Objectives To develop and examine the reliability and validity of a questionnaire assessing patient satisfaction with their osteoarthritis treatment.

Methods Item generation: Semi-structured interviews were performed among 20 French patients (6 suffering from arthritis of the knee, 6 of the hip and 8 of the hand), and 10 clinicians (5 rheumatologists and 5 general practitioners (GPs)). Interviews were recorded, transcribed and analysed. The first version of the questionnaire was produced in French and reviewed by 5 rheumatologists and 5 GPs. Content validity and cognitive debriefing were performed on 10 arthritis patients. The resulting arthritis treatment satisfaction (ARTS) questionnaire comprised 24 items.

Validation study: Psychometric analysis was performed to establish the structure (Principal Component Analysis (PCA) and multi-trait analysis) and the internal consistency (Cronbach alpha) of the questionnaire on a cross-sectional sample of 797 arthritis patients recruited by their clinicians. Patients were on average 67.5 years old (sd = 10.4), 64.5% were women, 26%had arthritis of the hip, 58%had arthritis of the knee, and all patients had suffered from arthritis for an average of 7 years (sd = 6.4).

Results PCA revealed a clear four dimensional structure. In order to reduce the number of items, 6 out of the 24 items were discarded because of moderate to low item convergent validity, and because they did not modify the internal reliability coefficient and content validity of the questionnaire. The final questionnaire comprised 18 items consisting of 4 dimensions measuring treatment advantages, treatment convenience, apprehensions about treatment and satisfaction with medical care. Item convergent and discriminant validity and the internal consistency reliability were satisfactory (Cronbach alpha of 0.84 for the global questionnaire which varied between 0.63 and 0.88 per dimension). PCA and multitrait analysis were reproduced on an independent sample of 137 patients, thus confirming the relevance of the 4 dimensional structure and internal consistency reliability of the questionnaire.

Conclusion Psychometric validity of this first treatment satisfaction questionnaire specific to arthritis is promising and further psychometric validation (clinical validity and reproducibility) is presently underway.