Bloodstream infections among children with juvenile idiopathic arthritis: a prospective study from the onset of disease.

Objective To describe the incidence and nature of bloodstream infections (BSI) among children with juvenile idiopathic arthritis (JIA) followed-up prospectively from disease onset. Methods The Social Insurance Institution's (SII) national register on individuals with reimbursement for medication of chronic diseases was used to identify children with JIA from 2004 through 2011 and their medications. The National Infectious Disease Register (NIDR) collects data of all blood culture positive samples from all microbiology laboratories in Finland. We combined the NIDR and SII registers to identify JIA patients with BSI. Clinical and laboratory data of each JIA-BSI patient were collected from hospital records. Results There were 1604 JIA patients and 6630 person-years of follow-up. Five patients had BSI. During the first 5 years after diagnosis the cumulative emergence of BSI was 0.38% [95% confidence interval (CI) 0.16% to 0.92%]. The incidence rates were 7.5/10 000 follow-up years for JIA (95% CI 2.4-17.6) and 2.8/10 000 follow-up years for the age-matched general population (95% Cl 2.7-2.9). The standardised incidence ratio was 3.0 (95% Cl 1.2 to 7.2). The causative bacteria were Streptococcus pneumoniae, Staphylococcus aureus, Escherichia coli and Fusobacterium necrophorum. Three patients were on anti-rheumatic drugs, including two on TNF inhibitors. All patients responded rapidly to antimicrobial therapy and recovered uneventfully. Conclusion Although BSI is rare among children with JIA, the incidence is 3-fold higher than among the general population.