoted by τ and γ is a ki ni may imply better psychometric properties. normalizing factor. Some authors claim that only the The objective of this study was to demonstrate, using Rasch model fulfils the axioms of fundamental measure- the Rasch measurement model [16], that the HAQ-II in- ment [19, 20]. strument was invariant across disease categories. That is, people with different diseases answer the items in the Prfg X ¼ x ¼ exp x β −δ − τ =γ ni i ki n ni same way (dependent only on their level of function) so k¼0 that scores can be interpreted in the same way. For example, a score of 2 for a person with rheumatoid Overall model fit was assessed using an item-trait arthritis (RA) will mean the same level of disability as a interaction chi-square statistic and Root Mean Square score of 2 for a person with systemic lupus erythemato- Error of Approximation (RMSEA) [21]. As reported by sus (SLE). Tennant and Pallant, large samples (N > 500), can lead to statistically significant chi-square tests without sub- Methods stantive misfit in simulated datasets, so we followed the All patients attending the rheumatology outpatient procedure suggested by Tennant and Pallant, by ran- clinics at the Wellington Regional Rheumatology Unit domly selecting five subsets of 500 participants and routinely complete a questionnaire, which consists of fitting these data to the Rasch model independently; and the Health Assessment Questionnaire-II (HAQ-II), by using the RMSEA index from the whole sample. 10 cm VAS for ‘pain’ and 10 cm VAS for ‘patient global’. RMSEA is a model fit index less likely than the Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 3 of 7 chi-square test to be affected by large samples. A value (F = 0.25) with a total sample of 882, given a (conservative) of 0.02 or less was accepted as indicating adequate Bonferroni corrected critical p-value of 0.0017, 5 categor- model fit [21]. ies of disease and 10 classes of scale location. This yielded Measurement precision was assessed using the a power of 84%. There are multiple approaches to deter- Person-Separation-Index (PSI), which can be interpreted mining DIF, but generally different methods have been in a similar way to Cronbach’s alpha. A PSI of 0.7 means shown to lead to similar findings [27]. that the score can distinguish between 2 strata of The distribution of HAQ-II scores and the relation- person-ability whereas a value of 0.9 suggests 4 distinct ship between raw HAQ-II scores and Rasch modelled groups of person-ability can be identified [22]. scores was assessed using SPSS v24. Rasch-modelled Individual item fit to the Rasch model was assessed scores were re-scaled to be between 0 and 3 (the raw with an item-trait interaction chi-square statistic and a score range) for ease of interpretation by clinicians normalized item-person interaction fit residual. A familiar with HAQ scores. This was accomplished Bonferroni-corrected p-value of less than 0.05 was using a linear transformation according to the rescal- taken to indicate misfit for the chi-square test; fit resid- ing formula below, where the range of the Rasch uals of greater than 2.5 are taken to indicate poor score was observed to be − 5.97 to 4.91 and the range discrimination of the item and fit residuals of less than of the rescaled score was 0 to 3. − 2.5 are taken to indicate excessive good discrimin- max − min ation (overfit). Unidimensionality was assessed by the rescaled rescaled ðÞ value− max Rasch proportion of independent t-tests of person estimates max − min Rasch Rasch þ max derived from contrasting sets of items (selected on the rescaled basis of positive or negative loading on the first factor of a principal components analysis of residuals) that Ethical approval was granted by the New Zealand were significant at the 0.05 level. Where fewer than 5% Health and Disability Ethics Committee without full of t-tests are significant at p<0.05, thedataissupport- review as part of its standing procedures for observa- ive of unidimensionality [23]. tional, low risk studies. The study was retrospectively For each item, invariance by disease category was registered with the Australian New Zealand Clinical Tri- assessed by a 2-way analysis of variance (ANOVA) of the als Registry (ACTRN12617001500347). standardized residuals for individuals grouped into 10 classes based on their Rasch-modeled latent trait (phys- Results ical disability) and the 5 disease categories [24, 25]. A From the 1000 consecutive patient visits over 24 months, statistically significant F-value for the disease factor we selected 882 unique patients with their first visit indicates a main-effect of disease category on fit to the during the observation period (since some patients Rasch model that is independent of the location of the visited more than once). About one third of all patient person on the latent trait. This is known as ‘uniform visits had rheumatoid arthritis (RA) (Table 1). Fitting the DIF’. A statistically significant F-value for the interaction data to the Rasch model led to an overall chi-square of between disease category and scale location indicates that 122 (df 90), p = 0.013 and RMSEA 0.02. The PSI was people with different diseases fit the Rasch model differ- 0.89, indicating approximately 4 distinct strata of ently depending on where they are on the latent trait. This person-ability can be distinguished with the HAQ-II. is known as ‘non-uniform DIF’. A Bonferroni-corrected Unidimensionality was confirmed using the equating p-value was used to account for multiple hypothesis test- t-tests procedure implemented in RUMM (3.78% of ing. The sample size calculation for a 2-way ANOVA with t-tests were significant at the 5% level). Each of the five 5 categories of disease and 10 classes of scale location is randomly selected subsets of 500 individuals showed somewhat complex; we used a post-hoc estimation of overall model chi-square p-value > 0.05, confirming that power in G*Power [26] to detect a medium effect the data fit the Rasch model. Table 1 Participant characteristics Diagnostic group N Percent female Age, years (mean, SD) Rheumatoid arthritis 342 72 60.4 (13.5) Other inflammatory arthritis 304 47 50.1 (14.9) Non-inflammatory disorder 84 87 56.6 (12.8) Auto-immune connective tissue disease 125 91 47.8 (16.1) Other 145 83 59.4 (18.6) Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 4 of 7 Table 2 Item location and fit statistics Item Location (SE) in logits Residual DF ChiSq (9 df) p-value 1 Getting on and off the toilet 2.30 (0.08) 0.16 681.78 13.67 0.13 2 Open car doors 2.00 (0.08) 0.14 680.89 5.87 0.75 3 Stand up from straight chair 0.89 (0.07) 0.37 684.47 12.27 0.20 4 Walk outdoors on flat ground 0.91 (0.07) 1.78 683.57 13.33 0.15 5 Wait in line for 15 min −0.25 (0.06) 1.88 675.52 11.62 0.24 6 Reach and get down a 5 lb. object −0.33 (0.06) 0.11 679.10 6.49 0.69 7 Go up 2 or more flights of stairs −0.44 (0.06) −2.67 675.52 18.72 0.03 8 Do outside work −1.21 (0.06) − 1.61 676.42 18.22 0.03 9 Lift heavy objects −1.83 (0.06) −4.02 687.15 14.99 0.09 10 Move heavy objects −2.05 (0.06) − 2.21 683.57 7.26 0.61 Individual item fit is shown in Table 2.While no HAQII Rasch score ¼ 0:05 þ HAQII  2:12−HAQII item demonstrated evidence of misfit at the Bonferro- 3 1:06 þ HAQII  0:24 ni-corrected p-value, 2 items showed evidence of overfit with fits residuals of less than − 2.5. Differential item functioning analysis is displayed in The distribution of Rasch modeled scores by disease cat- Table 3. One item (opening car doors) suggested egory is shown in Fig. 3. One way analysis of variance invariance was not present at a p-value close to the showed that there was a significant difference between the Bonferroni-corrected level of significance. Inspection of disease categories (F(4,877) = 6.46, p < 0.001). Post-hoc the item-characteristic curve suggested that mostly the tests using RA as the reference disease category showed ICC for each disease group overlapped, but patients with that RA patients have slightly more disability than patients RA found this item harder than other disease groups, with other inflammatory arthritis with a mean difference especially for higher levels of disability (to the right of 0.17 (95% CI 0.04 to 0.30, p = 0.004) and more disability the logit scale) (Fig. 1). However, there was no significant than patients with autoimmune connective tissue disor- DIF for any item observed in any of the five randomly ders with a mean difference of 0.24 (95% CI 0.07 to 0.42, selected samples of 500 individuals. p = 0.002). There were no differences in disability between A transformation from a raw HAQ-II score to a RA and the other two disease categories. Rasch modeled score (rescaled to also range from 0 to 3), but which is now strictly linear, was accom- Discussion plished by fitting a cubic equation to the relationship This study has shown that the HAQ-II instrument can be between the raw HAQ-II score and the Rasch mod- considered psychometrically generic amongst rheumatol- eled score (Fig. 2). This equation has an R of 0.99. ogy clinic patients. It shows minimal invariance for disease Table 3 ANOVA for Differential Item Functioning by Disease (item in bold suggests possible DIF at the Bonferroni-corrected level of 0.0017) Item Class Interval Disease Class Interval x Disease Total MS F (df 9) p MS F (df 4) p MS F (df 36) p MS F (df 40) p 1 1.42 1.61 0.108 0.74 0.84 0.498 1.46 1.65 0.010 55.53 1.573 0.015 2 0.56 0.62 0.778 4.17 4.61 0.001 0.85 0.93 0.573 47.25 1.307 0.101 3 1.31 1.42 0.173 1.61 1.75 0.137 0.69 0.74 0.859 31.26 0.849 0.735 4 1.43 1.38 0.192 3.04 2.94 0.019 0.82 0.78 0.809 41.52 1.005 0.465 5 1.5 1.47 0.152 2.6 2.55 0.038 0.54 0.53 0.989 29.88 0.734 0.889 6 0.68 0.75 0.660 1.13 1.25 0.287 0.94 1.03 0.409 38.24 1.06 0.373 7 2.01 2.7 0.004 0.17 0.22 0.926 0.85 1.13 0.269 31.22 1.046 0.395 8 1.91 2.43 0.010 0.65 0.83 0.506 1.06 1.34 0.087 40.76 1.296 0.108 9 1.66 2.38 0.012 2.4 3.43 0.009 0.33 0.47 0.996 21.57 0.773 0.843 10 0.82 1.04 0.402 1.69 2.14 0.074 0.45 0.56 0.981 22.81 0.726 0.896 Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 5 of 7 Fig. 1 The item-characteristic curve (ICC) for item 2 (opening car doors). This plots the expected response to item 2 based on the individuals’ level of disability (person location). The curves for each disease category are superimposed upon the Rasch model (gray line). DIF would be implied by a significantly different location of a disease-specific ICC. RA (rheumatoid arthritis), IA (inflammatory arthritis), INF (inflammatory disorder), AICTD (autoimmune connective tissue disease) category, which implies that responses to each item and can be reasonably incorporated into the PAS-II score for the total score can be interpreted in just the same way for patients with any rheumatic disease to produce meaning- these disease categories. Therefore, it is valid to directly ful and comparable scores. RAPID3 uses a different ver- compare physical disability between diseases, and it was sion of HAQ, which will require a similar analysis to found that patients with RA have slightly more disability confirm invariance by disease category. on average than patients with two other disease categories. We have also described a transformation of the raw The results make the HAQ-II instrument a useful indica- HAQ-II score that may be useful for aggregated data ana- tor of physical functioning in a general rheumatology lysis in audit or clinical research, since it is strictly linear clinic, where patients with several different diseases may on an interval scale, making it very suitable for parametric come for treatment. Furthermore, the HAQ-II instrument statistical analysis and mathematical manipulation. Fig. 2 The relationship between Rasch modeled scores and raw HAQ-II scores closely fits a cubic equation Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 6 of 7 Fig. 3 The distribution of Rasch-modeled HAQ-II scores by disease category The meaning of changes in HAQ scores within indi- Conclusions viduals or between groups is highly dependent upon the The HAQ-II instrument has good psychometric proper- linearity of the scale. A non-linear scale makes it very ties including invariance by disease, suggesting that the difficult to compare changes at different starting points measure can be used with confidence in general on the scale, as has been shown for the 10 cm Pain vis- rheumatology clinics. Although theoretically attractive, it ual analogue scale [28]. The conventional minimal im- is not yet clear whether transformation of raw scores to portant difference (MCID) for HAQ-DI in RA is 0.20 to a Rasch-modelled score confers practical advantages. 