use of cellular telephones in automobile crashes (Redelmeier and Tibshirani, 1997). Case-crossover studies control for variability in individual characteristics and are uniquely capable of identifying transient risk factors.
Researchers also employ ecologic study designs in which the investigator does not assign the exposure and in which the unit of analysis is a group rather than the individual subject. Injury epidemiologists frequently utilize ecologic designs to evaluate laws and regulations. These natural experiments allow comparisons of jurisdictions with a law to those without it, for example, comparing firearm fatality rates in two cities with different firearm laws (e.g., Sloan et al. ); other times, a single jurisdiction after a law has taken effect is compared to the same jurisdiction before the law (time-series design). Time-series designs may use simple before-and-after comparisons, such as comparing fatal motor vehicle crash rates among young people before and after the passage of a law lowering the legal blood alcohol level for young drivers (e.g., Hingson et al. ), but can also incorporate approaches that account for background secular trends. Interpretation of ecologic studies is difficult because, in addition to controlling all the normal sources of bias, researchers must also deal with ecologic bias and the failure of ecologic analyses to account for the distribution of confounders at the individual level.
Finally, case studies are a traditional research design that have yielded important advances in understanding causes of injury but frequently are misused. Case studies are vital when attempting to elucidate the physical mechanisms for excessive energy transmission. For example, the mechanisms by which occupants are injured during motor vehicle crashes were determined by thorough case investigations. However, case studies can be used only to ascertain directly observable proximate causes of injury. One common misuse of case studies is for study of risk factors for injury that are not directly observable, particularly indirect human factors such as fatigue or alcohol. Studies with comparison groups are the only scientifically valid method of quantifying the extent to which indirect risk factors contribute to injury causation.
The committee strongly recommends the utilization of rigorous analytical methods in injury research. Collaborations between research centers are critical for assembling populations and cohort groups necessary for conducting large-scale randomized controlled trials, cohort studies, and case-control studies.
There is a dearth of funding for research training in injury prevention, except for occupational injuries. Although the latter is supported by the National Institute for Occupational Safety and Health (NIOSH), a unit of the Centers for Disease Control and Prevention with a long-standing commitment to pre- and