OP0014\u2005NO ASSOCIATION BETWEEN THE PROPORTION OF ELDERLY PEOPLE AND TRIAL RETENTION IN RHEUMATOID ARTHRITIS AND OSTEOARTHRITIS TRIALS: A SYSTEMATIC REVIEW WITH META-REGRESSION ANALYSES

Abstract

Background The elderly (defined by an age of ≥65 years) are underrepresented in current rheumatoid arthritis (RA) and osteoarthritis (OA) trials.1 This is partly due to age-related exclusion criteria. Investigators might be reluctant to include more elderly people because they fear reduced trial retention. Objectives To evaluate whether the proportion of included elderly individuals (PE) is associated with trial retention in current RA and OA trials. Methods This study is registered with protocols.io (dx.doi.org/10.17504/protocols.io.uhaet2e). In our previous study,1 two systematic searches in MEDLINE had yielded randomized controlled trials (RCTs) in rheumatoid arthritis (RA) and osteoarthritis (OA) on any intervention published in 2016 and 2017. Here we included all trials reporting data on retention. We either extracted the PE from the research manuscript or estimated it from an assumed (truncated) normal distribution. We coded retention into proportional effect sizes (logit-transformed for analysis and back-transformed for reporting). Subsequently, multiple mixed effects meta-regression models with several covariates assessed whether there is an association between the PE and trial retention. Sensitivity analyses evaluated whether associations were connected to attrition due to lack of efficacy (LoE) or adverse events (AE). Results 243 RCTs comprising 48,288 participants were deemed eligible, and 227 RCTs provided complete data on overall retention and all covariates. Pooled trial retention (random effects) was 88% (95%-CI 87% to 90%; I2 = 90%). The PE was not associated with trial retention in the unadjusted (slope β = 0.0 [–0.0 to 0.0]; p = 0.97; Figure 1) or any protocolized adjusted model (p-values depending on the adjustment: 0.14 to 0.90). Of all included covariates, only study duration (longer study duration being associated with inferior retention: p < 0.001) and the type of intervention (surgical interventions averaging the highest retention; psychological interventions averaging the lowest retention: p < 0.001) were associated with trial retention. Post hoc analyses allowing for interaction revealed a small statistically significant association between the PE and trial retention in pharmacological (retention increasing with the PE) and physical/physiotherapeutic (retention decreasing with the PE) interventions. Further sensitivity analyses showed no significant associations between attrition due to LoE or AE and the PE in any model. Conclusion Trial retention in RA and OA trials is very high, and unaffected by the proportion of elderly. References [1] Palmowski A, Buttgereit T, Palmowski Y, et al. Applicability of trials in rheumatoid arthritis and osteoarthritis: A systematic review and meta-analysis of trial populations showing adequate proportion of women, but underrepresentation of elderly people. Semin Arthritis Rheum2018 [ePub ahead of print]. Acknowledgement This study is part of the GLORIA trial and project and has received funding from the European Union (Horizon 2020) under grant agreement number 634886. Disclosure of Interests Andriko Palmowski: None declared, Sabrina Mai Nielsen: None declared, Thomas Buttgereit: None declared, Yannick Palmowski: None declared, Maarten Boers Consultant for: Bristol-Myers Squibb, Teva, Novartis, Pfizer, GlaxoSmithKline, Robin Christensen Grant/research support from: AbbVie Inc, and the Oak Foundation, Speakers bureau: Roche, Frank Buttgereit: None declared