0.22 [29] but may be larger [30]. For HAQ-II, its authors suggest MCID of 0.34. However, MCID assume a linear Abbreviations ANOVA: Analysis of variance; DIF: Differential item functioning; HAQ: Health scale, which is clearly not the case for the raw scores. Assessment Questionnaire; ICF: International Classification of Health, More meaningful values of MCID should be directly de- Functioning and Disability; MCID: Minimal clinically important difference; termined using Rasch-modelled scores compared to pa- OMERACT: Outcome Measures in Rheumatology Clinical Trials; PAS: Patient Activity Scale; RA: Rheumatoid arthritis; RAPID3: Routine Assessment of tient perception of change. Patient Index Data 3; RMSEA: Root mean square error of approximation; The main limitation of this study is the semi-arbitrary RUMM: Rasch Unidimensional Measurement Models; SLE: Systemic Lupus way by which rheumatic diseases were grouped together. Erythematosus; SPSS: Statistics Package for the Social Sciences; VAS: Visual Analogue Scale; WHO: World Health Organisation It is possible that more distinct diseases may show dif- ferential item functioning which is not apparent when Funding two or more diseases are grouped together. On the other This work received no specific funding but was supported by the Hutt Valley hand, grouping similar diseases together may increase District Health Board and the University of Otago. the statistical power to show differences, although this assumes that the within-group diseases associate with Availability of data and materials physical functioning in a similar way. In addition, there The dataset used and analysed during the current study are available from the corresponding author on reasonable request. is some functional heterogeneity within some relatively defined diseases such as systemic lupus erythematosus and psoriatic arthritis. Overall, it is unclear whether a Authors’ contributions WJT conceived and designed the study, analysed the data and wrote the different approach to grouping diseases would have manuscript. KP designed the study, collected the data and critically reviewed yielded different results, and could be an avenue for fur- the manuscript. Both authors authorized submission of the manuscript for ther testing.. publication. Both authors read and approved the final manuscript. Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 7 of 7 Ethics approval and consent to participate 17. Wolfe F, Michaud K, Pincus T. Development and validation of the health Ethical approval was granted by the New Zealand Health and Disability Ethics assessment questionnaire II: a revised version of the health assessment Committee without full review as part of its standing procedures for questionnaire. Arthritis Rheum. 2004;50:3296–305. observational, low risk studies. 18. Andrich D, Sheridan B, Luo G. RUMM2030: Rasch Unidimensional Models for Measurement. Perth: RUMM Laboratory; 1997–2012. Competing interests 19. Boone WJ, Staver JR, Yale MS. The Rasch model and item response theory The authors declare that they have no competing interests. models: identical, similar, or unique? In: Rasch analysis in the human sciences. Dordrecht: Springer; 2014. 20. Perline R, Wright BD, Wainer H. The Rasch model as additive conjoint Publisher’sNote measurement. Appl Psychol Meas. 1979;3:237–55. Springer Nature remains neutral with regard to jurisdictional claims in published 21. Alan Tennant PJF. The root mean square error of approximation (RMSEA) as maps and institutional affiliations. a supplementary statistic to determine fit to the Rasch model with large sample sizes. Rasch Measurement Transactions. 2012;25:1348–9. Author details 22. Jr WF. Reliability Statistics. Rasch Measurement Transactions. 1992;6:238. Department of Medicine, University of Otago Wellington, PO Box 7343, 23. Smith EV. Detecting and evaluation the impact of multidimensionality using Wellington, New Zealand. Wellington Regional Rheumatology Unit, Hutt item fit statistics and principal component analysis of residuals. J Appl Meas. Valley District Health Board, Wellington, New Zealand. General Medicine 2002;3:205–31. Service, Capital and Coast District Health Board, Wellington, New Zealand. 24. Andrich D, Hagquist C. Real and artificial differential item functioning in Polytomous items. Educ Psychol Meas. 2014;75:185–207. Received: 24 October 2017 Accepted: 21 May 2018 25. Hagquist C, Andrich D. Is the sense of coherence-instrument applicable on adolescents? A latent trait analysis using Rasch-modelling. Personal Individ Differ. 2004;36:955–68. References 26. Erdfelder E, Faul F, Buchner A, GPOWER. A general power analysis program. 1. Kostanjsek N, Rubinelli S, Escorpizo R, Cieza A, Kennedy C, Selb M, Stucki G, Behavior Research Methods, Instruments, & Computers. 1996;28:1–11. Ustun TB. Assessing the impact of health conditions using the ICF. Disabil 27. Tang K, Canadian Arthritis Network work productivity G. Disease-related Rehabil. 2011;33:1475–82. differential item functioning in the work instability scale for rheumatoid 2. World Health Organization. International classification of functioning, arthritis: converging results from three methods. Arthritis care & research. Disability and Health: ICF. Geneva: WHO; 2001. 2011;63:1159–69. 3. Taylor WJ, Geyh S. A rehabilitation framework: the international classification 28. Kersten P, White PJ, Tennant A. Is the pain visual analogue scale linear of functioning, disability and health. In: Dean SG, Siegert RJ, Taylor WJ, and responsive to change? An exploration using Rasch analysis. PLoS editors. Interprofessional Rehabilitation: A Person-Centred Approach. One. 2014;9:e99485. Chichester: Wiley-Blackwell; 2012. 29. Wells GA, Tugwell P, Kraag GR, Baker PRA, Groh J, Redelmeier DA. Minimum 4. Tugwell P, Boers M. Developing consensus on preliminary core efficacy important difference between patients with rheumatoid-arthritis - the endpoints for rheumatoid arthritis clinical trials. OMERACT Committee. J patients perspective. J Rheumatol. 1993;20:557–60. Rheumatol. 1993;20:555–6. 