WIDAC - Council Meeting
(April 13, 2023) - Impaired Driving Data

The council heard that even after modifications to traffic fatality data involving drug testing results the narrative on the increasing prevalence of poly-drug incidences remained unchanged.

Here are some observations from the Thursday April 13th Washington Impaired Driving Advisory Council (WIDAC) Council Meeting.

My top 3 takeaways:

  • In October 2022, the Washington Traffic Safety Commission (WTSC) was presented with analysis indicating that data on poly-drugged driving incidents overreported cases involving cannabis.
    • Research Associate Max Roberts presented on the subject, showing statistics for “trends in different drug and/or alcohol statuses among drivers in fatal crashes in Washington from 2008 to 2020” which suggested “drivers and fatal crashes that are poly-drug positive ha[d] been increasing rapidly since 2011 reaching a high of 182 drivers in 2020.” As commission staff were “reviewing positive drug results to better understand impairment and to reevaluate the prevalence of poly-drug driving,” Roberts explained they had found the use of different testing methods for “identifying screening versus confirmatory results” meant, in “some circumstances screening and confirmatory results may be present and counted as multiple drugs.”
    • He’d reported that staff had begun reassessing these “drug positive drivers and fatal crashes from 2015 to 2020,” and revising down their estimates for the number of drug combinations and incidents. Previously, “44% of drug positive drivers and fatal crashes tested positive for two drugs, and only 23% tested positive for only one drug.” However, Roberts said “this switched and the majority, nearly 67%, of drug positive drivers and fatal crashes tested positive for only one drug, and 22% tested positive for two drugs.”
    • With these updates, Roberts had concluded that the “trends and the message remain generally unchanged: poly-drug positive drivers remain the most prevalent group in fatal crashes, and the number of poly-drug positive drivers has been steadily increasing over time.” He promised WTSC officials would continue to track the results and present further findings, predicting the public would “see some changes to Washington's alcohol and drug positive driving data in the future.” 
    • WTSC organizes WIDAC to hear from agencies and groups working to reduce impaired driving and assist with coordinating their work.
  • WTSC Research and Data Division Research Associate Max Roberts talked about Re-Evaluating The Prevalence of Poly-Drug-Positive Driving in Washington, covering current data, methodological changes, and revised statistics and trends.
    • “I do want to…address one thing about the proportional aspect of the impaired driving…fatalities that we're seeing,” Impaired Driving Program Manager Edica Esqueda remarked. She understood that increasing impaired traffic “fatalities [were] proportional to” an overall increase in deaths on Washington roadways and weren’t “necessarily spiking the numbers.” Impaired driving incidents had “always been in and around 50%” of all traffic fatalities, “and it remains around 50% of the whole, now,” Esqueda asserted (audio - 2m).
    • Roberts started his remarks by emphasizing the "novel approach that we are implementing" to "re-evaluate the prevalence of poly-drug positive driving in Washington." Drivers were considered to be “poly-drug,” he stated, because toxicology testing confirmed they were “positive for two or more drugs that are not alcohol,” or “for alcohol and one or more other drugs.” Roberts showed a graphic for the “number of drivers in fatal crashes positive for alcohol and/or drugs in Washington from 2012 to 2020” which indicated the “number of drivers in fatal crashes positive for one drug only, and alcohol only, has remained relatively stable over time. However, the number of drivers in fatal crashes that are poly-drug positive has been increasing rapidly since 2012, reaching a high of 183 in 2020.” He didn’t have the final statistics for 2021, but relayed that the “number preliminarily is 222” (audio - 6m)
      • “So the work that we are currently doing is reviewing positive drug results to better understand impairment and to reevaluate the prevalence of poly-drug driving,” Roberts stated. He used the example he’d offered in 2022 about a hypothetical “older motorcycle rider waking up on a Sunday morning getting ready to go for a ride” who died and was determined to have taken Tylenol, also known as acetaminophen, drank “a couple cups of coffee,” took non-impairing prescription medicine, and “happens to be a chronic cannabis user, but they did not smoke the morning before.” Such a person would test positive for several drugs, including carboxy-THC [11-Nor-9-carboxy-Δ9-tetrahydrocannabinol], a non-active metabolite from THC, he added. Circumstances like these “would be flagged in our data as a…poly-drug positive driver and…one might come to the conclusion that it was drug impairment that caused the fatal crash,” Roberts noted.
