Rheumatoid arthritis patients treated in trial and real world settings: comparison of randomized trials with registries.

Objective To investigate whether patients with RA enrolled in randomized controlled trials (RCTs) and observational studies may differ in terms of characteristics that could modify treatment effects, leading to an efficacy-effectiveness gap. Methods We conducted systematic literature reviews to identify RCTs and observational studies with RA, treated with rituximab, tocilizumab or etanercept. We extracted baseline characteristics and compared the data of RCTs and observational studies using fixed-effects meta-analyses for the RCTs and random-effects meta-analyses for the observational studies. We also assessed whether the baseline characteristics changed over time. Results Compared with patients enrolled in RCTs, those from observational studies were on average 3.0 years older (P < 0.001), suffered from RA for 3.1 years longer (P < 0.001), had 1.6 more prior disease modifying drugs (P = 0.001), and had a lower DAS-28 (difference -0.6, P < 0.001). CRP and ESR levels were slightly higher in RCTs. The HAQ-Disability Index (HAQ-DI) score was slightly lower in the RCT group. No differences were found in the percentages of included females or RF positivity. Over time, we found a significant decrease of - 0.08 in DAS-28 and a decrease of - 0.04 in HAQ-DI both in patients in RCTs and in patients from registries. Furthermore, ESR and CRP declined over time in RCT patients, but not in patients participating in observational studies. Conclusion There are substantial systematic differences in patient characteristics between RCTs and registries in RA. The efficacy seen in RCTs may not reflect real-world effectiveness