Objective: To assess the utility of laboratory markers of severe alcoholism in a cross-sectional population of male trauma center patients, who have a high rate of current alcohol dependence (CAD).
Methods: A total of 684 men were assessed for CAD by using standard criteria, all of whom had complete laboratory data. The ability of tests to predict CAD was ascertained by using Student's t test, sensitivity, specificity, and area under the curve (AUC) of receiver operating characteristic curves.
Results: Mean values of five tests were individually associated with a CAD diagnosis. Gamma-glutamyltransferase, aspartate aminotransferase, and mean corpuscular volume had sensitivities less than or equal to 0.51 and AUC less than or equal to 0.67. Blood alcohol concentration (BAC) and serum osmolality had sensitivities of 0.64 and 0.74 and AUC of 0.74 and 0.76, respectively. Each marker, when combined with BAC, showed little improvement in AUC over BAC alone.
Conclusions: Laboratory tests are not highly predictive of CAD in male trauma patients. A combination of BAC tests and interview screens is suggested for use in this patient population.