Injury case-control studies using "other injuries" as controls.

Studies of event-related (or "event-phase") interventions (such as ski helmets) can address injury at a specific body site (such as the head) by using as controls a group of people who experienced the same event (fall) and suffered injuries at other body sites (other than the head). The research question addressed by this type of study is the effect of an exposure or intervention (helmet) during the event phase (fall) of the causal chain. However, this is a valid case-control design only if the controls (skiers with other injuries) provide a reasonable proxy for the prevalence of exposure (helmet-wearing) in the underlying event-phase source population (skiers who fell). This assumption needs to be carefully assessed. Factors associated with both helmet-wearing and injury given a fall (eg, previous injury history, skiing inexperience, or risk-taking behavior) have considerable potential to create bias.