OBJECTIVES: To compare the risk of serious infection across time cohorts in patients with inflammatory arthritis (IA) initiating their first biologic/targeted synthetic DMARD (b/tsDMARD), to that of the general population. Secondarily, to compare the development in infection risk during treatment across diagnoses and examine risk dynamics during the course of b/tsDMARD treatment. METHODS: Patients with IA starting their first b/tsDMARD were included from the prospective NOR-DMARD study. Controls were randomly drawn from the general population. Cox regressions were used to compare the 12-month risk of serious infections across three time cohorts following initiation (2009-2011, 2012-2014, 2015-2018) and risk during the course of treatment at 6-month intervals up to 24 months. RESULTS: A total of 4309 patients (RA, 1581
PsA, 1032
SpA, 1696) and 86 640 controls were included. From 2009 through 2018, 51 serious infections occurred during the first year of b/tsDMARD treatment in RA patients [hazard ratio (HR) 2.42 (95% CI 1.83, 3.21)] compared with controls and 52 serious infections were observed in patients with PsA/SpA [HR 1.91 (95% CI 1.44, 2.52)]. There were no significant differences in 12-month risk of serious infections during b/tsDMARD exposure between time cohorts. PsA/SpA patients had a consistently lower risk of serious infection compared with RA patients. The risk of serious infections did not change during the treatment course. CONCLUSION: Patients with IA starting their first b/tsDMARD between 2009 and 2018 had a consistently higher 12-month risk of serious infection compared with controls. No change in the risk of serious infection across time cohorts of b/tsDMARD initiation was observed, nor during the treatment course.