      • “However, none of these substances have impairing qualities that contributed to, or were directly involved in the fatal crash,” he told the group, and when accounting for substances like this, it was reasonable to declare “drugs impairment were not actually involved, nor contributed, to the fatal crash.” Believing that “drug presence does not equal drug impairment,” he remarked that staff were working to “try and tease out some of these non-impairing circumstances to get a better understanding of the prevalence of poly-drug positive driving.”
    • Current Data (audio - 4m)
      • WTSC “manages our enhanced state traffic fatality data set, which is called the coded failed crash files, or CFC files,” which Roberts described “contain[ed] information from police traffic collision reports, death certificates, drivers licenses, and driving records, and toxicology results.” Studying CFC files from 2012 to 2020 with “all drivers that were involved in a fatal crash that had a positive toxicology drug result,” and excluding alcohol-only results, narrowed their review to “1,635 drivers over that nine year period” with positive tests for “130 unique drugs,” as well as “491 distinct drug combinations among these drivers.”
      • “The majority of drug positive drivers involved in a fatal crash tested positive for two drugs,” commented Roberts, constituting 41% of incidents, while those testing positive for a single drug was 24%. The remaining 35% had three or more positive drug test results, he added.
      • “THC and carboxy-THC was the most common drug combination representing about 21% of drug positive drivers and fatal crashes,” Roberts said. Although “THC is the primary active component of cannabis, which causes the user to become high,” carboxy-THC was “non-active…and in chronic cannabis users carboxy-THC can be detected in the blood for days or even weeks after consuming cannabis.” The next most common combination, “amphetamine and methamphetamine,” represented “about eight percent” of incidents and both substances could be actively impairing, though “amphetamine is an active metabolite of methamphetamine.” In total, “32% of drug positive drivers had only forms of cannabis,” Roberts stated.
    • Revision Methodology (audio - 8m)
      • Roberts went over ‘data modification steps,’ or how impairing substances were determined and when they were screened out of the updated count, such as being a non-active metabolite of another drug:
      • Potential for impairment - “We broke these out into three categories: impairing, potentially impairing, and non-impairing.” A substance was determined to be impairing “when there was clear evidence and understanding of the drug's capacity to impair driving abilities,” with “methamphetamine, benzodiazepines, THC, and Zolpidem, also known as Ambien,” offered as examples. Compounds were marked as potentially impairing “when there was mixed evidence on the drug’s capacity to impair driving, but the potential for impairment remained,” mentioning “selective serotonin reuptake inhibitor antidepressants, antihistamines, and pseudoephedrine.” And non-impairing substances which had been listed in drug test results were considered to have “no serious impairing effects associated with the drug” which covered “non-active metabolites such as carboxy-THC.”
      • When to screen out drugs - Using the standards above, 14 drugs were deemed “non-impairing” and removed from consideration of poly-drug fatalities. Some drug metabolites had impacts on the body, but other “non-active” compounds, carboxy-THC among them, lacked “any effect on the body” and were removed from the data “but only when they appeared on their own, and without a parent drug present.” 13 non-impairing pharmaceutical drugs were removed, he said, “but only if alcohol was not present as there could be potential for interactions to occur when these drugs are with alcohol.”
        • Roberts specifically noted that because carboxy-THC alone wouldn’t count as an impaired driving result under this standard, “feedback in other venues where we presented this information” had been suggested to keep carboxy-THC in impairment statistics since “THC can be metabolized quickly in the body, and with the time it takes to get a blood draw after a crash a driver who is positive for THC when they crashed may only show as being positive for carboxy-THC in their toxicology report.” He asked for “professional input” on whether to retain the THC metabolite in their drug impairment count at the conclusion of his remarks.
        • Find out more about cannabis metabolization and detection times, which has been found to involve factors besides how recently a person consumed cannabis, such as frequency of use, hydration, gender, or mode of ingestion.
        • 2023 research indicated use of full spectrum hemp items may result in detectable carboxy-THC, and a 2018 study on testing of oral fluids instead of blood noted “increased true positives and decreased false positives of suspected marijuana-related driving under the influence (DUI) arrests.”
      • Matching metabolites with parent drugs - Roberts added that “we match parent drugs with our metabolites” in order to “determine situations where actually only one drug was consumed.” The data was also changed to distinguish “screening versus confirmatory toxicology results,” the latter “used to determine if a specific drug is present and what level of that drug is in the blood.” He told attendees, “in some circumstances a screening and a confirmatory result was present at the same time” and “each was being counted as a separate drug.” Roberts summarized the steps of how they had analyzed and modified poly-drug fatality data.
