Using calculations to estimate blood alcohol concentrations for naturally occurring drinking episodes: a validity...

The purpose of this study is to determine the accuracy of retrospective estimated blood alcohol concentrations (eBACs) for re-creating intoxication resulting from a naturally occurring drinking event. This study identified five eBAC equations, applied them to self-report data and compared the results to actual blood alcohol concentration obtained by a breath test.

A convenience sample of 109 drinkers was recruited near drinking establishments and asked to provide breath alcohol concentration (BrAC) samples. Participants were contacted by telephone on the next waking day to gather data required for five algorithms that determine eBAC. BrAC and eBAC obtained from each equation were compared to determine the level of agreement between the two approaches.

eBACs correlated highly with each other (r > or = 0.99); R2 for all algorithms ranged from 0.54 to 0.55 with BrAC as the criterion. On average, eBAC equations overestimated BrAC. Regression analysis identified the amount of time spent drinking, number of standard drinks, weight and year in school as factors related to discrepancy.

These data indicate that, although all equations produce eBACs that are highly related, their relationship to BrAC does vary across equations. Using the best fitting equation, eBAC is more strongly correlated with BrAC when intoxication is less than 0.08 g/210 L of breath, and the magnitude of the relationship decreases as intoxication rises.