Time-dependent increase in risk of hospitalisation with infection among Swedish RA patients treated with TNF antagonists.

The degree to which treatment with tumour necrosis factor (TNF) antagonists may be associated with increased risks for serious infections is unclear. An observational cohort study was performed using prospectively collected data from the Swedish Biologics Register (ARTIS) and other national Swedish registers. First, in the ARTIS, all 4167 rheumatoid arthritis (RA) patients starting TNF antagonist treatment between 1999 and 2003 were identified. Secondly, in the Swedish Inpatient Register, all individuals hospitalised for any reason and who also carried a diagnosis of RA, between 1964 and 2003 (n = 44 946 of whom 2692 also occurred in ARTIS), were identified. Thirdly, in the Swedish Inpatient Register, all hospitalisations listing an infection between 1999 and 2003 were identified. By cross-referencing these three data sets, RRs for hospitalisation with infection associated with TNF antagonist treatment were calculated within the cohort of 44 946 RA patients, using Cox regression taking sex, age, geography, co-morbidity and use of inpatient care into account. Among the 4167 patients treated with TNF antagonists, 367 hospitalisations with infections occurred during 7776 person-years. Within the cohort of 44 496 RA patients, the RR for infection associated with TNF antagonists was 1.43 (95% CI 1.18 to 1.73) during the first year of treatment, 1.15 (95% CI 0.88 to 1.51) during the second year of treatment, and 0.82 (95% CI 0.62 to 1.08) for subjects remaining on their first TNF antagonist treatment after 2 years. Treatment with TNF antagonists may be associated with a small to moderate increase in risk of hospitalisation with infection, which disappears with increasing treatment duration.