CHAPTER 1
Background and Objective

Crash report data are the primary source for identifying crash types, involved persons and vehicles, contributing factors, and crash patterns and trends. Researchers and policy makers depend on crash report data to identify problems, select and evaluate countermeasures, and analyze the effectiveness of safety measures and policies designed to reduce motor vehicle crash frequency and severity. Despite their importance, crash report data are historical and are often incomplete or inaccurate because of misreporting (Imprialou and Quddus 2019). The term misreporting refers to a variety of issues, including underreporting, overreporting, errors in crash data recording, and misclassification.

Misreporting has a detrimental effect on policy makersʼ ability to develop effective road safety policies. Researchers have found that misreporting crash data makes it difficult to identify crash patterns, which in turn makes it difficult to develop safety countermeasures (Hauer and Hakkert 1988). Also, reporting errors lead to the misidentification of dangerous road segments, inaccurate projections of crash and fatality rate estimates, and the incorrect identification of factors contributing to crashes. Moreover, the calculation of road crash injury cost, cost–benefit ratios for evaluating safety interventions, and inferences made on the effectiveness of interventions are all affected by inaccuracies in crash injury data (Abay 2015; Watson, Watson, and Vallmuur 2015). In the past few decades, researchers have conducted extensive studies to identify the causes of misreporting and discover how to minimize the problem (Hauer and Hakkert 1988; Elvik and Mysen 1999; Ahmed, Sadullah, and Yahya 2019). Misreporting crashes is more common when impaired and distracted driving are involved (Miller et al. 2012; Arnold and Scopatz 2016).

The primary objectives of this report were to develop procedures to determine the extent of misreporting impaired and distracted driving and to develop a methodology to improve the reporting of impaired and distracted driving in motor vehicle crashes. The guidance and methodologies developed in this research will help states and jurisdictions identify the misreporting of impaired and distracted driving and improve crash data collection and analysis. Several factors contribute to the misreporting of impaired and distracted driving crashes, including crash report limitations and variability, law enforcement officer (LEO) practices in completing the crash reports, challenges associated with incorporating the test results, and states treating the crash report as a legal document admissible in court.

Law enforcement–reported crash reports are the primary data source for crashes and the data elements included, and their associated data quality directly affects the misreporting of distracted and impaired crashes. The Model Minimum Uniform Crash Criteria (MMUCC) are voluntary guidelines to encourage uniformity among states by providing a minimum set of crash data elements, along with their definitions and attributes. The 5th edition of the MMUCC, released in 2017, contains several person-level crash elements with attributes to help states better classify distraction and impairment. The MMUCC, 6th edition, was released during the period of

performance for BTSCRP Project BTS-20, “Strategies to Address Misreporting of Impaired and Distracted Driving in Motor Vehicle Crashes”; however, examples drawn from earlier versions of the MMUCC remain relevant. Element P18, “Distracted By” captures “distractions that may have influenced driver/non-motorist performance, involving both an action taken by the driver/non-motorist and the source of the distraction.” Elements P19 through P23 capture suspected driver impairment and any tests conducted. While the MMUCC provides states with guidance on data elements and attributes, this guidance is often modified to suit individual state needs.

Some researchers believe that differentiation between distraction and inattention is not consistent across agencies. Nie et al. (2021) reported the well-known fact that law enforcement crash report forms and data elements vary from state to state. A National Safety Council (NSC) review of crash reports from all 50 states in 2017 found large discrepancies in the types of data states collected related to impaired driving (NSC 2017). Specifically, six states did not provide fields or codes to record blood alcohol content (BAC) values greater than 0.01 g/dL, 17 states did not provide fields or codes to record BAC values from all drivers and non-motorists, 47 states did not provide specific fields for drug test type, and 32 states did not provide fields for specific drug types identified. With regard to distracted driving, some states relied only on the narrative to identify distraction. Twenty-six states lacked fields to capture texting, and 32 states lacked fields to capture hands-free cell phone use.

