Abstract
Background Serious infections contribute significantly to morbidity and mortality in systemic lupus erythematosus (SLE)
Methods We undertook a retrospective analysis of all SLE patients enrolled in the Australian Lupus Registry & Biobank admitted to hospital with infection between 2009–2020. Comparisons of length of stay (LOS) according to different patient and disease characteristics and clinical decisions regarding corticosteroids (CS), immunosuppression (IS) and anti-microbial therapy were performed using ANOVA (Kruskal Wallis test). Cox and related regression models were used to identify associations between hospital LOS and clinical variables.
Results 53 patients with 85 separate hospitalisations were identified. Patients had a mean (SD) age of 44.6 (14.8) years. Mean hospital LOS was 15.4 (19.7) days. Admission to the Intensive Care Unit (ICU) occurred in 11.8% of cases. IS was withheld during 29.4% of infections. There was considerable variation in whether CS were modified from a patient’s baseline dose. Use of pulse CS and weaning baseline CS during infection were strongly associated with longer LOS. Unsurprisingly, ICU admission, intravenous anti-microbial use and nosocomial infections were also linked with increased LOS. Withholding immunosuppression during infection did not reduce LOS. Patient factors including age, Charlson Co-Morbidity Index (CCI), preceding disease activity and pre-admission immunosuppression did not influence LOS.
Conclusions From this single centre study hospital LOS was primarily influenced by factors relating to the severity of infection. Factors possibly relating to inadequately controlled SLE, such as attempted tapering of usual CS dose and administering pulse CS – commonly used to treat severe disease flares, also significantly prolonged infection admissions. Further research is required to identify the optimal approach towards modifying baseline CS and immunosuppression when managing infections in SLE patients.
