Blood ethanol and serum osmolality in the trauma patient.

The relationships between blood alcohol concentrations and both "osmolar difference" (measured serum osmolality minus expected serum osmolality calculated from serum sodium, serum potassium, blood urea, and blood glucose concentrations) and total serum osmolality have been investigated in severely traumatized patients. Osmolar difference was found to be more strongly correlated to blood alcohol concentration than serum osmolality; however, the error in the slopes of the regression lines was essentially identical (approximately 17%). Consequently, indirect estimations of blood alcohol concentrations by calculating osmolar difference do not improve the accuracy obtained by the estimation by measurement of serum osmolality alone. It appears that the production of unidentified osmoles in traumatized patients significantly influences all indirect calculation of blood alcohol by osmometry. A direct determination of blood alcohol should be made whenever possible.