Motor vehicle crashes are the leading cause of death of children in developed and developing nations.1 Occupant safety in the form of seat belts, airbags, and child restraints is an important factor in decreasing this morbidity and mortality. For example, car seats are 71% effective in reducing infant deaths and 54% effective in reducing toddler deaths.2 Other road safety measures such as motorcycle and bicycle helmets have proven efficacy. Severe head injury is reduced by 67% by motorcycle helmets3 and by 85% by bicycle helmets.4 Passage and enforcement of mandatory use laws have contributed to significant reductions in traffic injuries and deaths.5

Traffic-related data have several different functions in the prevention of injuries. Firstly, traffic data provide estimates of the frequency and severity of traffic injuries as a cause of death, hospitalization, disability, and economic loss at international, national, regional, and community levels. Secondly, traffic data provide information about environmental, equipment, and personal risk factors that are important in generating countermeasures. Finally, traffic data provide information to determine the efficacy of interventions.

Two sources provide most of the information needed to study the impact of traffic injuries and trends in both developed and developing nations. The first provides data on mortality and morbidity derived from health professionals, which include death certificates, medical examiner reports, and hospital and emergency department records. The second general source of traffic injury data is crash reports from police.

In the USA, traffic data are entered at the municipal and state level on police accident report (PAR) forms. For example, data from these forms are used to generate the data for the Fatality Analysis Reporting System, a nationwide database of motor vehicle deaths and corresponding information. In addition, state Departments of Transportation use these crash data to generate state-specific reports. These forms are used by law enforcement to record, in part, which safety features were in use by the occupants, motorcyclists, and bicyclists during a motor vehicle crash. This information is, in turn, often used by traffic safety advocates to promote public awareness and enforcement campaigns to increase use of motor vehicle safety devices. The purpose of this study is to describe the safety device items currently included on state-based PAR forms and compare them with the traffic safety laws of these states. We expected that most of the safety devices that had a corresponding law would also be represented on the PAR forms.

METHODS

We collected PAR forms from all 50 states and the District of Columbia (states/DC) from the National Highway Traffic Safety Administration website.6 Updated forms were obtained from state-specific websites shortly thereafter. The report forms were examined to determine the presence and coding scheme for the following variables: airbags, seat belts, car seats, booster seats, motorcycle helmets, and bicycle helmets. Information on the traffic laws of the states/DC was obtained from the Insurance Institute of Highway Safety.7 The following variables were collected: seat belt enforcement status (either primary or secondary), motorcycle helmet law, bicycle helmet law, car seat law, and booster seat law. For motorcycle helmets, partial laws cover young riders and some adult riders (eg, 14–20 years of age). For bicycle helmets, laws varied requiring helmet use for persons 11, 15, 17 years of age and younger. For each state, we compared the presence of a traffic safety law with the presence of a code on the PAR form. Safety information was current to February 2007.

RESULTS

Table 1 shows the frequency of the laws by state/DC regarding the different traffic safety issues, along with the corresponding safety variables that were coded for in the PAR forms. Overall, 22 states/DC had a primary seat belt enforcement law, all states/DC had a car seat belt law, and 39 had a booster seat law. There were no states/DC that had a bicycle law that covered all riders, but 21 had a bicycle helmet law that covered some younger age riders. For the motorcycle helmet law, 21 states/DC had laws that covered all riders, 26 had laws that covered some riders, and four did not have a law requiring any rider to wear a helmet.

For the safety devices coded, 50 states/DC had a coding variable for seat belts, 50 had coding for airbags, and all 51 states/DC had a coding for car seats and for motorcycle helmets. Nine states/DC coded for booster seats. For bicycle helmets, 48 states/DC had a designation coded for helmets. Few states/DC had a separate code for bicycle helmet versus a motorcycle helmet, but whether the person in the accident was riding a “pedal cycle” or a motorcycle is coded for elsewhere in the PAR form.

DISCUSSION

The primary finding of our study is that most state PAR forms include variables on seat belt use and child restraint use, but most do not contain a variable to code for booster seats. In addition, many states/DC code for helmets, which includes bicycle helmets, but do not have a full or partial bicycle helmet law. To our knowledge this is the first paper to describe this important gap in the ability of state transportation agencies to collect and analyze these important safety device variables.

The Model Minimum Uniform Crash Criteria Guidelines provides recommendations for variables and coding schemes on PAR forms. There is no mandate to follow these voluntary guidelines, and the second edition from 2003 is scheduled for review in 2008.

The time line for revising PAR forms varies from state to state. It is not unusual that, as new state traffic safety laws are enacted, there is often a several year delay until those items are included on PAR forms. For example, Colorado updated their PAR form on 17 November 2005 and it went into effect on July 2006, yet that form is still not coding for booster seats, even though the law went into effect on 1 August 2004. Another example is that Connecticut passed a booster seat law in 2005, but as of February 2008, still does not have a booster seat variable on its PAR form. In addition, in Connecticut, the revisions of the PAR form are facilitated by a Department of Transportation traffic records coordinating committee which meets on an ad hoc basis. The last revision of the Connecticut PAR form was January 2001 with no firm deadline established to update and revise the current form.

National Highway Traffic Safety Administration recommends that each state undergo an assessment of their traffic records program at least every 5 years.8 We believe that the costs associated with revising PAR forms and providing the necessary training to complete the forms and enter the data is offset by improvement gained by understanding the safety device use patterns by a population of road users. We understand that, although safety device information is requested on the PAR form, police officers may not always fully or accurately complete the safety device coding variables. The completeness and validity of PARs vary substantially depending on the number of police officers in the area and the time they have available to investigate crashes.9 In addition, some laws have partial coverage—that is, they only apply to specific populations—adding another layer of complexity in collecting accurate information. Nevertheless, even with the limitations in the PAR form itself and in the vagaries of police officers completing the form, this information is critical to accessing the effect of state-based safety legislation.

The limitations of our study include the fact that state laws and PAR forms are being modified constantly, so it may be difficult to obtain an accurate assessment of the state of these variables. Nevertheless, our report provides a reasonable description of the situation.

CONCLUSIONS

Our study describes major deficiencies in safety device items included on state-based PAR forms. PAR forms are a critical source of data for injury prevention professionals worldwide. PAR forms do not reflect current traffic safety laws in the majority of states/DC coding. This deficiency hampers the efforts of the public health and public safety communities to adequately quantify the effect that these devices, and the laws surrounding these devices, has on the safety of the public. Revising the PAR forms in all states to include complete variables for safety devices should be an important priority, independent of the laws.