Topic Area 1: What are the Effects of Combinations of Medicines on Functional Abilities Key to Safe Driving?


  • How should polypharmacy be defined for future NHTSA research? (number of medications, prescription only or include over-the-counter medications, herbal remedies, and alcohol).
  • What classes of medications and combinations should be the focus in future NHTSA studies, based on their effects on functional abilities key to safe driving?
  • How should the test sample be defined (e.g., % male to female; age groupings; and at what age to start: 55+, 65+, 75+, 85+)?

The following points were provided as additional guidance from NHTSA attendees in the meeting:

  • It is important to separate the effects of the medications from the effects of the illness for which the medications are being taken.  In some cases, the medications will improve driving performance and in other cases they will cause driving performance to deteriorate.
  • The timing of medication administration in relation to the timing of driving needs to be taken into account.  Since older drivers do most of their driving during the day, it may be of no consequence that certain medications are taken at night, if those medications have no residual effects on driving performance during the day.
  • NHTSA has an infrastructure to support the prosecution and adjudication of alcohol-impaired driving, but not for drug-impaired driving.  So if someone is using both medications and alcohol, and is pulled over by law enforcement, he or she will be pursued based on the presence of alcohol.  There are also practical limitations in detecting the presence of drugs by labs used by law enforcement, whereas the presence of alcohol is much more easily detected.  Since older people are not typically in the population apprehended for drunken driving, alcohol and medication interactions are not a focus in the current discussion.

Defining Polypharmacy and Identifying Drug Classes for Future Research

  • Although the literature review highlighted the fact that the definition of polypharmacy varies from the concomitant use of 3 to 10 medications, the brainstorming experts agreed that assigning a number is a meaningless starting point.  A person could have congestive heart failure and diabetes and be taking 10 medications appropriately. Contrast that person with someone taking an over-the-counter sleep aid and a long-acting benzodiazepine who would be much more impaired taking these 2 medications than the aforementioned person taking 10. And, taking two drugs could be less-impairing than taking one drug, depending on the class of the drugs; thus, the number of drugs is almost meaningless.  The focus on polypharmacy should be on multiple classes of drugs—not on the number of drugs per se.
  • Also, rather than a focus on numbers of medications, the focus should be on the effects of medications (the symptom), as the causal element of the driving impairment.  Focus on level of alertness (a mental effect), and categorize drugs into a hierarchy for risk of impairing one’s level of alertness.  This approach could make it possible to say when combining a low-risk medication will increase the impairing effects of the high-risk medications.  Also, drugs with motor (physical) effects such as dizziness, and those with visual effects should be considered. 
  • Level of alertness is a global parameter for cognitive function, as level of alertness affects all of the higher critical functions—if there is a decrement in alertness, there will be a decrement in the “executive” functions.  However, level of alertness may be fine at the same time there is a decrement in executive functioning (e.g., as in strokes, dementia).  The downside to limiting the definition of cognition to “level of alertness” is that the spectrum where alertness is impaired is relatively small, and there is a larger segment where more subtle deficits would show up and would be missed.  But the effects of the medications (the impairments) must be related to increased crash risk in order for the research to have relevance for NHTSA. 
  • Experts recommended that classes of medications to be considered in any research on the topic of polypharmacy and driving should include—but not necessarily be limited to—the following: alpha blockers (potential for severe hypotension; common in the older male population); anti-diabetic agents (potential for hypoglycemia); sedating types of drugs such as the benzodiazepines, tricyclic antidepressants, and possibly the opioids; drugs that could potentiate hypoglycemic effects; and anti-hypertensive agents (because of their hypotensive effects).  
  • Extraneous factors such as fatigue/sleep deprivation will also affect a person’s driving performance, and need to be accounted for in a study of the effects of polypharmacy on driving.  Also, length of time on medication (defining stable dosing) may have an effect on driving, so that needs to be documented.  However, research should not be limited to those on stable doses, because it is the group that is not on stable dosing who is at highest risk.  Also, those presumed to be on stable doses, but who are not compliant (and therefore not stable at all) are more likely to be at high risk.
  • Although the potential effects of over-the-counter herbal remedies are increasingly being studied in other countries, and some of the mechanisms are beginning to be understood, the evidence on outcomes (adverse events and drug-drug interactions) isn’t strong enough to justify including these substances in research on polypharmacy and driving.
  • Over-the-counter (OTC) medications other than herbal remedies should be included in future research, because increasingly, the Food and Drug Administration (FDA) has been moving medications from prescription to OTC, particularly the nonsteroidal anti-inflammatory drugs (NSAIDs) and antihistamines, and there is ample science on which to recommend inclusion of these medications.
  • Setting aside the aforementioned NHTSA comments on this subject, there was not any clear-cut agreement among the panelists regarding whether alcohol should be included as an agent that causes interactions with medications.  In the practices of the CDRSs and physicians, older people are in the drinking-and-driving population and are using medications; they just are not picked up by law enforcement teams that generally operate at night (when older people are less likely to be on the road); and when they are picked up, it is often not reported. Alcohol-use screening of all admissions in an acute-care hospital in one physician’s practice showed that 20 percent used alcohol regularly, and the majority of the users were over age 60.  Alcohol used at night may still be present and interact with medications used during the day.
  • To enhance the cost-effectiveness of research on polypharmacy and driving, studies should be limited to drugs/classes of medications that have high exposure in the population of interest, in contrast to those that may be more impairing but have a much lower frequency of use.
  • To identify target drugs/drug classes of interest and their effects on driving, or to gain information that could alter practices for prescribing, it makes more sense to do a cross-sectional study than a longitudinal or an epidemiological study.  In the latter case, the magnitude of a study needed to ensure enough people who were on a given medication or class of medications, and who had an outcome of interest, would be excessive.
  • Concern was voiced among panelists that there are an overwhelming number of variables impinging upon the topic of polypharmacy and driving.  Between the medical conditions, and the physical conditions, studies may demonstrate significant associations but are unlikely to demonstrate a causal effect between medication use and crashes.  NHTSA participants acknowledged this, and reiterated that the hope was to do research to inform programmatic efforts—to inform drivers, physicians who prescribe, pharmacists who advise their customers, licensing agencies and Medical Advisory Boards, and possibly law enforcement, to be able to recognize signs of impairment.  The purpose of future research in this area is not to come up with a basis for restricting driving for individuals who take particular medications. 

Defining “Older Drivers” For Future Research

  • Most of the potential for problems with respect to polypharmacy and traffic safety is found among drivers age 70 and up; it is primarily in this group where multiple medications are being used, for multiple conditions.   In addition, since crashes are the outcomes of interest in database analyses, and crashes are less frequent in the “young-old” age group (e.g., 55 to 70), it makes more sense to focus research upon the older age groups.
  • At the same time, educational efforts should begin by age 55; people often start taking more than one medication around that age, and should be aware of potential interactions as well as effects on driving.  Thus, separate groups of people, e.g., age 65+ and those under age 65 should be included in future studies evaluating the impact of education and awareness campaigns.
  • It would be a significant omission to exclude drivers with cognitive impairment from studies of polypharmacy and driving, despite the fact that it may be difficult to determine whether they perform poorly because of their cognitive impairment or because of the medications they are taking.  There a large percentage of the older population with cognitive impairments, and many of the medications that put people at risk for crashes are cognitively impairing.