Qualitative and quantitative indices of mild hepatic encephalopathy

Background and Aims: The diagnosis of covert/mild overt HE remains a matter of debate. While practices are becoming more homogeneous, the semi-quantitative assessment of overt HE is not necessarily performed, and the interpretation of results when different neuropsychiatric tests are utilized remains poorly defined. The aim of the present study was to assess the impact of a qualitative approach to clinical HE diagnosis, and to describe the features of patients being diagnosed with covert HE on abnormalities in either the Psychometric Hepatic Encephalopathy Score (PHES) or the electroencephalogram (EEG). Method: 411 patients evaluated in our dedicated HE clinic between April 2009 and June 2023 [292 males (71 %), 60 ± 10 years, MELD = 14.4 ± 5.9, MELD-Na 18.6 ± 16.1] were included. Patients were qualified as: 1) unimpaired, when they were clinically normal and both the PHES and the EEG were normal; 2) having covert HE when they were clinically normal but the PHES and/or the EEG were abnormal, or 3) having overt HE based on the semi-quantitative modification of Conn’s criteria (Vilstrup et al., J Hep 2014). Patients were also classified as having/not having overt HE based on a qualitative impression of the clinician, prior to any formal assessment. Results: 137 (33 %) patients were unimpaired, 174 (42 %) had covert HE and 100 (25 %) overt HE; 122 (30 %) were qualified as having overt HE on qualitative assessment. Of the 100 patients with overt HE, 30 % were missed on qualitative assessment. In addition, 17 % unimpaired/covert HE patients were erroneously qualified as having overt HE on qualitative assessment. Across the HE spectrum, patients with an abnormal EEG were older (61.5 ± 9.1 vs 58.1 ± 10.3, p < 0.001), had higher MELD (15.8 ± 6.1 vs 12.8 ± 5.2, p < 0.001), higher MELD-Na (21.2 ± 18.9 vs 16.1 ± 13.7, p < 0.01) and higher ammonia levels (76.2 ± 43.6 vs 54.9 ± 41.6 µmol/l, p < 0.001) compared with their counterparts with normal EEG. Patients with an abnormal PHES were older (60.8 ± 9.3 vs 58.9 ± 10.0, p < 0.05), had lower educational attainment (9.0 ± 3.3 vs 10.6 ± 4.1 years, p < 0.0001), higher MELD (15.7 ± 6.4 vs 13.4 ± 5.3, p < 0.001), higher MELD-Na (22.0 ± 20.0 vs 16.3 ± 12.5, p < 0.01) and higher ammonia levels (70.9 ± 48.1 vs 59.7 ± 38.9 µmol/l, p < 0.05) compared with their counterparts with normal PHES. Amongst patients with covert HE diagnosed on one abnormal test only, those with abnormal PHES had lower educational attainment compared to those with abnormal EEG (8.2 ± 3.1 vs 10.2 ± 4.1 years, p < 0.002). Conclusion: Qualitative clinical evaluation of mild HE is unreliable, with significant proportions of false negatives and, unexpectedly, also false positives. The EEG and PHES work well in this context, with both being affected by age, and PHES also by educational attainment, despite adjustment based on local norms.