Harm reduction and drug-impaired driving: sharing the road?

As research increasingly demonstrates that drug-impaired driving (DID) poses significant risks to public health and safety (e.g. Asbridge, Hayden, & Cartwright, 2012; Elvik, 2013; European Monitoring Centre for Drugs and Drug Addiction, 2012), international jurisdictions are paying greater attention to developing and refining legal and other countermeasures to reduce DID (Watson & Mann, 2016). Legislative responses vary greatly across jurisdictions, but in general can be categorised as zero tolerance, per se with defined cut-offs for certain drugs (e.g. 5.0 μg/l for tetrahydrocannabinol in blood), behavioural impairment, or some combination of these approaches; related measures to support such laws include roadside drug testing by police and remedial programmes for convicted impaired drivers (Watson & Mann, 2016). We use ‘DID’ to refer to driving a motor vehicle while experiencing the acute effects of psychoactive drug use – including the use of varied licit and illicit drugs for recreational and/or therapeutic purposes – that may impact the psychomotor and cognitive skills involved in driving and potentially increase collision risk. Curiously, we have observed that the term ‘harm reduction’ is seldom used in published research and policy discussions about DID, nor has the topic of DID received much coverage at recent international conferences focussed on harm reduction. Our observations prompt us to ask: could this omission reflect avoidance of tensions that emerge when the traditional principles of harm reduction (developed for and applied to drug policy more broadly) are considered in relation to responses to DID? What would a more explicitly harm reduction-infused research agenda on DID look like and how might it inform responses to DID? In light of the growing attention to DID, the many prescription drugs with psychoactive effects, and changes to drug policy that may increase drug availability and access as well as, quite possibly, cases of DID (e.g. legalisation of medical and/or recreational cannabis; Hall & Lynskey, 2016), now is an important time for cross-disciplinary dialogues that encourage researchers to consider such questions and novel, and perhaps some controversial, DID research and initiatives. Our goal with this editorial is to start new conversations.

Harm reduction as a public health model for drug policy and interventions largely took shape over the 1980s and 1990s, primarily in response to HIV epidemics among people who inject drugs (Ball, 2007; Stimson & O’Hare, 2010). Over the years and with much debate, considerable agreement has been reached on the main principles of the model. These principles include, among others, targeting specific risks and harms, having evidence-based and cost-effective policies, treating people who use drugs with dignity and compassion, and contesting policies (including those that criminalise drug use) that increase harm (Harm Reduction International, 2010). While debate continues regarding whether the harm reduction label should apply to any intervention designed to reduce drug-related harm inclusive of interventions that aim to reduce drug use, or if the term should apply only to interventions designed to reduce drug-related harm without the aim of reducing drug use (Ball, 2007), a defining feature of harm reduction is demonstrating some degree of non-judgemental acceptance that individuals may continue to use drugs (i.e. abstinence is not necessarily the ultimate goal), albeit they may be encouraged to do so in ways that reduce the numerous health and social harms associated with drug consumption (Erickson, Riley, Cheung, & O’Hare, 1997). This logic readily applies to responses to DID that try to separate impairing levels of drug use from the behaviour of driving. Indeed, a longstanding connection already exists regarding alcohol and driving. Investigations of different interventions (including population-level policies and campaigns, educational programmes, and interventions aimed at drinking environments) designed to reduce various alcohol-related harms recognise the importance of reducing drink-driving rates in particular (e.g. Anderson, Chisholm, & Fuhr, 2009; Jones, Hughes, Atkinson, & Bellis, 2011; Lang, Stockwell, Rydon, & Beel, 1998; Martineau, Tyner, Lorenc, Petticrew, & Lock, 2013; Room, Graham, Rehm, Jernigan, & Monteiro, 2003). Drink-driving laws are viewed by many as ‘a classic example of harm reduction’ (Stimson & O’Hare, 2010, p. 92). For example, per se laws that set specified blood alcohol concentration limits for drivers, and random breath testing programmes that allow police to stop drivers to detect those driving above per se levels, have been found to effectively reduce rates of drink-driving (Fell & Voas, 2014; Homel, 1990; Mann et al., 2001). In short, the application of a harm reduction approach to drink-driving seems to have meant that there has been a specific focus on and extensive research to address this behaviour, without trying to prohibit or reduce drinking itself, as well as identification of programmes and policies that successfully reduce alcohol-related collisions, injuries and fatalities. Researchers and policymakers are certainly interested in achieving similar road safety improvements via establishing, for example, safer drug limits for driving (e.g. Wolff, 2016). However, it is presently unclear whether application of a harm reduction lens to DID would develop in the same way as it has for drink-driving given the limited state of knowledge about DID and possible policy responses that may be seen as compromising some core principles of the approach, as we discuss below.

