SHM Converge Scientific Abstract ePoster Gallery-Characterization of Patients With Liver Cirrhosis Readmitted Within 30 Days of Hospital Discharge: A Learning Health System Initiative

Characterization of Patients With Liver Cirrhosis Readmitted Within 30 Days of Hospital Discharge: A Learning Health System Initiative

Back to course

Pdf Summary

This study examined 2,261 patients with liver cirrhosis to identify factors linked to 30-day hospital readmission, with a focus on differences between mild and moderate-severe disease. Using electronic medical record data from December 2019 to March 2024, the authors applied generalized regression and a boosted tree ensemble model to evaluate which readily available clinical variables at presentation were most predictive of readmission.<br /><br />Among the cohort, 60% had mild cirrhosis and 40% had moderate-severe cirrhosis. The 30-day readmission rate was 7.9% in the mild group and 15.0% in the moderate-severe group, showing substantially higher risk with greater disease severity.<br /><br />Key predictors differed by severity. In patients with mild cirrhosis, the Elixhauser Comorbidity Index was the strongest predictor, highlighting the importance of overall comorbidity burden. In moderate-severe cirrhosis, hemoglobin was the most important predictor, suggesting anemia as a major risk marker. Other important variables across groups included kidney dysfunction markers such as BUN/albumin ratio, liver disease severity measures such as MELD-Na and CLIF-C AD scores, nutritional status indicators such as albumin/globulin ratio, and AST/ALT ratio, which may reflect underlying etiologies needing targeted treatment. Longer hospital stays relative to expected length of stay were also associated with increased readmission risk.<br /><br />The study concludes that early EMR-based identification of these factors could help clinicians stratify risk, improve discharge planning, and target transitional care interventions to reduce readmissions in patients with cirrhosis.

Asset Subtitle

Raja Goyal

Meta Tag

Author List Aman Khurana, Caroline Wojtas, David Jerez, Gustavo Ramos, Joel Baker, John Korin, Karen Hamad, Manuel Rosario, Raja Goyal, Robert A. Smith, Shivani Patel

Category Research

Concept Cirrhosis

Concept Hospital Readmission

Concept 30-Day Readmission

Concept Predictor

Concept Disease Severity

Distinguished Non-Finalist

Presenter Organization Sarasota Memorial Hospital/Florida State University School of Medicine

Presenting Author Raja Goyal

Track Quality Improvement

Keywords

liver cirrhosis

30-day readmission

electronic medical records

risk prediction

mild cirrhosis

moderate-severe cirrhosis

Elixhauser Comorbidity Index

hemoglobin

MELD-Na

CLIF-C AD score

Cirrhosis

Hospital Readmission

30-Day Readmission

Predictor

Disease Severity