RESUMO
INTRODUCTION: Although extensively characterized in the outpatient setting, systemic lupus erythematosus (SLE) in the hospitalization wards is still scarcely portrayed, particularly in the perspective of its evolution over the years. METHODS: Retrospective analysis of SLE patients hospitalized in the Department of Autoimmune Diseases of a university hospital during a 20-year period (1995-2015), describing hospitalization characteristics, causes and predictors of outcome. RESULTS: A total of 814 hospitalizations concerning 339 patients were analysed. The main causes of admission were flare (40.2%), infection (19.2%), diagnostic procedures (18.8%) and thrombotic events (5.4%). Therapy with cyclophosphamide (odds ratio (OR) 1.908, p = 0.047) was associated with admission due to infection, while antimalarials displayed a protective effect (OR 0.649, p = 0.024). Nearly 3.9% of patients required admission to an intensive care unit, with associated antiphospholipid syndrome (OR 7.385, p = 0.04) standing as a predicting factor for this outcome. Readmission at 30 days occurred in 5.8% of patients, with thrombocytopenia (OR 6.007, p = 0.002) and renal involvement (OR 3.362, p = 0.032) featuring as predicting factors. Eight patients died, with antiphospholipid syndrome (OR 26.814, p = 0.02) and thrombocytopenia (OR 31.523, p = 0.01) being associated with mortality. There was no significant variation in patients' demographics or admission causes across the 20-year period, except for a decrease in admissions due to thrombotic and musculoskeletal causes. Recently, an increase in the use of mycophenolate mofetil and lower doses of glucocorticoids were noted. CONCLUSION: While demographics of SLE hospitalizations have not markedly changed over the past 20 years, changes in therapy patterns were observed. Thrombocytopenia, antiphospholipid syndrome and renal involvement featured as predictors of poor outcome.