30. Ward MM, Guthrie LC, Alba MI. Clinically important changes in individual 5. Bellamy N, Kirwan J, Boers M, Brooks P, Strand V, Tugwell P, Altman R, Brandt K, and composite measures of rheumatoid arthritis activity: thresholds Dougados M, Lequesne M. Recommendations for a core set of outcome applicable in clinical trials. Ann Rheum Dis. 2015;74:1691–6. measures for future phase III clinical trials in knee, hip, and hand osteoarthritis. Consensus development at OMERACT III. J Rheumatol. 1997;24:799–802. 6. van der Heijde D, van der Linden S, Bellamy N, Calin A, Dougados M, Khan MA. Which domains should be included in a core set for endpoints in ankylosing spondylitis? Introduction to the ankylosing spondylitis module of OMERACT IV. J Rheumatol. 1999;26:945–7. 7. Merkel PA, Aydin SZ, Boers M, Direskeneli H, Herlyn K, Seo P, Suppiah R, Tomasson G, Luqmani RA. The OMERACT core set of outcome measures for use in clinical trials of ANCA-associated vasculitis. J Rheumatol. 2011;38:1480–6. 8. Schumacher HR Jr, Taylor W, Edwards NL, Grainger R, Schlesinger N, Dalbeth N, Sivera F, Singh JA, Evans R, Waltrip RW, et al. Outcome domains for studies of acute and chronic gout. J Rheumatol. 2009;36:2342–5. 9. Bruce B, Fries JF. The health assessment questionnaire (HAQ). Clin Exp Rheumatol. 2005;23:S14–8. 10. McDowell I, Newell C. Measuring Health. A guide to rating scales and questionnaires. Second edn. New York: Oxford University Press; 1996. 11. Fries JF, Bruce B, Cella D. The promise of PROMIS: using item response theory to improve assessment of patient-reported outcomes. Clin Exp Rheumatol. 2005;23:S53–7. 12. Unick GJ, Shumway M, Hargreaves W. Are we ready for computerized adaptive testing? Psychiatric services (Washington, DC). 2008;59:369. 13. Pincus T, Swearingen CJ, Bergman M, Yazici Y. RAPID3 (routine assessment of patient index data 3), a rheumatoid arthritis index without formal joint counts for routine care: proposed severity categories compared to disease activity score and clinical disease activity index categories. J Rheumatol. 2008;35:2136–47. 14. Wolfe F, Michaud K, Pincus T. A compositie disease activity scale for clinical practice, observational studies and clinical trials: the patient activity scale (PAS/PAS-II). J Rheumatol. 2005;32:2410–5. 15. Parekh K, Taylor WJ. The patient activity scale-II (PAS-II) is a generic indicator of active disease in patients with rheumatic disorders. J Rheumatol. 2010;37:1932–4. 16. Tennant A, Conaghan PG. The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis & Rheumatism. 2007;57:1358–62. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Health and Quality of Life Outcomes Springer Journals

Rasch analysis suggests that health assessment questionnaire II is a generic measure of physical functioning for rheumatic diseases: a cross-sectional study

Free
7 pages

Loading next page...
 
/lp/springer_journal/rasch-analysis-suggests-that-health-assessment-questionnaire-ii-is-a-M89wHfIda9
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Quality of Life Research; Quality of Life Research
eISSN
1477-7525
D.O.I.
10.1186/s12955-018-0939-2
Publisher site
See Article on Publisher Site

Abstract

Background: Versions of the Health Assessment Questionnaire (HAQ) are commonly used to measure physical functioning across multiple rheumatic diseases but there has been no clear demonstration that any HAQ version is actually generic. This study aimed to show that the HAQ-II instrument is invariant across different rheumatic disease categories using the Rasch measurement model, which would confirm that the instrument is generic. Methods: HAQ-II responses from 882 consecutive rheumatology clinic attendees were fitted to a Rasch model. Invariance across disease was assessed by analysis of variance of residuals implemented in RUMM2030. Rasch modeled HAQ-II scores across disease categories were compared and the mathematical relationship between raw HAQ-II scores and Rasch modeled scores was also determined. Results: The HAQ-II responses fitted the Rasch model. There was no substantive evidence for lack of invariance by disease category except for a single item (“opening car doors”). Rasch modeled scores could be accurately obtained from raw scores with a cubic formula (R 0.99). Patients with rheumatoid arthritis had more disability than patients with other kinds of inflammatory arthritis or autoimmune connective tissue disease. Conclusions: The HAQ-II can be used across different rheumatic diseases and scores can be similarly interpreted from patients with different diseases. Transforming raw scores to Rasch modeled scores enable a strictly linear, interval scale to be used. It remains to be seen how that would affect interpretation of change scores. Trial registration: ANZCTR ACTRN12617001500347. Registered 24th October 2017 (retrospectively registered). Keywords: Health assessment questionnaire, Psychometric properties, Disability Background disease is ‘activity limitations’. Activity limitations refer to According to the World Health Organisation (WHO) difficulties with day to day activities such as walking, talk- International Classification of Functioning, Health and ing, housework or self-care (for example). Activity limita- Disability (ICF), the effects of disease or injury are prin- tions are typically considered at the individual-level of cipally manifest as deficits of functioning [1]. Different functioning (that is, without reference to social context). aspects of functioning have been conceptualized within The WHO defines ‘Activity’ as ‘the execution of a task or the ICF model [2]. One aspect of functioning which is action by an individual’, which may interact with other intrinsically important to most people with rheumatic components of the ICF model including Environmental Factors that ‘make up the physical, social and attitudinal environment in which people live and conduct their lives’. * Correspondence: will.taylor@otago.ac.nz While activity limitations may be both influenced by and Department of Medicine, University of Otago Wellington, PO Box 7343, influence social context, for conceptual clarity and meas- Wellington, New Zealand Wellington Regional Rheumatology Unit, Hutt Valley District Health Board, urement, activity limitations are considered separate con- Wellington, New Zealand cepts from social context [3]. One important category of Full list of author information is available at the end of the article © The Author(s). 2018 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. Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 2 of 7 activity limitations concerns physical activities, which is The information is mainly used to inform point of care the typical concern of measures of so-called ‘physical clinical decision making. Data were obtained from 1000 functioning’ in the rheumatology literature. consecutive patient visits over 24 months and were ‘Physical functioning’ or ‘disability’ or a similar con- previously reported in an analysis of the PAS-II instru- cept has been endorsed by the Outcome Measures in ment [15]. Rheumatology Clinical Trials (OMERACT) group as a The HAQ-II is a 10-item version of the original core domain for outcome studies in every rheumatic HAQ-DI, with some new items to extend the range of disease it has considered [4–8]. While there are some assessed disability and was derived by fit to a Rasch disease-specific measures of physical functioning in measurement model [17]. Each item is rated on a rheumatology, the most commonly used instrument is 4-point scale (no difficulty, some difficulty, much diffi- the Health Assessment Questionnaire (HAQ) Disability culty, unable to do) and averaged over the number of Index and variants [9]. There are several advantages in answered items (must be at least 7) to obtain a total raw using the same instrument across different diseases [10]. score that can range from 0 to 3 (least to most disabled). In particular, direct comparisons can be made with The disease diagnoses were divided into 5 diagnostic regard to the severity of the functional deficit, which is categories (rheumatoid arthritis (RA), other inflamma- more difficult when disease-specific instruments are tory arthritis, auto-immune connective tissue diseases, used. It is likely that computer-adaptive testing (CAT) non-inflammatory disorders and others). “Other inflam- will be even better [11], but in most clinical situations matory arthritis” consisted of ankylosing spondylitis, that technology is not easily available [12]. psoriatic arthritis, gout and undifferentiated inflamma- In addition, versions of HAQ scores are one of the 3 tory arthritis. “Non-inflammatory disorders” consisted of components of the Routine Assessment of Patient Index regional pain syndromes, osteoarthritis and fibromyalgia Data 3 (RAPID3) [13] or Patient Activity Scale (PAS, syndrome. “Autoimmune connective tissue” diseases in- PAS-II) [14] which can be a useful monitor of health sta- cluded SLE, systemic sclerosis and undifferentiated con- tus in the clinic situation. The other two components nective tissue diseases. “Others” included conditions such are pain and global assessment of health status/disease polymyalgic rheumatica, inflammatory myositis, Sjogren’s activity. Treatment targets and thresholds for low dis- syndrome, Behcet’s disease, and plantar fasciitis. ease activity or remission have been identified for these Data were fitted to a polytomous unrestricted indices in rheumatoid arthritis. Since the three compo- partial-credit Rasch model using RUMM2030 software nents of these indices are potentially applicable to any [18]. The Rasch model is mathematically expressed disease where pain and functional deficit are key mani- below, and essentially means that the probability of any festations, it is possible that they may be generic [15]. particular response (x where X = x {0,1, …, m } associ- ni i For this to be the case, it would be helpful to confirm ated with the m + 1 successive category of item i)on that the HAQ instrument is also generic. We chose to any item i, is a function of the ‘ability’ (amount of trait, evaluate the HAQ-II variant of HAQ since it is shorter β ) of the person n and the ‘difficulty’ (amount of trait, than the original HAQ-DI (10 items versus 20 items) δ ) of the item. The thresholds between each of the m + i i and was developed using Rasch methodology, which 1 categories of each item are denoted by τ and γ is a ki ni may imply better psychometric properties. normalizing factor. Some authors claim that only the The objective of this study was to demonstrate, using Rasch model fulfils the axioms of fundamental measure- the Rasch measurement model [16], that the HAQ-II in- ment [19, 20]. strument was invariant across disease categories. That is, people with different diseases answer the items in the Prfg X ¼ x ¼ exp x β −δ − τ =γ ni i ki n ni same way (dependent only on their level of function) so k¼0 that scores can be interpreted in the same way. For example, a score of 2 for a person with rheumatoid Overall model fit was assessed using an item-trait arthritis (RA) will mean the same level of disability as a interaction chi-square statistic and Root Mean Square score of 2 for a person with systemic lupus erythemato- Error of Approximation (RMSEA) [21]. As reported by sus (SLE). Tennant and Pallant, large samples (N > 500), can lead to statistically significant chi-square tests without sub- Methods stantive misfit in simulated datasets, so we followed the All patients attending the rheumatology outpatient procedure suggested by Tennant and Pallant, by ran- clinics at the Wellington Regional Rheumatology Unit domly selecting five subsets of 500 participants and routinely complete a questionnaire, which consists of fitting these data to the Rasch model independently; and the Health Assessment Questionnaire-II (HAQ-II), by using the RMSEA index from the whole sample. 10 cm VAS for ‘pain’ and 10 cm VAS for ‘patient global’. RMSEA is a model fit index less likely than the Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 3 of 7 chi-square test to be affected by large samples. A value (F = 0.25) with a total sample of 882, given a (conservative) of 0.02 or less was accepted as indicating adequate Bonferroni corrected critical p-value of 0.0017, 5 categor- model fit [21]. ies of disease and 10 classes of scale location. This yielded Measurement precision was assessed using the a power of 84%. There are multiple approaches to deter- Person-Separation-Index (PSI), which can be interpreted mining DIF, but generally different methods have been in a similar way to Cronbach’s alpha. A PSI of 0.7 means shown to lead to similar findings [27]. that the score can distinguish between 2 strata of The distribution of HAQ-II scores and the relation- person-ability whereas a value of 0.9 suggests 4 distinct ship between raw HAQ-II scores and Rasch modelled groups of person-ability can be identified [22]. scores was assessed using SPSS v24. Rasch-modelled Individual item fit to the Rasch model was assessed scores were re-scaled to be between 0 and 3 (the raw with an item-trait interaction chi-square statistic and a score range) for ease of interpretation by clinicians normalized item-person interaction fit residual. A familiar with HAQ scores. This was accomplished Bonferroni-corrected p-value of less than 0.05 was using a linear transformation according to the rescal- taken to indicate misfit for the chi-square test; fit resid- ing formula below, where the range of the Rasch uals of greater than 2.5 are taken to indicate poor score was observed to be − 5.97 to 4.91 and the range discrimination of the item and fit residuals of less than of the rescaled score was 0 to 3. − 2.5 are taken to indicate excessive good discrimin- max − min ation (overfit). Unidimensionality was assessed by the rescaled rescaled ðÞ value− max Rasch proportion of independent t-tests of person estimates max − min Rasch Rasch þ max derived from contrasting sets of items (selected on the rescaled basis of positive or negative loading on the first factor of a principal components analysis of residuals) that Ethical approval was granted by the New Zealand were significant at the 0.05 level. Where fewer than 5% Health and Disability Ethics Committee without full of t-tests are significant at p<0.05, thedataissupport- review as part of its standing procedures for observa- ive of unidimensionality [23]. tional, low risk studies. The study was retrospectively For each item, invariance by disease category was registered with the Australian New Zealand Clinical Tri- assessed by a 2-way analysis of variance (ANOVA) of the als Registry (ACTRN12617001500347). standardized residuals for individuals grouped into 10 classes based on their Rasch-modeled latent trait (phys- Results ical disability) and the 5 disease categories [24, 25]. A From the 1000 consecutive patient visits over 24 months, statistically significant F-value for the disease factor we selected 882 unique patients with their first visit indicates a main-effect of disease category on fit to the during the observation period (since some patients Rasch model that is independent of the location of the visited more than once). About one third of all patient person on the latent trait. This is known as ‘uniform visits had rheumatoid arthritis (RA) (Table 1). Fitting the DIF’. A statistically significant F-value for the interaction data to the Rasch model led to an overall chi-square of between disease category and scale location indicates that 122 (df 90), p = 0.013 and RMSEA 0.02. The PSI was people with different diseases fit the Rasch model differ- 0.89, indicating approximately 4 distinct strata of ently depending on where they are on the latent trait. This person-ability can be distinguished with the HAQ-II. is known as ‘non-uniform DIF’. A Bonferroni-corrected Unidimensionality was confirmed using the equating p-value was used to account for multiple hypothesis test- t-tests procedure implemented in RUMM (3.78% of ing. The sample size calculation for a 2-way ANOVA with t-tests were significant at the 5% level). Each of the five 5 categories of disease and 10 classes of scale location is randomly selected subsets of 500 individuals showed somewhat complex; we used a post-hoc estimation of overall model chi-square p-value > 0.05, confirming that power in G*Power [26] to detect a medium effect the data fit the Rasch model. Table 1 Participant characteristics Diagnostic group N Percent female Age, years (mean, SD) Rheumatoid arthritis 342 72 60.4 (13.5) Other inflammatory arthritis 304 47 50.1 (14.9) Non-inflammatory disorder 84 87 56.6 (12.8) Auto-immune connective tissue disease 125 91 47.8 (16.1) Other 145 83 59.4 (18.6) Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 4 of 7 Table 2 Item location and fit statistics Item Location (SE) in logits Residual DF ChiSq (9 df) p-value 1 Getting on and off the toilet 2.30 (0.08) 0.16 681.78 13.67 0.13 2 Open car doors 2.00 (0.08) 0.14 680.89 5.87 0.75 3 Stand up from straight chair 0.89 (0.07) 0.37 684.47 12.27 0.20 4 Walk outdoors on flat ground 0.91 (0.07) 1.78 683.57 13.33 0.15 5 Wait in line for 15 min −0.25 (0.06) 1.88 675.52 11.62 0.24 6 Reach and get down a 5 lb. object −0.33 (0.06) 0.11 679.10 6.49 0.69 7 Go up 2 or more flights of stairs −0.44 (0.06) −2.67 675.52 18.72 0.03 8 Do outside work −1.21 (0.06) − 1.61 676.42 18.22 0.03 9 Lift heavy objects −1.83 (0.06) −4.02 687.15 14.99 0.09 10 Move heavy objects −2.05 (0.06) − 2.21 683.57 7.26 0.61 Individual item fit is shown in Table 2.While no HAQII Rasch score ¼ 0:05 þ HAQII  2:12−HAQII item demonstrated evidence of misfit at the Bonferro- 3 1:06 þ HAQII  0:24 ni-corrected p-value, 2 items showed evidence of overfit with fits residuals of less than − 2.5. Differential item functioning analysis is displayed in The distribution of Rasch modeled scores by disease cat- Table 3. One item (opening car doors) suggested egory is shown in Fig. 3. One way analysis of variance invariance was not present at a p-value close to the showed that there was a significant difference between the Bonferroni-corrected level of significance. Inspection of disease categories (F(4,877) = 6.46, p < 0.001). Post-hoc the item-characteristic curve suggested that mostly the tests using RA as the reference disease category showed ICC for each disease group overlapped, but patients with that RA patients have slightly more disability than patients RA found this item harder than other disease groups, with other inflammatory arthritis with a mean difference especially for higher levels of disability (to the right of 0.17 (95% CI 0.04 to 0.30, p = 0.004) and more disability the logit scale) (Fig. 1). However, there was no significant than patients with autoimmune connective tissue disor- DIF for any item observed in any of the five randomly ders with a mean difference of 0.24 (95% CI 0.07 to 0.42, selected samples of 500 individuals. p = 0.002). There were no differences in disability between A transformation from a raw HAQ-II score to a RA and the other two disease categories. Rasch modeled score (rescaled to also range from 0 to 3), but which is now strictly linear, was