S2854 Improving Hospital Readmissions for Upper GI Bleed Among Cirrhosis Patients

Introduction: Cirrhosis is a chronic condition which is associated with a high mortality rate. The high rates of patients with decompensated cirrhosis contributes to the rising healthcare costs and is independently related to mortality. Although the readmission rates among patients with decompensated cirrhosis is not extensively described in the literature, there are studies that elucidate predictors of readmission. One study by Morales et al showed that the rate of early hospital readmissions (<30 days) for patients admitted for decompensated cirrhosis was 29.5%. This study revealed that predictors of early readmission among this patient population are: male gender, Model for End-Stage Liver Disease-sodium score (MELD-Na) ≥15 and Charlson index ≥7. Methods: This is a single-center cohort study with the aim to develop an intervention protocol to reduce 30-day readmission rates among cirrhosis patients admitted for upper gastrointestinal bleed (UGIB). This study compares retrospective analysis with data from a prospective intervention. The prospective component involves closely monitoring decompensated cirrhosis patients admitted for UGIB, ensuring they receive appropriate inpatient care (including GI referrals) and that outpatient follow-up appointments with the Liver Diagnostic Clinic (LDC) are scheduled prior to discharge. The retrospective component consists of a chart review to assess early readmission rates among cirrhosis patients and serves as a control group for our study. The primary outcome of this study is to assess 30-day readmission rates for UGIB among cirrhosis patients. Results: The prospective portion includes n = 20 and n = 102 for the retrospective portion. In the retrospective phase, 42.2% of patients had an appointment scheduled in LDC upon discharge and 58.1% attended. The 30-day UGIB readmission rate for this group was 8.8%. In the prospective phase, 80% of patients had an appointment scheduled, and 50% of patients attended their appointments. The 30-day UGIB readmission rate for this group was 5%. Conclusion: Our preliminary data suggests that there is a 3.8% difference in 30-day readmissions between the retrospective and prospective groups. Our limitations include a single-center design and a small prospective sample. Our findings support the potential benefit of structured transitional care interventions in reducing readmissions among high-risk cirrhosis patients.