*Purpose: Inpatient frailty assessment and its utility in clinical care is understudied in liver transplant (LT) candidates. Study objectives were to quantify physical frailty among patients admitted for cirrhosis-related complications and identify whether frailty predicted mortality and 30-day readmissions.

*Methods: These are data from a prospective, multi-center cohort study ongoing at two large transplant centers in Pennsylvania. Frailty was assessed using the liver frailty index (LFI), a validated measure in cirrhosis, by trained research staff or physical therapists during inpatients cirrhosis admissions prior to hospital discharge. Age, sex, race, clinical variables, medical comorbidities, admission reason, MELD-Na prior to discharge, death, transplantation, and readmission data were obtained from the electronic health records. Multivariable Cox proportional hazards models were fit for mortality with LT as a competing risk; logistic regression models were fit for non-home discharge with total LFI score as the exposure variable.

*Results: A total of 149 patients had LFI assessments (n=87 UPMC, n=62 UPenn). Median age was 59 [IQR 44-65), median MELD was 22 [IQR 16-33]; 42 (28%) were waitlisted for LT. A total of 71 (48%) of patients had alcohol liver disease, 39 (26%) had NASH, 13 (8.7%) had HCV. The most common reasons for the index admission were volume overload/ascites (23%) and hepatic encephalopathy (17%); 105 (70%) had an inpatient paracentesis. The median LFI score was 4.7 [IQR 4.2-5.5], which was above the 4.5 frailty threshold; 94 (64%) of patients were frail, 48 (32%) were pre-frail, and 7 (4.7%) were robust. A total of 20% required rehab or nursing home upon discharge. Univariable factors associated with frailty were older age (OR 1.04, 95% CI: 1.01-1.07 per year), paracentesis (OR 2.6, 95% CI 1.2-5.3), and chronic kidney disease (OR 4.00, 95% 1.8-9.2). After median follow-up of 6 months [IQR 3-8 months], 16 (10.7%) patients died, 47 (31.5%) were transplanted. Total LFI score was associated with increased odds of non-home discharge (OR 2.1, 95% CI 1.32-3.33) and increased mortality (sHR: 1.7, 95% CI 1.02-2.94) after adjustment for MELD. An LFI threshold of 5 was associated with significantly higher mortality.

*Conclusions: Frailty measured by LFI predicted mortality after cirrhosis-related admissions independent of MELD and nonhome discharge. Inpatient frailty assessment prior to hospital discharge can be used for risk-stratification and targeted interventions to improve physical fitness, physical activity, and nutrition. Larger prospective studies are needed to validate these relationships.

« Back to 2020 American Transplant Congress