Quadratic multiple discriminant analysis of 25 commonly ordered laboratory tests resulted in correct classification of 100% of nonalcoholics without overt liver disease, 98% of alcoholism treatment program patients with presumed mild liver involvement, 96% of alcoholics with liver disease, and 89% of nonalcoholics with liver disease. Direct comparison of the biopsy-verified alcoholic and nonalcoholic liver disease groups resulted in 100% discrimination, and removal of traditionally evaluated liver tests from the battery of 25 tests did not substantially alter the original classification accuracy. Alcoholic and nonalcoholic liver disease was still 100% differentiable when equated for number of patients with cirrhosis, hepatitis, and hepatitis combined with cirrhosis or fibrosis. Additional utility of the quadratic discriminant approach was demonstrated when 83% alcoholic and 83% nonalcoholic liver disease cases were diagnosed correctly in a prospective manner. In contrast, use of aspartate aminotransferase to alanine aminotransferase ratios (ie, SGOT to SGPT) identified correctly 75% and 33% of patients, respectively.
Ryback RS, Eckardt MJ, Felsher B, Rawlings RR. Biochemical and Hematologic Correlates of Alcoholism and Liver Disease. JAMA. 1982;248(18):2261–2265. doi:10.1001/jama.1982.03330180021025
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: