Measure of activity performance of the hand (MAP-Hand) questionnaire: linguistic validation, cultural adaptation and psychometric testing in people with rheumatoid arthritis in the UK

Yeliz Prior, Alan Tennant, Sarah Tyson, Ingvild Kjeken, Alison Hammond

Developed in the Norway, the Measure of Activity Performance of the Hand (MAP-Hand) assesses 18 activities performed using the hands. It was developed for people with rheumatoid arthritis (RA) using patient generated items, which are scored on a 0-3 scale and summarised into a total score range (0 to 54). This study reports the development and psychometric testing of the British English MAP-Hand in a UK population of people with RA. Recruitment took place in the National Health Service (NHS) through 17 Rheumatology outpatient clinics. Phase 1 (cross-cultural adaptation) involved: forward translation to British English; synthesis; expert panel review and cognitive debriefing interviews with people with RA. Phase 2 (psychometric testing) involved postal completion of the MAP-Hand, Health Assessment Questionnaire (HAQ), Upper Limb HAQ (ULHAQ), Short-Form 36 (SF-36v2) and Disabilities of the Arm Shoulder Hand (DASH) to measure internal consistency (Cronbach's alpha); concurrent validity (Spearman's correlations) and Minimal Detectable Difference (MDC95). The MAP-Hand was repeated three-weeks later to assess test-retest reliability (linear weighted kappa and Intra-Class Correlations (ICC (2,1)). Unidimensionality (internal construct validity) was assessed using (i) Confirmatory Factor Analysis (CFA) (ii) Mokken scaling and (iii) Rasch model. The RUMM2030 software was used, applying the Rasch partial credit model. In Phase 1, 31 participants considered all items relevant. In Phase 2, 340 people completed Test-1 and 273 (80%) completed Test-2 questionnaires. Internal consistency was excellent (α = 0.96). Test-retest reliability was good (ICC (2,1) = 0.96 (95% CI 0.94, 0.97)). The MAP-Hand correlated strongly with HAQ20 (rs = .88), ULHAQ (rs = .91), SF-36v2 Physical Functioning (PF) Score (rs = -.80) and DASH (rs = .93), indicating strong concurrent validity. CFA failed to support unidimensionality (Chi-Square 236.0 (df 120; p < 0.001)). However, Mokken scaling suggested a probabilistic ordering. There was differential item functioning (DIF) for gender. Four testlets were formed, resulting in much improved fit and unidimensionality. Following this, testlets were further merged in pairs where opposite bias existed. This resulted in perfect fit to the model. The British English version of the MAP-Hand has good validity and reliability in people with RA and can be used in both research and clinical practice.