576 Relationship Between Hospital Facility Characteristics and End-Stage Liver Disease Readmission Rates in California

Abstract

infection on/during admission while 70% remained infection-free. During index hospitalization,77 died or went to hospice & 20 were transplanted. Therefore, 481 pts (age 56 yrs, 53% men,37% HCV) were available for analysis. Three month outcomes of these 481; 77% were alive without transplant, 11% were transplanted while 12% died. Readmissions: 50% of patients were readmitted at least once within 3 months; 56% had 1 readmission, 28% had 2, and 16% had ≥3 readmissions. Readmission rates were similar between those admitted with/without an infection (51% vs 50%,p=0.8) during the index hospitalization. Most readmissions were infection related (40%) followed by liver-related complications (17% anasarca, 11% GI bleed, 9% TIPS, 7% AKI, 5% HE) & 11% others. Two regression models were created; Model using index hospitalization day of admission values: A high MELD score (OR 1.1, p=0.001), lactulose use (corollary for HE; OR 1.6, p=0.05) & male gender (OR 1.6, p=0.05) predicted readmission while high serum albumin was protective (OR 0.57, p= 0.004). Infection as the reason for index admission, diabetes or admission WBC count did not impact 3-month readmissions. Model using index hospitalization discharge values: The only predictors of 3-month readmissions on discharge were MELD score (OR 1.1, p=0.002) and SBP prophylaxis use (OR 1.95, p=0.05). Nosocomial infections, diabetes, infections on admission and other medications were not significantly predictive. Conclusions: Readmissions occurred in 50% of hospitalized cirrhotics within 3 months of hospital discharge in this prospective multi-center study regardless of infections as their cause for index hospitalization. Without intervention, admitted cirrhotics are likely to require readmission, most commonly for infection related complications. Strategies to prevent hospital readmissions are urgently needed, especially in higher MELD patients with encephalopathy who have lower serum albumin and are on SBP prophylaxis.