accom- Discussion plished by fitting a cubic equation to the relationship This study has shown that the HAQ-II instrument can be between the raw HAQ-II score and the Rasch mod- considered psychometrically generic amongst rheumatol- eled score (Fig. 2). This equation has an R of 0.99. ogy clinic patients. It shows minimal invariance for disease Table 3 ANOVA for Differential Item Functioning by Disease (item in bold suggests possible DIF at the Bonferroni-corrected level of 0.0017) Item Class Interval Disease Class Interval x Disease Total MS F (df 9) p MS F (df 4) p MS F (df 36) p MS F (df 40) p 1 1.42 1.61 0.108 0.74 0.84 0.498 1.46 1.65 0.010 55.53 1.573 0.015 2 0.56 0.62 0.778 4.17 4.61 0.001 0.85 0.93 0.573 47.25 1.307 0.101 3 1.31 1.42 0.173 1.61 1.75 0.137 0.69 0.74 0.859 31.26 0.849 0.735 4 1.43 1.38 0.192 3.04 2.94 0.019 0.82 0.78 0.809 41.52 1.005 0.465 5 1.5 1.47 0.152 2.6 2.55 0.038 0.54 0.53 0.989 29.88 0.734 0.889 6 0.68 0.75 0.660 1.13 1.25 0.287 0.94 1.03 0.409 38.24 1.06 0.373 7 2.01 2.7 0.004 0.17 0.22 0.926 0.85 1.13 0.269 31.22 1.046 0.395 8 1.91 2.43 0.010 0.65 0.83 0.506 1.06 1.34 0.087 40.76 1.296 0.108 9 1.66 2.38 0.012 2.4 3.43 0.009 0.33 0.47 0.996 21.57 0.773 0.843 10 0.82 1.04 0.402 1.69 2.14 0.074 0.45 0.56 0.981 22.81 0.726 0.896 Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 5 of 7 Fig. 1 The item-characteristic curve (ICC) for item 2 (opening car doors). This plots the expected response to item 2 based on the individuals’ level of disability (person location). The curves for each disease category are superimposed upon the Rasch model (gray line). DIF would be implied by a significantly different location of a disease-specific ICC. RA (rheumatoid arthritis), IA (inflammatory arthritis), INF (inflammatory disorder), AICTD (autoimmune connective tissue disease) category, which implies that responses to each item and can be reasonably incorporated into the PAS-II score for the total score can be interpreted in just the same way for patients with any rheumatic disease to produce meaning- these disease categories. Therefore, it is valid to directly ful and comparable scores. RAPID3 uses a different ver- compare physical disability between diseases, and it was sion of HAQ, which will require a similar analysis to found that patients with RA have slightly more disability confirm invariance by disease category. on average than patients with two other disease categories. We have also described a transformation of the raw The results make the HAQ-II instrument a useful indica- HAQ-II score that may be useful for aggregated data ana- tor of physical functioning in a general rheumatology lysis in audit or clinical research, since it is strictly linear clinic, where patients with several different diseases may on an interval scale, making it very suitable for parametric come for treatment. Furthermore, the HAQ-II instrument statistical analysis and mathematical manipulation. Fig. 2 The relationship between Rasch modeled scores and raw HAQ-II scores closely fits a cubic equation Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 6 of 7 Fig. 3 The distribution of Rasch-modeled HAQ-II scores by disease category The meaning of changes in HAQ scores within indi- Conclusions viduals or between groups is highly dependent upon the The HAQ-II instrument has good psychometric proper- linearity of the scale. A non-linear scale makes it very ties including invariance by disease, suggesting that the difficult to compare changes at different starting points measure can be used with confidence in general on the scale, as has been shown for the 10 cm Pain vis- rheumatology clinics. Although theoretically attractive, it ual analogue scale [28]. The conventional minimal im- is not yet clear whether transformation of raw scores to portant difference (MCID) for HAQ-DI in RA is 0.20 to a Rasch-modelled score confers practical advantages. 0.22 [29] but may be larger [30]. For HAQ-II, its authors suggest MCID of 0.34. However, MCID assume a linear Abbreviations ANOVA: Analysis of variance; DIF: Differential item functioning; HAQ: Health scale, which is clearly not the case for the raw scores. Assessment Questionnaire; ICF: International Classification of Health, More meaningful values of MCID should be directly de- Functioning and Disability; MCID: Minimal clinically important difference; termined using Rasch-modelled scores compared to pa- OMERACT: Outcome Measures in Rheumatology Clinical Trials; PAS: Patient Activity Scale; RA: Rheumatoid arthritis; RAPID3: Routine Assessment of tient perception of change. Patient Index Data 3; RMSEA: Root mean square error of approximation; The main limitation of this study is the semi-arbitrary RUMM: Rasch Unidimensional Measurement Models; SLE: Systemic Lupus way by which rheumatic diseases were grouped together. Erythematosus; SPSS: Statistics Package for the Social Sciences; VAS: Visual Analogue Scale; WHO: World Health Organisation It is possible that more distinct diseases may show dif- ferential item functioning which is not apparent when Funding two or more diseases are grouped together. On the other This work received no specific funding but was supported by the Hutt Valley hand, grouping similar diseases together may increase District Health Board and the University of Otago. the statistical power to show differences, although this assumes that the within-group diseases associate with Availability of data and materials physical functioning in a similar way. In addition, there The dataset used and analysed during the current study are available from the corresponding author on reasonable request. is some functional heterogeneity within some relatively defined diseases such as systemic lupus erythematosus and psoriatic arthritis. Overall, it is unclear whether a Authors’ contributions WJT conceived and designed the study, analysed the data and wrote the different approach to grouping diseases would have manuscript. KP designed the study, collected the data and critically reviewed yielded different results, and could be an avenue for fur- the manuscript. Both authors authorized submission of the manuscript for ther testing.. publication. Both authors read and approved the final manuscript. Taylor and Parekh Health and Quality of Life Outcomes (2018) 16:108 Page 7 of 7 Ethics approval and consent to participate 17. Wolfe F, Michaud K, Pincus T. Development and validation of the health Ethical approval was granted by the New Zealand Health and Disability Ethics assessment questionnaire II: a revised version of the health assessment Committee without full review as part of its standing procedures for questionnaire. Arthritis Rheum. 