Methods: ACTION was a 2-year, non-interventional, international, multicenter, cohort study that evaluated the retention and effectiveness of IV ABA in adults with moderate-to-severe RA. Pts who received ≥1 prior biologic and enrolled in countries with sufficient pt numbers to explore between-country effects were included in this analysis. Pts were stratified by their baseline BMI.⁷ Crude 2-year retention rate was estimated using the Kaplan–Meier method. The effect of BMI was analyzed through a multivariate Cox proportional hazard model clustered for site effects with conditional imputation of missing data for covariates. Hazard ratios and corresponding 95% CI were adjusted for socio-demographic variables, disease characteristics, comorbidities at initiation and treatment characteristics. Pts were considered adherent to ABA if the ratio of the number of infusions received to the number expected was between 80 and 120%.   Results: 1009/1131 (89.2%) evaluable pts had received ≥1 prior biologic; 995 were included in the analysis. Pts with higher BMI had shorter disease duration, more severe disease (assessed by TJC, DAS28, HAQ-DI), were more likely to have comorbidities, but were less likely to be RF or anti-cyclic citrullinated peptide positive. Crude retention rates and adjusted HR are shown in Table 1. Similar proportions of pts were considered adherent to ABA therapy in all BMI groups as follows: BMI <25 kg/m² = 82.8%, BMI 25–<30 kg/m² = 79.5%, BMI 30–<35 kg/m² = 85.0%, BMI ≥35 kg/m² = 84.1%.   Table 1