Objectives: The objective of this study was to estimate the number of crash deaths specifically attributable to alcohol-impaired driving, with a focus on various strategies for introducing vehicle-based solutions. If alcohol detection systems are standard in all new vehicles, how many lives could be saved in the near term, and how long will it take to essentially eliminate alcohol-impaired driving? Alternatively, if such systems are offered as an option, how many lives could be saved?

Methods: Fatal crashes in the United States during 2015-2018 were classified by the highest driver blood alcohol concentration (BAC) and the corresponding age category of that driver. Based on the estimates of relative risk (RR) for a given driver group, eliminating alcohol in the driver's blood should lower risk by the attributable proportion, 1 - 1/RR. Multiplying this quantity by the number of deaths for the driver group yielded the estimated number of lives potentially saved if the BACs were reduced to zero.

Results: Systems that restrict drivers with any BAC could prevent nearly 12,000 deaths per year, while systems that restrict BAC to less than 0.08 g/dL could prevent more than 9,000 deaths. Within 3 years of a mandate for vehicle-based alcohol detection systems, it is expected that the annual lives saved would be between 1,000 and 1,300. Within 6 years, it would be between 2,000 and 2,600 lives saved per year, and within 12 years it would be between 4,600 and 5,900 lives saved per year. A system required only for those convicted of alcohol-impaired driving could save between 800 and 1,000 lives per year. A system available only to fleets of vehicles could save between 300 and 500 lives per year.

Conclusions: Vehicle-based interventions will not immediately solve the problem of alcohol-impaired driving, but they are an important component of the overall strategy. Even if required as standard equipment in all new vehicles, it will take 12 years before there are enough to reach more than half of their potential. If vehicle-based interventions are instead voluntary or introduced as options, then progress toward a solution will be much slower.