    • Revised Data (audio - 3m)
      • Turning to changes to the drug positive and poly-drug statistics from 2012 to 2020, Roberts described how “after applying our methodology, 240 drivers were removed representing a 15% decrease in drug positive drivers,” including those “drivers that were positive for only carboxy-THC or other non-active metabolites.”
      • “The number of unique drugs in the data decreased from 130…to 114,” explained Roberts, while the “number of distinct drug combinations decreased from 491 to 360.” Although “in our original data the majority, or 41%, of drug positive drivers in fatal crashes tested positive for two drugs, and 24% tested positive for only one drug,” he said when “applying our methodology, the majority, about 69% of drug positive drivers in fatal crashes tested positive for only one drug” and the “proportion of drivers testing positive for one drug nearly tripled from the original data” along with “about half as many that tested positive for two drugs.” Additionally, Roberts commented “31% were positive for two or more drugs, which was a notable reduction from the 76% that were positive for two or more in the original data.”
      • “So in the original data the combination of THC and carboxy THC was the most frequent drug combination appearing among 21% of drivers, but after modifying the drug data” for the parent-metabolite combination Roberts remarked how “THC carboxy/THC was still the most frequent drug combination” increasing to “about 32% of drug positive drivers.”
      • “And then finally about 34% of drug positive drivers were positive for only forms of cannabis, which was up slightly from the 32% in the original data,” Roberts concluded.
    • Alcohol (audio - 2m)
      • Adding alcohol back into the impaired driving data, Roberts said there had been “713 drivers that were positive for alcohol only” over that nine year period for “any [blood alcohol content] result greater than zero.” He stated this “brings our total to 2,108 drivers in fatal crashes that were positive for drugs and/or alcohol.” Alcohol by itself was the most frequently occurring drug, Roberts acknowledged, “representing about 34%” of fatal traffic incidents, moreover “56% of alcohol positive drivers were positive for only alcohol” while “about one in five alcohol positive drivers that were also positive for THC/carboxy-THC.”
      • Alcohol and THC/carboxy-THC was the second most common drug combination, accounting for 12%” of the drivers involved in fatal crashes, Roberts reported, “and then THC/carboxy-THC was the third most common drug combination representing nine percent of drivers, followed by methamphetamine/amphetamine, which was the fourth most common at five percent.”
    • Conclusions (audio - 5m)
      • Sharing “updated trends” based on staff revisions, Roberts remarked “we can see a considerable reduction across all nine years in the number of poly-drug positive drivers involved in fatal crashes, but it still remains the most prevalent category.” A corresponding increase in positive results for drivers having only one substance in their systems was also observed, he noted, sharing some specific changes in the most recently available data for 2020 where there’d been a “35% decrease among poly-drug, a 143% increase in one drug, and a 15% increase in alcohol only” incidents.
      • Roberts’ interpretation was that the “trends and the message remain generally unchanged after modifying the data, and that is that poly-drug positive drivers remain the most prevalent drug and alcohol positive group in fatal crashes and the number of poly-drug positive drivers has been increasing steadily over time.” He added that a “new discovery was the prevalence of one drug only drivers who became the second most prevalent group in fatal crashes surpassing alcohol only positive drivers in every year after 2013.”
      • WTSC staff intended to use this methodology going forward on traffic fatality data for Washington State, said Roberts; “this will mean updating our state data files, recent data briefs on drug and alcohol positive driving, as well as our dashboards.” Moreover, “we are anticipating releasing a larger report later this year that will cover everything here and much more,” he added. Even with this revised approach, he stressed “the overall trends and message about poly-drug driving remain the same.”
      • Robserts identified three limitations on the analysis:
        • Length of time from crash to blood draw
          • Roberts understood that “if a driver is under the influence of THC while driving and then crashes, but doesn't get a blood draw for several hours, we would not see THC in the toxicology report, and instead we would only see carboxy-THC.” The metabolite was “challenging because it may be our best indicator to recent cannabis use while driving,” but could “also show in chronic users, even if they have not recently consumed cannabis. So in those circumstances, we may presume they were under the influence of THC when they crashed when in fact they were not, so there's a dilemma there.”
