Using Parallel Streams of Evidence to Inform Guideline Development: The Case of the 2021 American College of Rheumatology Management of Rheumatoid Arthritis Guideline

Objective: We aim to describe an evidence synthesis approach using parallel streams of evidence that informed the development of the 2021 American College of Rheumatology (ACR) guideline for the management of rheumatoid arthritis (RA). Methods: We developed the evidence synthesis approach using parallel streams of evidence in multiple rounds of discussion, piloting, feedback, and revisions. A number of working groups involving ACR staff, content experts, and methodologists coordinated to develop and implement the approach. Results: We used a major stream of evidence that identified evidence specific to the clinical questions being addressed in the guideline (ie, we were able to match relevant articles to specific questions). We also used additional streams that identified data that applied across multiple questions. We describe in this article the different steps of the major stream, ie, screening and tagging, matching articles to question clusters, matching articles to individual questions, data abstraction and analysis, and Grading of Recommendations Assessment, Development and Evaluation (GRADEing). We then describe how we packaged the parallel streams of evidence into standardized structured tables to facilitate formulating the recommendations. These tables included information for the following factors: desirable effects, undesirable effects, certainty of evidence, valuation of outcomes, cost of interventions, and cost-effectiveness of interventions. The approach allowed us to match eligible articles for 47 of 81 clinical questions. We identified no eligible articles that addressed the remaining 34 questions. Conclusion: We were successful in using parallel streams of evidence to inform the development of the 2021 ACR guideline for the management of RA.