Effect of order of presentation of a generic and a specific health-related quality of life instrument in knee and hip osteoarthritis: a randomized study

Summary Objective Current guidelines recommend using both a generic and a specific instrument to measure quality of life (QoL) among people with chronic diseases. However, the two questionnaires may not be independent, which raises the issue of whether the order in which they are completed influences their value. We aimed to assess the effect of order of presentation of a generic (SF36) and a specific (Osteoarthritis Knee and Hip QoL [OAKHQOL]) QoL instrument administered to patients with knee and hip osteoarthritis (OA). Methods We recruited 341 outpatients from rheumatology and orthopaedic surgery clinics. Demographic and clinical data and responses to the SF36 and OAKHQOL self-administered questionnaires were collected at baseline and 10 days after inclusion; surgical patients were also assessed 6 months post-surgery. The order of presentation of the instruments was randomized at inclusion. Results The order of instrument presentation had no significant effect on response rate, number of dimensions with a floor or a ceiling effect, or questionnaire scores. In one of 13 dimensions (social support dimension of the OAKHQOL), test–retest reliability was slightly better when the generic SF36 questionnaire was presented first (intra-class correlation coefficient (ICC) 0.84 vs 0.55). The analysis of sensitivity to change and discriminant ability did not favour one group over the other. Conclusion The order of presentation of a generic and a specific QoL questionnaire to patients with OA had an only marginal effect on the quality of responses and the QoL scores obtained.