The National Highway Traffic Safety Administration (NHTSA) has adopted a new method to estimate missing blood alcohol concentration (BAC) test result data. This new method, multiple imputation, will be used by NHTSA's National Center for Statistics and Analysis (NCSA) to improve the scope of alcohol involvement statistics generated by the Fatality Analysis Reporting System (FARS). The old estimation method used by NHTSA calculated the chance that a driver, pedestrian or a pedalcyclist with unknown or missing alcohol results had a BAC in each of the three categories: 0, 0.01 to 0.09, or 0.10 and greater. Beginning with the 2001 data, NCSA will use multiple imputation to estimate missing BAC values in FARS. Multiple imputation offers NHTSA significant advantages over the old method in analyzing and reporting estimates of alcohol involvement. Instead of estimating alcohol involvement by the three aforementioned categories, the new method will estimate BAC along the entire range of plausible values (0 to 0.94 g/dl). Estimating missing BAC this way will enable NHTSA to report the extent of alcohol involvement at any BAC level. This research note is intended to inform NHTSA's partners of this methodology change, and to compare estimates of alcohol involvement using the old and new methods, as well as to address anticipated questions.
Abstract