        • Data linkage with [emergency medical services] and hospitals
          • This aided in determining “when certain drugs have been administered by emergency staff…this could involve drugs such as ketamine, fentanyl and morphine” and WTSC researchers lacked “comprehensive data linkages” making them “unable to determine every circumstance where some of the drugs are administered by emergency staff.”
        • Analyzing change in drug data among non-motorists
          • “About half of our fatally injured non-motorists are flagged as being impaired” meaning that “applying the data modifications I have shown here might lead to a decrease in that number.”
  • Several questions were posed around drug concentrations, data sources, and approaches for disseminating documents on poly-drug impairment data to lawmakers, press, and the public.
    • Asked if researchers were further scrutinizing drug concentrations in blood aside from alcohol, Roberts answered that “at the moment we are not, we do have that available but when we ultimately update our data files, we will only be including levels for alcohol, THC, and potentially carboxy-THC just because there is so much debate and there's not really a level that we can declare” where it definitely “becomes impairing” (audio - 3m).
      • WTSC Research Director Staci Hoff said there was an unreleased research paper “that talks about therapeutic levels of drugs and what I envision is a simple flag that” mentions when drug positive results were “over the therapeutic dose.”
      • Someone claimed that for THC testing, "we're often seeing levels decrease to below per se" and also believed “cocaine has a very short half-life, an hour. So, if you're considering adding carboxy back in I would consider adding” metabolites for cocaine in the drug positive results.
      • Amy Freedheim, Senior Deputy Prosecuting Attorney in the King County Prosecuting Attorney’s Office, argued that in impaired driving prosecutions, her staff looked for “really strong evidence…that carboxy-THC is based on THC at the time of the collision or driving.” She agreed that “just to be on the safe side, and because the recreational marijuana community…fights us so much about impairment - it's probably better to under report.”
    • When the issue of Federal Motor Carrier Safety Administration data on drug testing of commercial truck drivers involved in fatal collisions was raised, Hoff stated that information wasn’t included but they hoped to gain access to the Drug Recognition Expert (DRE) data system, as federal tests were outsourced to NMS Labs, separate from the Washington State Patrol (WSP) laboratory facilities. She further recognized state officials had more “flexibility” in presenting statistics “to make our data meaningful for us,” and data from other agencies could be limited. Roberts noted federal authorities only shared data identifying test results as "prohibited or not prohibited," something that complicated information around cannabis, which remained a schedule 1 controlled substance. He described how federal officials were “in possession of a lot of really good data about impairment that they're not sharing with the states” (audio - 3m).
    • Hoff noted the difficulty differentiating “what we've been saying versus what we're gonna say, so language matters and we are being very careful about that.” The report mentioned by Roberts would include more methodology details, she commented, as well as making the raw data itself available. Revising drug impairment data was the “right thing to do, but it's hard to change a trend…to back step what we've been communicating for 15 years about this” (audio - 1m).
    • Daniel Cooke, Licensing, Endorsements, and Traffic Safety Administrator at the Washington State Department of Licensing, asked whether past CFC files would be revised to reflect the new approach, which Hoff affirmed would happen (audio - <1m).
    • Linda Thompson, Washington Association for Substance Misuse and Violence Prevention (WASAVP) Board Member and Greater Spokane Substance Abuse Council (GSSAC) Executive Director, wondered about getting a more concise briefing document. Hoff promised a “one-pager, [and] long pager” publications about the data would be released. Thompson explained such materials would be useful when she spoke with lawmakers about impaired driving, and then complained about passage ofSB 5123, legislation prohibiting pre-employment drug screening for cannabis in many job fields. “That's an issue for…prevention about can you really work on a job” if a test shows one has previously used cannabis. She lamented a perceived “trend is it's not a big deal to use pot, or drink really, or use something else and I think there's so many ways we could educate” the public against an attitude that “it's not that big a deal” (audio - 3m).
      • As passed by the legislature, SB 5123 wouldn’t modify any workplace drug testing rights or mandates on employers related to screening for on-the-job impairment, and at time of publication, the measure was awaiting action by Governor Jay Inslee.
    • WTSC Executive Director Shelly Baldwin pointed out that when she’d been asked “what drugs do you see [reported in crashes], and I'd show them the whole list and they see…aspirin on it and they'd be like ‘well, how trustworthy is this list?’” She complimented Roberts and other researchers on changes she believed would help the “credibility” of their statements on the topic (audio - 1m).

Information Set