Casual blood-ethanol estimations in patients with chronic liver disease.

Patients attending a clinic for diseases of the liver were tested for blood-ethanol by a gas chromatographic technique sensitive to about 5 mg/dl (1 mmol/1). Of 172 patients (51 men, 121 women) 36% gave a history of heavy drinking (greater than 80 g ethanol/day; equivalent to 8 fl oz of whisky or 1 litre of wine) and 13% had ethanol in the bloodstream at values of 8-400 mg/dl. 42 patients (24%) had the liver-biopsy changes of alcoholic liver disease, and 17 of these had ethanol in the blood at one time or another. Nearly half (22/49) of all patients admitting heavy drinking also had detectable blood-ethanol. In all cases but 1 where blood-ethanol was found, a drinking history was admitted on first attendance, and alcoholic liver disease was nearly always found on subsequent biopsy. Blood-ethanol and admission of drinking were most constantly found in association with alcoholic steatosis and hepatitis. Both features were less commonly present in cases of alcoholic cirrhosis. Only 1 patient of 22 with "cryptogenic" cirrhosis on biopsy was found to have both ethanol in the blood and an alcoholic history, although 5 had an alcoholic history alone. The value of serial blood-ethanol estimations in the treatment of alcoholics and the detection of relapses is demonstrated. The findings confirm the relatively low frequency of alcoholism as a contributor to cirrhosis in the United Kingdom. Alcohol does not seem a major cause of cryptogenic cirrhosis. Casual blood-ethanol estimation is a useful and objective adjunct to techniques of investigating diseases of the liver.