Background/Purpose: For individuals with systemic lupus erythematosus (SLE), hospital readmission rate is associated with quality of care and prognosis. Our objective was to evaluate independent risk factors for readmission and determine the major reasons for readmission in a contemporary and nationally representative sample of SLE patients.

Methods: We used the Nationwide Readmissions Database (NRD) from the Healthcare Cost and Utilization Project (HCUP), which includes all-payer nation-wide data from US non-federal hospitalizations across 28 states, to identify adults with SLE who were discharged from hospital to home during January-November of 2016 and 2017. Subjects were identified using SLE ICD-10-CM primary and secondary diagnosis codes (M32.1-M32.9). 30-day all-cause readmissions were identified. Comorbidities were defined using the Elixhauser classification. Demographics (gender, age group, insurance, median income quartile by ZIP code, and population density), hospital-level variables (bedsize and teaching hospital), and SLE manifestation-related ICD-10-CM diagnosis codes (glomerular disease, pleurisy, pericarditis, seizure, ischemic stroke, autoimmune hemolytic anemia, leukopenia, and thrombocytopenia) were recorded. Survey-specific statistical methods were used to obtain weighted frequencies and percentages of study covariates. A multivariable survey-specific logistic regression model was used to identify factors associated with readmission adjusting for the covariates mentioned above. We identified the main cause of readmission using primary ICD-10-CM diagnosis codes for readmission and the HCUP Clinical Classifications Software, Refined.

Results: There were 132,400 hospitalized adults with SLE discharged home during the study period. 88.3% were female, with median age of 51.0 years (IQR 38.7-61.9). 18,973 (14.3%) people were readmitted within 30 days of discharge from their index hospitalization. Median hospital stay was 3.3 days (IQR 1.7-5.9) for readmitted patients and 2.5 (IQR 1.3-4.3) in patients without readmission. The predictors with largest magnitude of effect for readmission in an adjusted regression model were autoimmune hemolytic anemia [odds ratio (OR) 1.86, 95% CI 1.51-2.29], glomerular disease (OR 1.41, 95% CI 1.32-1.50), pericarditis (OR 1.40, 95% CI 1.19-1.66), heart failure (OR 1.38, 95% CI 1.28-1.49), and age 18-30 (OR 1.37, 95% CI 1.24-1.51, vs age ≥65) (Figure 1). The most common causes for readmission were sepsis (n=1,446, 7.6%), SLE (n=1,407, 7.4%), heart failure (n=668, 3.5%), pneumonia (n=605, 3.2%), and complication of surgical/medical care, injury (n=565, 3.0%) (Table 1).

Conclusion: In this first nationally representative study of SLE readmissions, we identified independent risk factors for readmission in adult SLE patients discharged to home. The strongest identified risk factors were related to select SLE manifestations such as cytopenias and comorbidities such as diabetes and cardiopulmonary disease. Certain demographic features also contributed to risk, such as young age and public insurance.

Multivariable Logistic Regression Model for Risk of 30-day Readmission in US Adults with Systemic Lupus Erythematosus, 2016-2017

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-thirty-day-hospital-readmissions-in-systemic-lupus-erythematosus-in-the-us-a-nationwide-study/