Prediction of incident overt hepatic encephalopathy using individual and composite measures of minimal hepatic encephalopathy: The PrO-MHE Study.

INTRODUCTION AND OBJECTIVES: Psychometric hepatic encephalopathy score (PHES) is the gold standard for minimal hepatic encephalopathy (MHE), predicts overt HE (OHE), but remains cumbersome for routine care. Point-of-care (POC) tests such as Stroop (EncephalApp) and Animal Naming Test (ANT) are simpler but assess narrower cognitive domains. We evaluated whether integrating POC domains into a composite score could improve the prediction of incident OHE. PATIENTS AND METHODS: 153 outpatients underwent PHES, Stroop, and ANT. The primary endpoint was incident OHE within 180 days; death and liver transplantation were competing events. Cognitive predictors were analysed using Fine-Gray competing-risks regression with MELD and prior OHE as prespecified covariates. A composite cognitive risk score (CCRS) was derived from the principal component of standardized Stroop and ANT scores and rescaled from 0 to 10 RESULTS: MHE was present in 48.4% of cases. During follow-up, 29 (18.9%) developed incident OHE. In univariable models, PHES, Stroop time, and the CCRS were associated with incident OHE, whereas ANT was not. In multivariable models adjusted for MELD and prior OHE, PHES was the strongest predictor of incident OHE (sHR 0.77; 95% CI 0.66-0.90; p=0.001).Stroop (per 10-second prolongation) independently predicted OHE (sHR 1.05; 95% CI 1.03-1.07; p < 0.001). CCRS independently predicted OHE (sHR 1.37; 95% CI 1.04-1.82; p = 0.028) and pragmatically stratified patients into risk-based strata. CONCLUSIONS: In a competing-risks framework, PHES best predicts incident OHE. Among POC tests, Stroop outperformed ANT. A novel CCRS predicts OHE, risk-stratifies, but does not approximate PHES performance.