Abstract
Screening for hepatic encephalopathy (HE) that does not cause obvious disorientation or asterixis (minimal HE [MHE]/grade 1 HE) is important. We examined if the animal naming test (ANT
) (maximum number of animals listed in 1 minute) is useful in this context. In total, 208 healthy controls, 40 controls with inflammatory bowel disease, and 327 consecutive patients with cirrhosis underwent the ANT
. Patients were tested for MHE by the psychometric HE score, and 146 were assessed by electroencephalography; 202 patients were followed up regarding the occurrence of overt HE and death. In the healthy controls, ANT
was influenced by limited education (<8 years) and advanced age (>80 years, P < 0.001). Using an age and education adjusting procedure, the simplified ANT
(S-ANT
) was obtained. An S-ANT
of <10 animals was abnormal. Of the patients, 169 were considered unimpaired, 32 as having HE ≥grade 2, and 126 as having MHE/grade 1 HE. This group had lower S-ANT
than unimpaired patients (12 ± 0.4 versus 16 ± 0.7, P < 0.001) and higher S-ANT
than those with HE ≥grade 2 (4 ± 0.9). In grade 1 HE the S-ANT
was lower than in MHE. Following receiver operating characteristic analysis (Youden's index), 15 animals produced the best discrimination between unimpaired and MHE/grade 1 HE patients. Thus, a three-level score (0 for S-ANT
≥15, 1 for 10 ≤ S-ANT
< 15, 2 for S-ANT
<10) was obtained. This score was correlated both to the psychometric HE score (P < 0.0001) and to electroencephalography (P = 0.007). By sample random split validation, both S-ANT
and its three-level score showed prognostic value regarding the 1-year risk of overt HE and death. No inflammatory bowel disease control had S-ANT <15.
The S-ANT
is an easily obtainable measure useful for the assessment of HE. (Hepatology 2017;66:198-208).