Article Text

Download PDFPDF

CS-27 Mortality among patients with systemic lupus erythematosus hospitalized with sepsis varies widely among U.S. hospitals
  1. Michael M Ward1,
  2. Maria Tektonidou2 and
  3. Abhijit Dasgupta1
  1. 1National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, U.S
  2. 2First Department of Propaedeutic Internal Medicine, Joint Academic Rheumatology Program, Medical School, National and Kapodistrian University of Athens, Athens, Greece


Background Sepsis is a serious infection that is often difficult to treat. We investigated if the risk of mortality in patients with systemic lupus erythematosus (SLE) hospitalized with sepsis varies among hospitals.

Methods We used the National Inpatient Sample (2002–2011) to obtain data on outcomes of adults age 18 to 64 with SLE who were admitted with a diagnosis of sepsis. We included 424 hospitals that had at least 5 hospitalizations of patients with SLE and sepsis. We abstracted data on demographic features, diagnoses, and mortality for these patients and for patients without SLE hospitalized with sepsis at the same set of hospitals. We used machine learning methods to derive the expected risk of in-hospital mortality for each patient, based on demographic features, insurance status, comorbidities, and diagnoses related to organ failure. We then computed the observed/expected ratio for in-hospital mortality separately for SLE and non-SLE patients by hospital. To further adjust for case-mix, the ratio of these two ratios (RR) was used to evaluate the relative likelihood of death among those with SLE compared to those without SLE at the same hospital.

Results We included 4001 hospitalizations of patients with SLE and sepsis, and 2 02 888 hospitalizations of patients with sepsis without SLE. The number of SLE hospitalizations ranged from 5 to 66 per hospital. Across all hospitals, 11.5% of patients with SLE and 13.1% of patients without SLE died during the hospitalization. The RR was greater than 1.0 in 44% of hospitals, and greater than 2.0 in 16% of hospitals, indicating increased risk of mortality by a factor of two or more among patients with SLE compared to those without SLE at the same hospital. Based on classification tree analysis, large urban hospitals were less likely to have RR >2 than smaller hospitals (13% vs 22%), particularly those in the midwest or northeast (35%).

Conclusions Risks of mortality due to sepsis among patients with SLE varied widely among U.S. hospitals. Risks were higher for patients with SLE at smaller hospitals.

Acknowledgements Supported by the Intramural Research Program, NIAMS/NIH.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.