To the Editor: Thompson and associates are to be commended for their carefully executed case–control study (May 25 issue),1 which attempts to establish the protective effect of bicycle safety helmets. Their study contains several sources of potential bias, however, that may limit the validity of the study results. Most important, 22.6 percent of Thompson's case patients (subjects with a head injury) had been involved in collisions with moving automobiles, as compared with only 12.5 and 3.9 percent in each of the two control groups. The subjects with a head injury were also less likely to have fallen onto soft surfaces….
All Science Journal Classification (ASJC) codes