Predictors of mortality in children with systemic lupus erythematosus

Background Systemic lupus erythematosus (SLE) is a multisystem chronic disease with a relatively high mortality rate in children, despite improvements in prognosis and survival rate over the past decade. Studies on the predictors of mortality in children with SLE, especially in low-and middle-income countries, are limited. Objective To determine the predictors of mortality of children with SLE. Methods This was case-control study using data from medical records of children with SLE at Dr. Sardjito Hospital, Yogyakarta, Indonesia, between 2009 and 2017. Subjects were children aged < 18 years diagnosed with SLE. Cases were those who died within one year of diagnosis; the controls were those who were discharged alive. From subjects' medical records, we collected clinical data including age, sex, date of diagnosis, nutritional status, anti-dsDNA antibody, antinuclear antibody (ANA), hypertension, disease activity based on the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) score, proteinuria, thrombocytopenia, mortality/survival outcome, date of death, cause of death, and clinical data including fever, seizures, antibiotic used, microbial culture outcomes, and infection-related diagnoses. We performed bivariate analysis of the association between predictor variables (SLEDAI score, proteinuria, infection, hypertension, and seizures) and mortality outcome (survival or death), followed by logistic regression analysis. Results Eighty-four patients with SLE were included, of which 72 were female. Median age at diagnosis was 14 (range 4-18) years. Twenty-three patients (27%) died within one year after diagnosis. The most common causes of death were infection and renal failure in 8/23 and 7/23 subjects, respectively. On bivariate analysis, the variables significantly associated with mortality were hypertension (OR 3.34, 95% CI 1.22 to 9.14) and infection (OR 3.71; 95% CI 1.36 to 10.12). Seizures, proteinuria, and SLEDAI score were not found to be significantly associated with mortality. On logistic regression analysis, infection was the only significant predictor of mortality (OR 3.22; 95% CI 1.15 to 9.05). Conclusion Among the factors studied, infection is significantly associated with mortality in children with SLE.