California was one of the first states to mandate the use of bicycle safety helmets in the United States. On 1 January 1994, a piece of legislation that required bicyclists ages 17 years and under to wear helmets while riding on public bicycle paths and roads became effective. The age limitation created a setting for a case-control comparison between young riders and adults. The primary objective of this study was to examine whether the bicycle safety helmet legislation in California was associated with any statistically significant reductions in either or both types of head injuries among Youth riders. A case-control design was used to make direct comparisons between the two age groups across the two time periods. Since a direct measurement of risk exposure was not available, the proportion of each of the three injury types to the total number of injuries was used as the study outcome measure. It was assumed that the use of proportions would control for major changes in exposure from annual variations of bicycle use in the general population. The relative proportions were expected to remain reasonably constant unless an intervention, such as a bicycle safety helmet law, was introduced in the study period. Thus the study objective was to detect any significant reductions in the proportions of head injuries among Youth bicycle riders associated with the helmet law. Two methods of analyses were applied to the injury data to meet the objective. The first was an aggregate data analysis approach using the Pearson chi-squared test for independence and odds ratios. This approach tested whether the relative proportions for the two age groups were significantly different across the two periods. The second method was a pooled disaggregate data fitting technique using Multinomial Logit (MNL) models. The MNL models examined the likelihood of each injury type outcome before and after the legislation and allowed for interactions between the age and year variables with the other ones to measure the explanatory power of each variable for the outcomes. Conclusions about the California bicycle helmet legislation were drawn from the results of these two methods. For the covering abstract see ITRD E136182.
Samenvatting