Incorporating adjustments for variability in control group response rates in network meta-analysis: a case study of biologics for rheumatoid arthritis

• DOI: 10.1186/s12874-019-0837-2
• Publication Date: 2019-10-16T13:09:37Z
• Authors: Chris Cameron; Abhishek Varu; Arthur Lau; Mahdi Gharaibeh; Marcelo Paulino; Raina Rogoza
• Keywords: Medicine (General); Biological Products; Statistical methods; Network Meta-Analysis; Biologics; Control Groups; 3. Good health; Arthritis, Rheumatoid; 03 medical and health sciences; R5-920; 0302 clinical medicine; Meta-Analysis as Topic; Antirheumatic Agents; Systematic review; Humans; Rheumatoid arthritis; Network meta-analysis; Research Article

Abstract Background The importance of adjusting for cross-study heterogeneity in control group response rates when conducting network meta-analyses (NMA) was demonstrated using a case study involving comparison biologics the treatment moderate-to-severe rheumatoid arthritis. Methods Bayesian NMAs were conducted American College Rheumatology (ACR) 50 based upon set randomized controlled trials (RCTs) identified by recently completed systematic review literature. In addition to performance an unadjusted NMA, model meta-regression fit data. Model evaluated, and findings from both analyses compared with regard clinical interpretations. Results ACR data total 51 RCTs 16,223 patients analyzed. Inspection variability considerable differences between studies. NMA incorporating adjustment this associated average change 38.1% magnitude ORs comparisons, over 64% odds ratio changed 15% or more. Important changes interpretations drawn comparisons improved modeling approach. Conclusions comparing moderate severe arthritis, failure adjust cross-trial arm can lead biased estimates comparative efficacy treatments.