2004;50:3296–305. observational, low risk studies. 18. Andrich D, Sheridan B, Luo G. RUMM2030: Rasch Unidimensional Models for Measurement. Perth: RUMM Laboratory; 1997–2012. Competing interests 19. Boone WJ, Staver JR, Yale MS. The Rasch model and item response theory The authors declare that they have no competing interests. models: identical, similar, or unique? In: Rasch analysis in the human sciences. Dordrecht: Springer; 2014. 20. Perline R, Wright BD, Wainer H. The Rasch model as additive conjoint Publisher’sNote measurement. Appl Psychol Meas. 1979;3:237–55. Springer Nature remains neutral with regard to jurisdictional claims in published 21. Alan Tennant PJF. The root mean square error of approximation (RMSEA) as maps and institutional affiliations. a supplementary statistic to determine fit to the Rasch model with large sample sizes. Rasch Measurement Transactions. 2012;25:1348–9. Author details 22. Jr WF. Reliability Statistics. Rasch Measurement Transactions. 1992;6:238. Department of Medicine, University of Otago Wellington, PO Box 7343, 23. Smith EV. Detecting and evaluation the impact of multidimensionality using Wellington, New Zealand. Wellington Regional Rheumatology Unit, Hutt item fit statistics and principal component analysis of residuals. J Appl Meas. Valley District Health Board, Wellington, New Zealand. General Medicine 2002;3:205–31. Service, Capital and Coast District Health Board, Wellington, New Zealand. 24. Andrich D, Hagquist C. Real and artificial differential item functioning in Polytomous items. Educ Psychol Meas. 2014;75:185–207. Received: 24 October 2017 Accepted: 21 May 2018 25. Hagquist C, Andrich D. Is the sense of coherence-instrument applicable on adolescents? A latent trait analysis using Rasch-modelling. Personal Individ Differ. 2004;36:955–68. References 26. Erdfelder E, Faul F, Buchner A, GPOWER. A general power analysis program. 1. Kostanjsek N, Rubinelli S, Escorpizo R, Cieza A, Kennedy C, Selb M, Stucki G, Behavior Research Methods, Instruments, & Computers. 1996;28:1–11. Ustun TB. Assessing the impact of health conditions using the ICF. Disabil 27. Tang K, Canadian Arthritis Network work productivity G. Disease-related Rehabil. 2011;33:1475–82. differential item functioning in the work instability scale for rheumatoid 2. World Health Organization. International classification of functioning, arthritis: converging results from three methods. Arthritis care & research. Disability and Health: ICF. Geneva: WHO; 2001. 2011;63:1159–69. 3. Taylor WJ, Geyh S. A rehabilitation framework: the international classification 28. Kersten P, White PJ, Tennant A. Is the pain visual analogue scale linear of functioning, disability and health. In: Dean SG, Siegert RJ, Taylor WJ, and responsive to change? An exploration using Rasch analysis. PLoS editors. Interprofessional Rehabilitation: A Person-Centred Approach. One. 2014;9:e99485. Chichester: Wiley-Blackwell; 2012. 29. Wells GA, Tugwell P, Kraag GR, Baker PRA, Groh J, Redelmeier DA. Minimum 4. Tugwell P, Boers M. Developing consensus on preliminary core efficacy important difference between patients with rheumatoid-arthritis - the endpoints for rheumatoid arthritis clinical trials. OMERACT Committee. J patients perspective. J Rheumatol. 1993;20:557–60. Rheumatol. 1993;20:555–6. 30. Ward MM, Guthrie LC, Alba MI. Clinically important changes in individual 5. Bellamy N, Kirwan J, Boers M, Brooks P, Strand V, Tugwell P, Altman R, Brandt K, and composite measures of rheumatoid arthritis activity: thresholds Dougados M, Lequesne M. Recommendations for a core set of outcome applicable in clinical trials. Ann Rheum Dis. 2015;74:1691–6. measures for future phase III clinical trials in knee, hip, and hand osteoarthritis. Consensus development at OMERACT III. J Rheumatol. 1997;24:799–802. 6. van der Heijde D, van der Linden S, Bellamy N, Calin A, Dougados M, Khan MA. Which domains should be included in a core set for endpoints in ankylosing spondylitis? Introduction to the ankylosing spondylitis module of OMERACT IV. J Rheumatol. 1999;26:945–7. 7. Merkel PA, Aydin SZ, Boers M, Direskeneli H, Herlyn K, Seo P, Suppiah R, Tomasson G, Luqmani RA. The OMERACT core set of outcome measures for use in clinical trials of ANCA-associated vasculitis. J Rheumatol. 2011;38:1480–6. 8. Schumacher HR Jr, Taylor W, Edwards NL, Grainger R, Schlesinger N, Dalbeth N, Sivera F, Singh JA, Evans R, Waltrip RW, et al. Outcome domains for studies of acute and chronic gout. J Rheumatol. 2009;36:2342–5. 9. Bruce B, Fries JF. The health assessment questionnaire (HAQ). Clin Exp Rheumatol. 2005;23:S14–8. 10. McDowell I, Newell C. Measuring Health. A guide to rating scales and questionnaires. Second edn. New York: Oxford University Press; 1996. 11. Fries JF, Bruce B, Cella D. The promise of PROMIS: using item response theory to improve assessment of patient-reported outcomes. Clin Exp Rheumatol. 2005;23:S53–7. 12. Unick GJ, Shumway M, Hargreaves W. Are we ready for computerized adaptive testing? Psychiatric services (Washington, DC). 2008;59:369. 13. Pincus T, Swearingen CJ, Bergman M, Yazici Y. RAPID3 (routine assessment of patient index data 3), a rheumatoid arthritis index without formal joint counts for routine care: proposed severity categories compared to disease activity score and clinical disease activity index categories. J Rheumatol. 2008;35:2136–47. 14. Wolfe F, Michaud K, Pincus T. A compositie disease activity scale for clinical practice, observational studies and clinical trials: the patient activity scale (PAS/PAS-II). J Rheumatol. 2005;32:2410–5. 15. Parekh K, Taylor WJ. The patient activity scale-II (PAS-II) is a generic indicator of active disease in patients with rheumatic disorders. J Rheumatol. 2010;37:1932–4. 16. Tennant A, Conaghan PG. The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis & Rheumatism. 2007;57:1358–62.

Journal

Health and Quality of Life OutcomesSpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off