Effects of abstinence on the ability of clinical laboratory tests to identify male alcoholics.

Quadratic discriminant analysis of 25 commonly ordered clinical laboratory tests resulted in correct classification of 100% of nonalcoholic inpatients (n = 41), 100% of medical ward alcoholics as alcoholic (n = 63), and 95% of alcoholism treatment program participants as alcoholic (n = 274). Admission laboratory values for a random sample of 138 additional alcoholism treatment program participants were used to correctly classify 96% of them as alcoholics. Repeat analysis of these same laboratory tests after some 27 days of hospitalization resulted in 94% of the 138 patients still being classified as alcoholic, even though significant improvements in individual hepatic and hematologic laboratory values were noted. Classification was not changed with reanalysis after long-term abstinence (7 or 24 months), suggesting widespread and persistent alcohol-related alterations in organ system functioning with chronic alcohol abuse. Quadratic discriminant analysis of laboratory tests is an effective means of identifying people with present or past alcohol abuse but is not yet of value in detecting compliance with a treatment goal of abstinence.