Lusk et al. respond.

We thank Schimek for his thoughtful letter. We agree that the American Association of State Highway and Transportation Officials (AASHTO) encourages bike paths; however, AASHTO’s definition of a bike path is a shared use path for pedestrians, skaters, wheelchair users, joggers, and other nonmotorized users. AASHTO discourages bicycle-only cycle tracks because, as noted by Schimek, AASHTO discourages bike lanes on the roadway separated from travel lanes by parked cars or raised barriers. Therefore, we agree with Schimek that the design of cycle tracks, such as those in Minneapolis, Minnesota, and New York City, are not included in AASHTO guidelines. As we discussed, our analysis suggested that these cycle tracks may have a higher crash rate than the other studied streets with cycle tracks because they are on busy urban arterials, which would include many intersections. At intersections, cyclist injuries increase with traffic volume, more specifically with the number of right and left turns by motor vehicles. In large cities, cyclists do need protection from motor vehicles, especially at intersections. Cycle tracks are associated with greater cycling and also, as concluded in a recent review of the literature, well-designed cycle tracks reduce cyclist injuries. We agree that three of the four comparison rates we cited are based on self-reports for specific populations. We had suggested that adequate comparisons for bicycle crash rates in US cities are lacking. Schimek generated a new bicycle injury rate for comparison, but it includes the overall estimated number of injured cyclists in the United States divided by the overall estimated length traveled by bicycle in the United States. We think that Schimek’s injury rate for the entire United States is not an appropriate comparison with the crash rate on studied cycle tracks, especially in urban New York City and Minneapolis. It is still noteworthy that Schimek’s injury rate estimate of 3.5 for all the states in the United States is higher than the overall crash rate of 2.3 that we found for the 19 cycle tracks across the United States. There is a need for more research on facility design, crashes, and bicycle activity in the United States. Our study does add to the best available evidence, which suggests that cycle tracks can reduce the risk of crashes while providing a more comfortable biking facility. We think it is time to change the AASHTO guidelines to improve roadways to take into account bicyclists needs, and to protect bicyclists from motor vehicles. j