Re: "Using head-on collisions to compare risk of driver death by frontal air bag generation: a matched-pair cohort study".

As reported in a recent Journal article, Braver et al. (1) studied head-on crashes of two passenger vehicles and estimated the risk of driver death for sled-certified air bags compared with first-generation air bags. In doing so, they extended our analysis (2) by identifying which vehicles were certified by sled tests (rather than using model year as a marker, as we did) and by including pickup trucks, sport utility vehicles, and minivans as well as cars. However, Braver et al. stated that our study (2) ‘‘estimated the impact of newer styles of frontal air bags on overall risk of front-occupant deaths in motor vehicle crashes rather than the specific effects on deaths in frontal crashes’’ (1, p. 547). In fact, we did report risk ratios for death comparing secondgeneration air bags with first-generation air bags in frontal car crashes in tables 2 and 3 of our published study: these risk ratios are summarized in table 1 of this letter. We defined a crash as frontal if the primary impact angle was 11, 12, or 1 o’clock, with 12 o’clock indicating the front of the car, because front air bags are designed to deploy within 30 degrees of the 12-o’clock position (3). Braver et al. (1) reported an adjusted risk ratio of 1.04 (95 percent confidence interval: 0.85, 1.29) for driver death in cars with sled-certified air bags compared with firstgeneration air bags. This value can be compared with our estimate of 0.93 (95 percent confidence interval: 0.79, 1.10) for all front occupants (table 1). We also reported estimates by seat position (driver or passenger) and restraint use (table 1). Letters to the Editor 885