The analysis performed in the Compton et al. study omits the crucial helmet/impact speed interaction variable that captures the neck injury enhancing effects of helmet use. As a result, the author's produce estimates of the neck injury inducing (reducing) effects of helmets that are understated (overstated).
The resulting bias in their evaluation of the effect of helmets on cervical spine injuries is obvious from the illogical conclusion reached in their paper. In particular, the authors find the statistically significant result that the use of a helmet reduces the odds of cervical injury by 22% in motorcycle accidents. The authors do not delineate any mechanism that substantiates their main finding. Yet, they accept these incredulous results at face value.
To the best of my knowledge, no proponent of the effectiveness of motorcycle helmets has advanced an argument for how helmets protect the cervical spine. Even if a mechanism is advanced, no matter how farfetched, the omission of the critical helmet-impact speed interaction variable still produces results that systematically understate the potential for helmets to cause neck injuries. Thus, the authors' main conclusion expressed in the subtitle of their paper, "Debunking the Myth", is at best presumptuous.
Jonathan P. Goldstein is a professor in the Department of Economics at Bowdoin College in Brunswick, ME.