Somatosensory evoked potentials in subclinical portosystemic encephalopathy: a comparison with psychometric tests

We prospectively studied the role of somatosensory evoked potentials (SEPs) and psychometric tests in the assessment of subclinical portosystemic encephalopathy (PSE) in 61 cirrhotic patients with grade 0 PSE and 20 controls. Six additional uneducated controls underwent only psychometric tests. Median nerve-evoked cortical responses were recorded for N20-N65 interpeak latencies (IPLs). Psychometric tests were conducted within 4 hours of SEP testing. Seven (26.9%) controls and 30 (49.2%) cirrhotic patients had abnormal psychometric test results (writing sample tests, 20; five-point star tests, 17; number-connection tests, 19; and following-a-track tests, 18); seven controls and 25 cirrhotic patients had 6 or fewer years of education; 7 controls and 28 cirrhotics were older than 50 years of age. Cirrhotic patients (47.6 +/- 8.3 milliseconds) had higher N20-N65 IPLs than controls (40.2 +/- 3.0 milliseconds; P < .001). Twenty-nine (47.5%) cirrhotic patients had abnormal N20-N65 IPLs. Of the 26 cirrhotic patients with more than 6 years of education, 15 (57.7%) had abnormal N20-N65 IPLs and 5 (19.2%) had abnormal psychometric test results (P = .005). Our data show that N20-N65 IPLs of SEPs are helpful in the assessment of subclinical PSE; 47.5% of cirrhotic patients had subclinical PSE. Poorly educated and older subjects tended to have abnormal psychometric test results. SEPs were not affected by education and age and were more sensitive than psychometric tests in the assessment of subclinical PSE in better-educated cirrhotic patients.