Patients’ preferences for osteoarthritis treatment: a systematic review of discrete-choice experiments

Given the significant opportunities and challenges of osteoarthritis (OA) treatment, as well as the increasing importance of preference data in clinical and policy decisions, several attribute-based preference studies (primarily discrete-choice experiments) have been conducted to assess preferences for OA treatment. As no overview of these studies is currently available, this study was designed to systematically review and summarize attributes-based preference studies for OA treatment. A systematic search in PUBMED was conducted to identify all peer-reviewed discrete-choice experiments or conjoint analyses that investigated patients or professionals preferences for OA treatment. Analysis was conducted using narrative synthesis and summary statistics. We identified 10 relevant articles with the majority of studies being published after 2012 (n=7). Most studies used a discrete-choice experiment (n=6) and were conducted in the USA (n=3) or in the UK (n=3). Studies were primarily conducted to assess patients (n=7) or physicians (n=3) preferences for OA treatment, suggesting that respondents were mainly concerned by the benefits and risks of treatment. Cost and route of administration were also shown to be significant predictors of preference in some studies. Some shortcomings of current practice with regard to DCEs applications in OA treatment were identified: (1) although most studies used appropriate methods to select attributes and levels, a greater involvement of patients in the preparation of the survey would be needed; (2) some studies did not report important information making it difficult to judge whether it was done appropriately and (3) heterogeneity in responses was only poorly investigated in the statistical analyses. In conclusion, this review suggests that attribute-based preference methods (such as discrete-choice experiment) represents a useful method to examine the preferences for OA treatment and that potential benefits and risk of adverse events were shown to be the most influential characteristics of OA treatment.