Laboratory-confirmed bloodstream infection in systemic lupus erythematosus: Risk profiling and short-term mortality

Abstract

Objectives To delineate laboratory-confirmed bloodstream infection (LCBI), analyze risk factors for its occurrence and predictors for its short-term mortality in systemic lupus erythematosus (SLE) patients. Methods A single center, retrospective, case-controlled study was performed in 159 SLE patients (2013–2019) to identify risk factors of LCBI by comparing patients with LCBI (n = 39) to those without infection (n = 120). The predictors associated with 30-day mortality in LCBI patients were also analyzed. Results Altogether 40 bacteria strains were isolated in 39 LCBI patients with a predominance of the gram-negative bacilli (24 strains, 60.0%). Escherichia coli and Staphylococcus aureus were the leading Gram-negative and Gram-positive microorganisms, respectively. Occurrence of LCBI was independently predicted by: SLE disease duration >4 years, SLEDAI score >4 points, glucocorticoids dose >7.5 mg/d and the previous or concomitant occurrence of autoimmune hemolytic anemia (AIHA) or thrombotic microangiopathy (TMA). Based on the identified risk factors, we developed a matrix model for the risk of future LCBI. The 30-day mortality (39 cases) was 23.1% and healthcare-associated LCBI was a predictor for 30-day mortality in SLE patients compared with community-acquired LCBI. Conclusion Longer duration, higher disease activity and glucocorticoids dose, and occurrence of AIHA or TMA were risk factors of LCBI in SLE and its poor short-term prognosis may attribute to healthcare-associated LCBI.