Opioids at the Health and Transportation Safety Nexus

Background: Drug overdoses and motor vehicle (MV) crashes are leading causes of unintentional injury death in the US, resulting in over 100,000 fatalities in 2017. Research has established that opioids affect driving ability and that crash-related injuries often result in opioid prescribing. Despite known associations, current approaches for studying these intertwined public health problems typically involve separate analyses using discrete databases. Purpose: To assess collection of relevant data elements and evaluate the linkage potential of prescription drug monitoring programs (PDMP) with crash databases, and to determine knowledge gaps that can be addressed through effective linkage. Methods: Standardized templates were used to abstract specific data elements and attributes of MV crash and PDMP databases for all 50 states and DC. Abstracted PDMP elements included accessibility of PDMP data and schedules of controlled substances monitored in each state, while crash-related elements included whether crash reports document the type of drug test administered at the scene and the granularity of test results recorded. Results: A majority of PDMPs (94%) are authorized to release data for research purposes. Schedules II-V controlled substances are tracked in 76% of PDMPs, with the remaining tracking II-IV. Drug-related elements captured in crash reports varied considerably by state. Eighty-six percent of states document the type of drug test administered; however, 54% of states only record whether a drug test was positive or negative, with less than a third of states citing specific drugs. Collection of personal identifiers is required in all crash and PDMP databases, suggesting high potential for effective linkage. Conclusions: Lack of integration between crash and PDMP databases hinders advancement of the evidence base on the interconnected causes of unintentional injury death. While crash reports and PDMPs possess their own sets of strengths and weaknesses, linkage of these two data sources could fill critical research gaps