Time-to tertiary care predicts mortality in alcohol-associated hepatitis in patients with cirrhosis

Background: Slovakia ranks first in the world in prevalence of decompensated cirrhosis, with alcohol-associated liver disease (ALD) being the leading etiology. The severe acute phenotype of ALD, alcohol-associated hepatitis (AH) is a distinct clinical syndrome characterized by recent onset of jaundice in the context of heavy alcohol consumption. The majority of patients with AH are admitted to the hospital. In our previous analysis of hospitalized cirrhotics, we found an increased mortality during the COVID-19 lockdown and postulated as main cause a pandemic-associated distortion of time-to tertiary cirrhosis care (TTT). In this proof-of-concept study we aimed to evaluate if the time-to-tertiary care (defined by the time spent in another hospital before referral to a tertiary care center) predicts mortality in patients with acute decompensation of cirrhosis triggered by AH. Methods: Since 2014, our tertiary hospital with liver transplant program has been running the registry RH7 of all the consenting adults admitted for advanced chronic liver disease (ACLD). We included patients from RH7 database between 07/2014 and 5/2020, with AH as the trigger of acute decompensation of cirrhosis who were admitted to the hospital. Severe AH was defined by MDF>31 or MELD>20;acute-on-chronic-liver failure (ACLF) was calculated by EASL-CLIF criteria. We divided the cohort to two groups: 1) patients admitted directly to a tertiary care center (Fig.1 - Central), and 2) patients referred to a tertiary care center after first being hospitalized in a lower-complexity hospital (Fig.1 - Subregional). We recorded demographics, clinical characteristics, time interval between the first and second admission (TTT= days spent in another hospital before admission to a tertiary care), and 30- and 90-day mortality. Results: Of 1,109 patients admitted for decompensated cirrhosis, we included 219 patients with AH decompensating cirrhosis, median age 50 years, 37% females;mean MELD and MDF were 23 and 46, respectively;76 % had severe AH. Subtypes of acute decompensation were pure acute decompensation in 53%, ACLF-1 in 24%, ACLF-2 in 18%, and ACLF -3 in 5%. One hundred and thirteen patients (52%) were admitted primarily to a tertiary care center (Central), 106 patients (48%) were referred from another hospital (Subregional). Median TTT was 15 days (1-95). Thirty- and 90-day mortality was significantly higher in Subregional as compared with Central group at 35.5% vs 11% and 51.4% vs 24.5%, respectively (p<0,0021) (Fig.1) Conclusion: Time-to-tertiary care is an independent predictor of mortality in patients with acute decompensation of cirrhosis triggered by AH. This difference in mortality is sustained for years after discharge.