Comparison of 6 tests for diagnosing minimal hepatic encephalopathy and predicting clinical outcome: A prospective, observational study.

Abstract

Current guidelines recommend the assessment for minimal HE in patients with liver cirrhosis. Various efforts were made to find tools that simplify the diagnosis. Here, we compare the 6 most frequently used tests for their validity and their predictive value for overt hepatic encephalopathy (oHE), rehospitalization, and death. One hundred thirty-two patients with cirrhosis underwent the Portosystemic Encephalopathy-Syndrome-Test yielding the psychometric hepatic encephalopathy score (PHES), Animal Naming Test (ANT), Critical Flicker Frequency (CFF), Inhibitory Control Test (ICT), EncephalApp (Stroop), and Continuous Reaction Time Test (CRT). Patients were monitored for 365 days regarding oHE development, rehospitalization, and death. Twenty-three patients showed clinical signs of HE grade 1-2 at baseline. Of the remaining 109 neurologically unimpaired patients, 35.8% had abnormal PHES and 44% abnormal CRT. Percentage of abnormal Stroop (79.8% vs. 52.3%), ANT (19.3% vs. 51.4%), ICT (28.4% vs. 36.7%), and CFF results (18.3% vs. 25.7%) changed significantly when adjusted norms were used for evaluation instead of fixed cutoffs. All test results correlated significantly with each other ( p <0.05), except for CFF. During follow-up, 24 patients developed oHE, 58 were readmitted to the hospital, and 20 died. Abnormal PHES results were linked to oHE development in the multivariable model. No other adjusted test demonstrated predictive value for any of the investigated endpoints. Where applicable, the diagnosis of minimal HE should be made based on adjusted norm values for the tests, exclusively. The minimal HE tests cannot be equated with one another and have an overall limited value in predicting clinical outcomes.