Summary:Alcoholism is a common disease; it is found in 10% to 15% of all patients admitted to general hospitals. There is no single characteristic finding, but on the other band, changes s compared with normal values have been reported in the literature for more than 30 frequently assayed clinical chemical and haematological parameters. In the project reported here all 24 clinical chemical parameters and all 8 haematological parameters frequently assayed were studied in each of 82 hospitalized inen with a confirmed diagnosis of alcoholism.The diagnosis of alcoholism was made on the basis of the Munich Alcoholism Test (MALT) together with the following standardized assessments and examinations: past history, an alcohol questionnaire, general physical examination and neurological examination. All forms were filled in completely.All Steps in the clinical laboratory investigations were standardized, and all were subject to ongoing reliability control. The clinical problem is usually not to differentiate alcohol abusers or alcoholics from healthy persons but rather to identify the alcoholics among a population of patients with a variety of illnesses.For this reason 70 patients from two hospitals who were clearly neither alcohol abusers nor alcoholics were studied in exactly the same manner s the alcoholics.In this combined group of 152 hospitalized patients significant differences were found in the distribution of the values for the alcoholics and the non-alcoholics for the following clinical chemical and haematological parameters: at the 0.1% level γ-glutamyltransferase, aspartate minotransferase, urea, creatinine and mean corpuscular volume (MCV), and at the 1% level glutamate dehydrogenase, alanine aminotransferase and alkaline phosphatase. From these eight parameters those combinations of between two and six parameters were selected that discriminated best between the alcoholics and the non-alcoholics. Using conventional decision limits the following was found: Fpr the alcoholics two or more of the res lts for the following five parameters were outside the decision limits given in parentheses: γ-glutamyltransferase (^28 U/l), aspartate aminotransferase (^18 U/l), alanine aminotransferase (^22 U/l), MCV (^96 fl), creatinine (^66.3 μπιοΐ/ΐ). The diagnostic sensitivity (alcoholics) is 85%, the diagnostic specificity (non-alcoholics) is 64%.For the non-alcoholics three or fewer results were outside the decision limits given in parentheses: γ-glutamyltransferase (<28 U/l), aspartate aminotransferase (<18 U/l), MCV (<96 fl), creatinine (>66.3 μιηοΐ/ΐ), urea-N (>5.0 mmol/1 With the decision limits given above and this combination of parameters the diagnostic specificity (non-alcoholics) is 96% and the diagnostic sensitivity (alcoholics) is 50%.This procedure thus enables the detection and exclusion of alcoholism in men on the basis of clinical laboratory findings. By optimizing the decision limits, the diagnostic test criteria sensitivity, specificity and efficiepcy can be iniproved. Optimized decision limits for the di...