ABSTRACT
Recently I talked with epidemiologists who were perplexed by their inability to find differences between injured and uninjured military trainees. It turned out that the young men in question had all been selected for an elite group of uniformly young, fit, intelligent males to be trained for especially hazardous and demanding tasks. The training programme was incredibly vigorous and had a high injury rate. But the search for personal risk factors that would predict which members of this homogenous group would be injured was doomed to failure, and distracted the investigators from performing a descriptive study of the circumstances of injury that would have had far greater potential for injury reduction. (A hypothetical example illustrating the same phenomenon would look at 100 identical siblings in a building that caught fire and killed half of them and then focus on personal risk factors rather than comparing injured vs. uninjured in terms of where they were when the fire broke out in relation to the point of origin of the fire and likely escape routes.)
Thus, we need to avoid the temptation to overemphasize personal risk factors when the environment and circumstances of injury dictate not only who is hurt but also the best means to prevent the injury.