What do we know about the underlying beliefs and behaviours of DID? Compared to drink-driving, there is much that is currently unknown about DID that might be useful for informing efforts to reduce harm. To date, most relevant studies about DID-related perceptions, beliefs and behaviours have tended to engage younger adults and focus on illicit drugs, and have under-recognised groups of people who drive after taking prescribed or therapeutic drugs (for a discussion about the latter, e.g. Voas, DuPont, Shea, & Talpins, 2013). Commonly reported perceptions include DID is not as dangerous (i.e. likely to result in traffic collisions) as drink-driving and that risk of police apprehension for DID is relatively low or lower than the chances of getting caught for drink-driving (e.g. Danton, Misselke, Bacon, & Done, 2003; Darke, Kelly, & Ross, 2004; Degenhardt, Dillon, Duff, & Ross, 2006; Matthews, Bruno, Dietze, Butler, & Burns, 2014; Matthews et al., 2009; Terry & Wright, 2005). Studies have also documented relatively high degrees of willingness or intent to engage in DID, sometimes along with a lesser intent to drink and drive (Fischer, Rodopoulos, Rehm, & Ivsins, 2006; Furr-Holden, Voas, Kelley-Baker, & Miller, 2006; Swift, Jones, & Donnelly, 2010). Given the growing number of jurisdictions worldwide that are debating or have implemented cannabis legalisation (e.g. Pardo, 2014), people who use cannabis have been a key population to study, and it appears that many users do not view their driving as impaired after cannabis use, although they may perceive a heightened risk of collision when cannabis and alcohol use are combined (e.g. Fischer et al., 2006; Swift et al., 2010; Terry & Wright, 2005). Other notable variation in DID beliefs has been observed within and between sample populations, depending on the type(s) of drug being asked about. For example, among a sample of 273 nightclub attendees in Australia, 71% perceived that driving under the effects of heroin was ‘very dangerous’ compared to 59% who perceived the same regarding alcohol (Degenhardt et al., 2006). In contrast, in a qualitative study from Scotland involving 26 people who reported having been addicted to heroin and had engaged in DID, driving under the influence of heroin when the drug is taken to ‘feel stable’ (as opposed to suffering from withdrawal symptoms) was perceived to be less dangerous than other types of DID; hallucinogenic drug use and polydrug use were viewed by these participants as especially risky when coupled with driving (McIntosh, O’Brien, & McKeganey, 2008). While such self-report studies suggest that there are potentially multiple DID subgroups that may exhibit different risk profiles of interest, epidemiological investigations appear to suggest the same. Notably, a growing evidence base demonstrates that alcohol and different drugs are associated with different increases in collision risk (Elvik, 2013; Romano, Torres-Saavedra, Voas, & Lacey, 2014).

Important from a harm reduction perspective, study participants report various strategies used to reduce DID or compensate for drug-related impairment. Examples of such strategies include nominating designated drivers, taking public transportation, limiting drug consumption, limiting or not engaging in polysubstance use, not driving after using certain drugs (e.g. hallucinogens), driving slowly, stopping driving if feeling impaired, and trying to ‘sober up’ by delaying driving or ingesting food or water (e.g. Duff & Rowland, 2006; McIntosh et al., 2008; Swift et al., 2010); sometimes behavioural strategies that involve driving while still potentially under the influence of drugs are opted for, despite accompanying perceived safety risks, simply because of the convenience and cost-effectiveness of driving to certain destinations when there is a lack of other options (e.g. readily available public transit). In some contexts, DID, therefore, appears to be a fairly ‘normalised’ behaviour among people who use drugs, particularly if it is common among one’s peer group (Duff & Rowland, 2006). It would be worthwhile to investigate whether any personal and routine ‘safer’ DID practises may effectively reduce road safety risks – although recommending some of the abovementioned practises, even where informed by evidence, may be controversial.

As with our understanding of underlying perceptions and behaviours, our understanding of approaches that effectively prevent DID is much less developed for drugs other than alcohol. Areas that need significant clarification include technical and legal matters related to detection and enforcement. For instance, while some evidence suggests that random roadside drug testing programmes hold promise in terms of increasing DID detection and associated perceptions about getting caught that may, in turn, reduce DID, few programmes have been rigorously evaluated and consensus on reliable, scientifically-derived legal limits for different drugs has been elusive (e.g. Watson & Mann, 2016; Wolff, 2016). It is imperative that more research is conducted that aids our understanding of reliable drug testing and detection techniques, and drug thresholds that relate to impairment, and how this knowledge might improve responses to DID. Relatedly, in some studies, people who use drugs have expressed the idea that increasing the certainty of detection is likely a more effective DID deterrent strategy than broad educational or public awareness campaigns (Degenhardt et al., 2006; Swift et al., 2010; Terry & Wright, 2005), but more research is needed here as well.

How to merge deterrence-based policy options with harm reduction principles remains an open-ended discussion, but one worthy of engagement. In particular, zero-tolerance DID laws that appear to exhibit ‘prohibitionist logic’ raise valid concerns about unfairly criminalising and targeting people who use drugs (e.g. Roth, 2015). Settings where DID laws draw distinctions between illicit and prescription drugs and make special exceptions for the latter (e.g. the United Kingdom; see http://www.gov.uk/drug-driving-law), may raise concerns about equity. Further, the finding that DID appears particularly entrenched among people who use drugs frequently or exhibit problematic use has led some authors to propose that DID-prevention interventions target levels of drug use (Matthews et al., 2009; McIntosh et al., 2008; Swift et al., 2010), a suggestion that might also have to be reconciled with the harm reduction tenet that accepts an individual’s choice to continue using drugs. That said, applying principles from harm reduction may also carry the benefit of increasing the perceived legitimacy of laws and other efforts to address DID-related harms, and thus may increase acceptance of and compliance with those efforts (e.g. Owens & Boorman, 2011).

In sum, as DID is increasingly recognised as a major public health concern, we wish to draw attention to neglected opportunities to understand and address DID utilising the harm reduction paradigm. Although there is much to learn, an approach to DID that does not proscribe drug use may advance practises and policies that will improve road safety. A major potential benefit of such an approach is that avoiding prohibitionist logic may increase engagement in and acceptance of DID prevention measures among people who use drugs. We recommend active collaboration between road safety researchers and harm reduction researchers, service providers and people with lived experience in DID research and policymaking as such efforts will contribute unique insights important for the design of legal, educational and other interventions that could effectively prevent DID and associated harms. New research should span different social, cultural and legal contexts to examine potentially localised patterns of DID while recognising that there are many types of psychoactive substance use (licit and illicit, recreational and therapeutic) that can impair driving.

Declaration of interest

No potential conflict of interest was reported by the authors.

Acknowledgements

Dr. Tara Marie Watson is funded by a Canadian Institutes of Health Research Postdoctoral Fellowship.