Generation of an osmolality gap-ethanol nomogram from routine laboratory data

Abstract

To develop a practical procedure for producing a nomogram describing the relationship between the osmolality gap and serum ethanol concentrations that is appropriate for the formulas and analytical methodologies in use at any laboratory. Retrospective review of serum sodium, glucose, urea nitrogen, ethanol, and osmolality values and the use of spreadsheet software in the generation of a nomogram. Emergency department in a university-affiliated county hospital. Two hundred three adult patients presenting with signs of alcohol intoxication. When either a common criterion (gap [mosmol/kg H2O] = ethanol [mmol/L]; range, -10 to +10) or a published nomogram was used for determination of consistency of the osmolality gap with the ethanol level, large percentiles of data were classified as inconsistent. Regression analysis and the method of truncated normal distributions were used to prepare a nomogram that better represented the data collected with our formulas and analytical methodologies. Interlaboratory analytical bias and the choice of the formula used for calculated osmolality limit the utility of published formulas or nomograms for the assessment of excess osmolality gap. Any laboratory can produce their own nomogram from routine patient data using the described spreadsheet procedure, without the need for measurement of volatiles other than ethanol.