A living network meta-analysis of treatments for rheumatoid arthritis: Novelty at the frontier of living evidence
2University of Calgary, Canada
3Ottawa Hospital Research Institute, Canada
Background: Choosing between the many effective disease-modifying therapies (DMARDs) for rheumatoid arthritis requires timely, high-quality systematic review of the best available evidence. ‘Living’ systematic reviews fast-track evidence synthesis without compromising rigorous systematic review methods. Network meta-analysis (NMA) is an attractive option for comparing rheumatoid arthritis treatments, but the development of ‘living’ NMAs presents unique methodological challenges.
Objectives: Our aim is to establish a sustainable, living NMA off DMARDs for rheumatoid arthritis and to facilitate the evolution of living NMA methods.
Methods: We are conducting three linked Cochrane living systematic reviews and living NMAs of DMARDs for rheumatoid arthritis using methods recommended by the Cochrane Living systematic Review Network, supplemented by new methods developed for this review. The three reviews will compare all current DMARD therapies for three separate populations of rheumatoid arthritis patients, grouped by prior therapy received (DMARD-naïve, DMARD-inadequate response, biologic DMARD- inadequate response).
Results: Several novel and emerging methods for the production of a living NMA have been incorporated and are described in five categories (“The 5 A’s”).
1. Accessibility: All raw data extracted will be formatted and hosted with the model code on the Open Science Framework in order to promote research efficiency and transparency.
2. Automation: Novel computer code has been developed to automatically read and format extracted data and generate GRADE tables.
3. Alliances: International collaborations aim to provide ‘global evidence for local recommendations’.
4. Applicability: These living NMAs will provide a direct living evidence base for living rheumatoid arthritis guidelines.
5. Atomisation: Identification, categorisation and extraction of data from RCTs is performed as ‘microtasks’ by a trained ‘crowd’ of volunteers.
Conclusions: Near real-time incorporation of evidence into systematic reviews is now possible, with incorporation into living guidelines as the next step. Novel methodologies will streamline the NMA process and inform living NMA methods.
Patient, public and/or healthcare consumer involvement: Trained ‘crowd’ of volunteers contributed to step 5.