Comparative Efficacy Randomized Controlled Trials in Rheumatology Guidelines

Background Comparative efficacy randomized controlled trials (RCTs) compare two active interventions in a head-to-head design. They are useful for informing clinical practice guidelines, but the degree to which such trials inform clinical practice guidelines in rheumatology is unknown. Methods The American College of Rheumatology (ACR) and European Alliance of Associations for Rheumatology (EULAR) websites were searched from January 1, 2017, to June 12, 2021, for clinical practice guidelines. RCTs referenced by each guideline were identified, and information regarding design and outcomes were extracted. Clinical practice recommendations from each guideline were also analyzed. Results Fifteen ACR- and nine EULAR-endorsed guidelines were included, which cited 609 RCTs and provided 481 recommendations. Referenced RCTs enrolled an average of 418 patients (SD 985), most commonly evaluated biologic/targeted synthetic disease-modifying antirheumatic drugs (70.1%), and infrequently used a head-to-head design (28%). A minority of recommendations received a high level of evidence (LOE) by the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) methodology (2.9%) or an "A" grade by the Oxford Centre for Evidence based Medicine Standards (OCEBM) methodology (28.9%). LOE was higher for recommendations informed by RCTs (P < 0.001) or head-to-head RCTs (P = 0.008). Many recommendations received a strong recommendation despite low (8 [2.6%]) or very low (25 [8.3%]) LOE. Conclusion Less than one in six rheumatology guideline recommendations are informed by head-to-head RCTs. Recommendations that were informed by head-to-head RCTs were more likely to have a high LOE by both GRADE and OCEBM. Efforts to introduce more comparative efficacy RCTs should be undertaken.