Predictors of serious infections in rheumatoid arthritis-a prospective Brazilian cohort.

Background: Infections increase mortality and morbidity and often limit immunosuppressive treatment in rheumatoid arthritis patients.

Objective: To analyze the occurrence of serious infections and the associated factors in a cohort of rheumatoid arthritis patients under real-life conditions.

Methods: We analyzed data from the REAL, a prospective observational study, that evaluated Brazilian RA patients, with clinical and laboratory data collected over a year. Univariate and multivariate analyses were performed from the adjustment of the logistic regression model Generalized Estimating Equations (GEE), with the primary outcome being the occurrence of serious infection, defined as need for hospitalization or use of intravenous antibiotics for its treatment.

Results: 841 patients were included with an average follow-up time of 11.2 months (SD 2.4). Eighty-nine serious infections occurred, corresponding to 13 infections per 100 patient-years. Pulmonary fibrosis, chronic kidney disease (CKD) and central nervous system disease increased the chances of serious infection by 3.2 times (95% CI: 1.5-6.9), 3.6 times (95% CI: 1.2-10.4) and 2.4 times (95% CI: 1.2-5.0), respectively. The use of corticosteroids in moderate doses increased the chances by 5.4 times (95% CI: 2.3-12.4), and for each increase of 1 unit in the health assessment questionnaire (HAQ), the chance increased 60% (95% CI: 20-120%).

Conclusion: The use of corticosteroids at moderate doses increased the risk of serious infection in RA patients. Reduced functionality assessed by the HAQ and comorbidities were other important factors associated with serious infection in this cohort.