Rates, Diagnoses, and Predictors of Unplanned 30-Day Readmissions of Critical Care Survivors Hospitalized for Lung Involvement in Systemic Lupus Erythematosus: An Analysis of National Representative US Readmissions Data

Introduction/objectives: Systemic lupus erythematosus (SLE) is a chronic autoimmune disease that frequently involves the lungs, contributing to significant morbidity in hospitalized patients. Critical care survivors with lung involvement in SLE are at particularly high risk for unplanned hospital readmissions, which can reflect the complexity of their disease, which often affects multiple organs and requires immunosuppressive therapy that increases infection risk. Severe pulmonary complications, critical illness sequelae, and challenges in medication adherence or follow-up care further contribute to their vulnerability. These factors result in frequent complications and flare-ups, making unplanned readmissions common in this population. This study assessed rates, most common reasons, and predictors of all-cause and SLE-related 30-day readmission among critically ill patients hospitalized for lung involvement in SLE.

Methods: We analyzed the 2021 National Readmissions Database. Critically ill non-elective adult hospitalizations for lung involvement in SLE were identified for analysis using a combination of the ICD-10 diagnostic code for SLE with lung involvement (M32.13) and presence of any procedure codes for mechanical ventilation, tracheostomy, extracorporeal membrane oxygenation, or bronchoscopy. Non-lung-related SLE admissions, non-SLE-related lung disorders, patients with concomitant COPD, history of COVID-19 or severe asthma, patients transferred in from other hospitals or admitted for <24 hours, and patients with a DNR order were excluded. We used χ2 tests to compare baseline characteristics between readmissions and index hospitalizations. Stata ranking commands were used to identify the most recurrent diagnoses associated with 30-day readmissions. We used multivariate Cox regression analysis to identify independent predictors of readmissions.

Results: Out of 3,472 index hospitalizations analyzed, 2,641 were discharged alive. Five hundred ninety-three (593; 22.5%) readmissions occurred within 30 days. Lung involvement in SLE was the most common reason for readmission (137; 23.1% of readmissions). Approximately 31.9% (189) of readmissions were due to other SLE-related complications. Readmissions were associated with higher inpatient mortality (18 (3.1%) versus 43 (1.6%); P=0.022), longer hospital stay (8 versus 5.2 days; P<0.001), younger mean age (26 versus 31 years; P=0.010), higher mean hospital costs (US $84,830 versus $64,628; P<0.001), and higher prevalence of heart failure (146 (24.6%) versus 526 (19.6%); P=0.024), CKD (435 (73.3%) versus 1,573 (58.6%); P<0.001), and anemia (138 (23.2%) versus 432 (16.1%); P=0.003) compared with index hospitalizations. Age ≥60 years (adjusted hazard ratio (AHR): 1.22; P=0.028), multiple (≥3) procedures during the initial admission (AHR: 2.57; P=0.003), discharge AMA (AHR: 1.68; P=0.047), lack of insurance/self-pay (AHR: 1.23; P=0.034), another coexisting autoimmune disorder (AHR: 1.19; P=0.041), index hospitalizations in the highest income quartile (AHR: 2.05; P=0.006), hyperlipidemia (AHR: 1.89; P=0.026), coexisting kidney disease (AHR: 1.56; P=0.017), and heart failure (AHR: 1.11; P=0.031) were significantly correlated with 30-day readmissions.

Conclusions: SLE lung readmissions were associated with worse outcomes than index hospitalization. Age ≥60 years, multiple procedures, discharge AMA, lack of insurance, kidney disease, and heart failure are significant predictors of readmission.