Are retrospectively reconstructed blood alcohol concentrations accurate? Preliminary results from a field study

Objective: The purpose of this study was to evaluate the relationship between blood alcohol concentration (BAC) derived from an in vivo breath test and retrospective estimates of BAC (eBAC) for the same drinking event. The relationship was expected to be lower at higher levels of BAC as a result of inaccuracies in the recall of self-report data. Method: The sample consisted of 44 persons (64% male) who were walking away from a bar district adjacent to a university campus. After giving informed consent, participants provided a breath sample and contact information. The next day a research assistant telephoned the participants and conducted a brief interview, using the equation prescribed by the National Highway Traffic Safety Administration, to obtain the self-report data necessary to calculate the eBAC for the previous evening. Results: The actual BACs and eBACs were not significantly different, and they were highly correlated (r = 0.84, n = 43). BAC did predict the magnitude of discrepancy between the two, and at BACs over 0.08, the relationship was nonsignificant. Conclusions: Although self-report data can be used to approximate the BAC obtained during a naturally occurring drinking event, the relationship is moderated by measured level of intoxication. Retrospective calculations are less accurate when estimating higher BACs.