Maxime Verhoeven

22 Chapter 2 Medline and Embase. The search combined terms relating to early RA, terms for cs-, b, and tsDMARD and remission, and publications limited to the last 5 year and English language. More details about the research question and search terms can be found in the Supplementary Data S1. We defined more intensive, remission-induction strategies as initiating treatment with a bDMARD or a tsDMARD, both with or without a csDMARD, or initiating a csDMARD with moderately-high dosed GCs, with delayed tapering (not ‘bridging therapy’), or starting ≥2 csDMARDs. The single csDMARD-initiating strategy was defined as starting treatment with a single csDMARD, with or without GCs as bridging therapy, according to the current guidelines. All titles and abstracts were screened by MMAV. If the reviewer was unsure about in-/excluding an abstract, it was discussed with one other co-author (PMJW) and one co-investigator (MdH) to reach consensus, and in case of remaining doubt based on title/abstract, the publication was included for full text evaluation. Full text screening was performed using the same strategy. The following selection criteria were used; 1) human studies, 2) (very) early RA patients, 3) remission-induction strategy arm (according to definition of remission- induction strategy, see above), 4) single csDMARD-initiating strategy arm (according to definition of single csDMARD-initiating strategy, see above) and, 5) results presented regarding the comparison of a remission-induction strategy and a single csDMARD- initiating strategy on an outcome of remission. Remission was defined as remission according to a validated disease activity index or the Boolean definition. 1 Randomized controlled trials (RCTs) as well as cohort studies with appropriate correction for multiple confounders were selected. Long-term extension studies of trials satisfying the above criteria were also selected to investigate long-term effects of remission-induction strategies on e.g., radiographic progression. Data extraction and outcome assessment The following data of studies was extracted: publication year, study design, patients’ baseline characteristics (age, gender, rheumatoid factor (RF) status, health assessment questionnaire (HAQ), symptom duration, disease activity score assessing 28 joints (DAS28)), a description of the single csDMARD-initiating strategy and the remission- induction strategy, the number of patients per arm, a description of the remission outcome, the number of patients achieving remission per arm, a description of missing data and other remarks deemed necessary. In case of a study evaluating long-term outcomes of a remission-induction strategy, we extracted additional information (if available) for the follow-up duration, outcome for disease activity, medication use and radiographic progression.