Case Crossover Designs: Utility for Studying Child Injury Risk

A case crossover design is a methodology that is well-suited to studying the etiology of acute events, such as injuries (Maclure & Mittleman, 2000). Such designs are best utilized for events with an abrupt rather than an insidious onset. The design differs from a matched case-control approach in that each person in the study serves as their own control, thereby eliminating the influence of between-subject confounding variables (e.g., mental health). In a case crossover design, proximal factors (e.g., child activity level) preceding an acute event (e.g., injury) can be compared to factors preceding a time-matched non-event (e.g., a non-injury) for the same person in order to isolate variables that lead to an event occurrence. The authors will present results from a secondary data analysis of a study using a case crossover design to examine risk of unintentional childhood injury. Immediate risk of unintentional injury was predicted by two proximal factors, including caregiver type (i.e., mothers versus fathers) and caregiver supervision level. Caregivers reported on circumstances preceding injury events as well as circumstances preceding non-injury events that were matched by day and time to injury events. Results revealed that children were at increased risk for injury when supervised by fathers and that mothers' supervision was a stronger predictor of injury than fathers' supervision. The case crossover design allowed us to examine proximal risk for injury while controlling for several between subjects variables that may influence child injury risk (e.g., child impulsivity).