Performance of the RABBIT infection risk score in an Argentinian rheumatoid arthritis cohort

Abstract

Patients with rheumatic autoimmune diseases have a higher risk of infections compared with age-and sex-matched controls. In Latin America, there are no validated tools to assess the risk of serious infection. The objectives were to estimate the incidence of serious infections in a cohort of rheumatoid arthritis (RA) patients followed for 12 months and to validate the RABBIT risk score for serious infections. Patients with RA were included and followed for 12 months. Baseline sociodemographic data, comorbidities, RA characteristics, and vaccination status were recorded. The baseline RABBIT risk score was calculated. Serious infections were documented, describing site and time since enrollment. Six hundred five patients were included (13 centers). The incidence of serious infection was 5% (95% CI 3-7). The most frequent sites were respiratory and urinary (90%). Performance of RABBIT risk score: patients with no infection during follow-up had a median score of 1.2 (IQR 0.8-2.1) and patients with infection 5.1 (IQR 2.15-12.6) p 0.00001. ROC curve analysis: AUC 0.86 (95% CI 0.8-0.94), best cut-off 2.85 (sensibility 75%, specificity 85%). The incidence of serious infections was 5% during the follow-up. The RABBIT score performed excellently in our patients. Key Points • The RABBIT risk score for serious infections showed an excellent performance in a population different (Latin America) from the original one included in the German registry. • This may assist rheumatologists in selecting drugs for patients according to the individual risk of infection, in a fast and simple way.

Keywords: Biological products; DMARD; Glucocorticoid; Infections; Rheumatoid arthritis.