Assessment of Cirrhotic Patients by the EncephalApp Fails to Predict Low-Grade Hepatic Encephalopathy.

Abstract

Introduction: An early detection of low-grade hepatic encephalopathy (HE) is of high importance. The aim of the study was to compare a neuropsychological with a psychophysical test on the basis of the psychometric hepatic encephalopathy score (PHES) regarding effectiveness in diagnosing minimal HE (MHE).

Methods: In our prospective controlled observational study, we examined a total of 103 patients with liver cirrhosis for HE. The PHES, CFF, and EncephalApp were performed in all patients. Graduation was based on the result of the PHES. Patients without evidence for HE 1&2 according to the mental state (West-Haven criteria) with a PHES <-4 value points and no clinical symptoms were defined as having MHE. Patients were considered as HE0 when in the PHES none of the psychometric subtest results was abnormal or with a PHES ≥-4 value points. Patients with clinical symptoms were considered HE 1&2 patients. Different cut-off values were determined, and their specificity and sensitivity were calculated.

Results: Ninety-six of the involved patients had liver cirrhosis and 25 acted as a healthy control group. The ROC analysis for the classification resulted in an AUC of 0.806, with the highest Youden index for the cut-off time >224 s, for which the sensitivity was 82% and the specificity 75%. Cases of withdrawals were seen in 10.74% of all tested patients.

Conclusion: The EncephalApp distinguishes well between HE0 and MHE but has its limitations in grading higher forms of HE. Diagnosis using only the EncephalApp is not sufficient.