AbstractBackgroundThis study aimed to compare the cardiovascular safety of interleukin‐6 inhibitors (IL‐6i) and Janus Kinase inhibitors (JAKi) to tumour necrosis factor inhibitors (TNFi).MethodsWe conducted a retrospective cohort study using population‐based electronic databases from Hong Kong, Taiwan and Korea. We identified newly diagnosed patients with rheumatoid arthritis (RA) who received b/tsDMARDs first time. We followed patients from b/tsDMARD initiation to the earliest outcome (acute coronary heart disease, stroke, heart failure, venous thromboembolism and systemic embolism) or censoring events (death, transformation of b/tsDMARDs on different targets, discontinuation and study end). Using TNFi as reference, we applied generalized linear regression for the incidence rate ratio estimation adjusted by age, sex, disease duration and comorbidities. Random effects meta‐analysis was used for pooled analysis.ResultsWe identified 8689 participants for this study. Median (interquartile range) follow‐up years were 1.45 (2.77) in Hong Kong, 1.72 (2.39) in Taiwan and 1.45 (2.46) in Korea. Compared to TNFi, the adjusted incidence rate ratios (aIRRs) (95% confidence interval [CI]) of IL‐6i in Hong Kong, Taiwan and Korea are 0.99 (0.25, 3.95), 1.06 (0.57, 1.98) and 1.05 (0.59, 1.86) and corresponding aIRR of JAKi are 1.50 (0.42, 5.41), 0.60 (0.26, 1.41), and 0.81 (0.38, 1.74), respectively. Pooled aIRRs showed no significant risk of cardiovascular events (CVEs) associated with IL‐6i (1.05 [0.70, 1.57]) nor JAKi (0.80 [0.48, 1.35]) compared to TNFi.ConclusionThere was no difference in the risk of CVE among RA patients initiated with IL‐6i, or JAKi compared to TNFi. The finding is consistent in Hong Kong, Taiwan and Korea.