Extant LEO practices in completing crash reports add another layer of complexity that needs to be accounted for. A common example is impairment testing practices. In states that do not distinguish between the nature of impairment, or where prosecutors choose not to apply enhanced penalties for drug involvement (in addition to alcohol involvement), officers may not investigate drug impairment if alcohol results are positive. If the driverʼs BAC is above the general legal limit of 0.08 percent, that may be as far as the official record goes unless law enforcement or medical professionals find a compelling reason to test further. In the hospital setting, doctors may suspect other drugs are responsible for a patientʼs condition and order additional tests to gain information that a LEO in the field may not or could not have considered. One reason is that typical blood- or urine-based drug tests are more expensive and take longer to process than breath tests for alcohol impairment. Another reason is that the per se limits for alcohol impairment are well established by research and case law, whereas drug impairment is a more complex situation medically and legally. The situation with drugs other than alcohol is also complicated by abuse of prescription medication, over-the-counter (OTC) medication, and an ever-evolving array of substances marketed as legal before states can regulate them.

Standards are most uniform for drugs that are not allowed in any detectable quantity (e.g., hallucinogenic narcotics) or for a particular class of driver (e.g., teens, commercial drivers). In those cases, laboratory tests at the limit of detection will be reportable and are more likely to lead to charges that would result in a record in one or more law enforcement reporting systems (e.g., crash record, citation, arrest), as well as in court and in the driverʼs department of motor vehicle history if the driver is convicted.

For other drugs or when a combination of drugs is in use, the completeness of the data depends on factors such as funding available for testing in the local jurisdiction, availability of a drug recognition expert to narrow the list of tests, severity of the crash, and presence of victims other than the impaired driver. Crashes under nearly identical circumstances can be treated differently depending on where the crash occurred and when, in the agencyʼs funding cycle, the crash took place. Most alcohol test results are from breath tests and are readily available to the officer completing the crash report. A blood or urine test can be conducted for alcohol as well. Typically, this happens with fatally injured drivers and passengers. Testing can also be done by request from drivers who cannot or do not want to blow into the breathalyzer or who are injured and immediately transported for medical care. Some statesʼ crash reports offer a “suspected

alcohol use” data element: if the blood or urine test result does not exceed the BAC threshold, the crash report will need to be amended, which does not always happen. Amendment is even more important for the suspected drug use data element, which gets coded; however, that suspicion may never be confirmed or denied in fact.

Blood or urine drug test results in crash reports also suffer from delays in reporting more than tests for the presence of alcohol. Drug test results come from a laboratory (or in some cases, a hospital) and may be significantly delayed or never actually received in time for analysis. Similarly, the drug test could be completed, but the officer never gets the information from the lab.

While this disconnect is more pertinent to drug tests, delays can happen with alcohol lab tests as well. Sometimes officers have difficulty obtaining drug or alcohol test results from the hospitalʼs labs. Officers may need to request that the hospital draw blood more than once for the state crime lab to test for different substances. Officers may also need to subpoena the hospitalʼs lab test results because they are patientsʼ private records and there is no medical necessity to report results to law enforcement. Under- and overreporting drug impairment can happen from both officers and hospitals if analyses are based solely on the initial officerʼs impression because laboratory confirmation is pending or missing from the record.

In states where the crash report is treated as a legal document and admissible in court, officers may be trained to avoid entering any information that they cannot support through their own observations, witness statements, or direct evidence. Even if the officer suspects that the driver was distracted, that information is unlikely to appear in the crash report, in any form, if the officer has concerns about being challenged in the legal system. Other jurisdictions specifically limit the ways crash reports can be used as evidence so that officers are more comfortable recording their opinions on crash-contributing circumstances, including driversʼ attention and level of distraction. While state crash report forms vary in content (e.g., how the officer can code for the presence of driversʼ actual interaction with sources of distraction), all of them have a section for the officerʼs narrative description where the officer could, if they choose, indicate their opinion about a driverʼs distracted state. Use of the data elements and narrative varies by state and by department within states.

In summary, impaired and distracted driving crashes are misreported for myriad reasons. The following chapters review the misreporting of impaired and distracted driving and the development of a methodology to improve the reporting of alcohol-involved, drug-involved, and distracted driving in motor vehicle crashes. A literature review and stakeholder survey results are presented first. Next, the report steps through the process of developing methods to estimate misreporting, supported by case studies using reporting data from Wisconsin, Connecticut, and Kentucky. Lastly, the report will present a summary of the research findings.