Exploring New Drug Signals of Abuse in the Human Laboratory

Driving while under the influence of drugs (i.e., drugged driving) has been identified as a national research priority by both the National Institute on Drug Abuse and the Office of National Drug Control Policy. The impairing effects of alcohol on driving performance have been studied for decades, and its pharmacodynamic and pharmacokinetic relationship with driving performance are well understood. In contrast, comparatively few controlled studies on driving while under the influence of psychoactive drugs have been conducted, and, for some drug classes, no data are available. Importantly, the National Highway Traffic Safety Administration reported that a full 33 percent of fatally injured drivers tested positive for at least one drug. Furthermore, ~10.6 million people reported driving under the influence of illicit drugs at least once in the year preceding the National Survey on Drug Use and Health (2010). A primary aim of the proposed studies is to examine the effects of widely used and abused prescription drugs on driving performance, including drugs from the opioid and benzodiazepine classes. Mu opioid agonists are now the most widely prescribed class of drug in the United States, with sales increasing dramatically over the past 15 years. Along with this increasing availability, a prescription opioid epidemic has grown that is associated with substantial morbidity and mortality. Few studies have examined the effects of commonly prescribed opioids on driving performance. Benzodiazepines are also widely prescribed for their benefits in treating anxiety and sleep disorders. However, benzodiazepines are also widely abused and misused for their sedative actions either alone or in combination with other agents. The proposed studies will employ a sophisticated driving simulator (NADs miniSim™) to characterize the impairing effects of these common drugs of abuse (along with other key test conditions) in a tightly controlled laboratory setting. While high-performance simulators can yield a wealth of data (e.g., the output for this model can monitor more than 60 driving performance variables), the importance of thoughtful variable selection and data reduction to ensure the relevance and generalizability of data are critical. Outcome selection here has been guided by the consensus recommendations from the International Council on Alcohol, Drugs & Traffic Safety (ICADTS) for behavioral outcomes in drugged driving studies and will include reaction times, steering responses, lateral control measures, and obstacle avoidance strategies. These studies will evaluate dose response relationships by examining multiple active doses spanning the therapeutic-to-supratherapeutic range in order to model outcomes for both prescribed drug taking and drug abuse/misuse. Moreover, both negative (placebo) and positive (alcohol) control conditions will be included in order to provide a framework for estimating degree of impairment and associated risk. Alcohol is the ideal positive control as it has been fully evaluated, and laws defining exposure/impairment relationships have been established based upon this wealth of data. These studies will produce a careful characterization of driving performance/impairment in a controlled laboratory setting while simultaneously collecting an array of clinically relevant multi-dimensional assessments (i.e., safety outcomes, self-reported drug effects and other psychomotor cognitive assessments [e.g., roadside-type sobriety testing]. Moreover, once these studies are completed, this system will allow for collection of pilot data for other agents of interest (e.g., marijuana) and examination of drug interactions identified as relevant to U.S. public health and safety by key surveys, such as state and national roadside testing studies. In summary, these studies will use state-of the-art technology to yield new empirical data on drugged-driving performance following exposure to two very widely prescribed and abused classes of agents (opioids and benzodiazepines). These findings will be highly r