Hepatic encephalopathy predicts increased inpatient mortality in decompensated nash cirrhosis

Introduction

The demographics of liver disease are changing, with rising levels of non-alcoholic steatohepatitis (NASH) and alcohol-related liver disease (ArLD). Cirrhosis complications account for a large proportion of associated early mortality. There is limited published data describing mortality patterns in contemporary clinical populations with decompensated cirrhosis. The aim of this study is to investigate the predictive factors of increased in-hospital mortality in this cohort.

Methods

Retrospective data were collected on consecutive patients admitted with decompensated cirrhosis at a tertiary liver centre (August 2021-December 2022). Demographics, disease aetiology, cirrhosis complications and mortality data were recorded and descriptive analyses undertaken. Univariate analyses using chi-square and Mann Whitney U-test and a multivariate binary logistic regression analysis using SPSS software were undertaken to identify predictors of inpatient mortality in the overall cohort and in subgroups based on aetiology of disease and ethnicity.

Results

217 acute admissions with decompensated cirrhosis were identified. Median age at admission was 53 years (interquartile range 15); men were 57%. Main disease aetiologies were ArLD (71%), NASH (14%) and HCV (8%). Inpatient mortality was 22%. There was no significant difference in age, gender or disease aetiology between those who died in hospital and those who survived to discharge. There were significantly higher levels of hepatic encephalopathy (HE) (77%;27%, p < 0.001), jaundice (83%;61%, p < 0.001), renal dysfunction (79%;27%, p < 0.001) and infection (75%;43%, p < 0.001) in those who died compared to those who survived. On multivariate regression analysis, odds ratios (OR) for inpatient mortality reached significance (p < 0.001) for HE (OR 7.03 [95% confidence interval 2.75–17.96]) and renal dysfunction (OR 5.91 [2.20–15.90]). HE was the only cirrhosis complication reaching statistical significance in all subgroup analyses and was highly predictive of inpatient mortality in NASH patients (OR 16.24 [1.27–207.09], p = 0.03) versus non-NASH (OR 5.90 [2.48–13.98], p < 0.001) and in patients of South Asian ethnicity (OR 13.6 [2.06–90.02], p = 0.007) versus all other ethnicities (OR 6.31 [2.38–16.68], p < 0.001).

Conclusion

HE, jaundice, ascites, renal dysfunction and infection occur significantly more frequently in those who die in hospital than those who survive to discharge. HE is the cirrhosis complication most highly predictive of inpatient mortality and this is most marked in those with NASH cirrhosis and of South Asian ethnicity. This data can help identify those at highest risk of in-hospital mortality, guiding earlier treatment escalation as well as palliative care involvement when appropriate.