Hi. It's 1245, so I'm going to start my presentation. My name is Dr. Michelle Waters. I'm a medical officer with the National Transportation Safety Board. Am I too loud? Okay. Is that better? Thank you. The title of my presentation is Beyond DOT Drug Testing, Toxicological Evaluation in NTSB Accident Investigations. So the learning objectives, for one, I'd like to present to you how at NTSB we look at the medical aspects of one of our transportation accidents and want to show you why we do go beyond DOT in terms of where DOT leaves off and what we need in terms of finding more sufficient information in terms of our toxicological findings. And finally, I'd like you as physicians and healthcare practitioners to recognize that it's beyond even illicit drugs in terms of our concern for safety-sensitive functions in accident causation. So the NTSB is an independent federal agency. So we were created from the Air Commerce Act in 1926, where Congress charged the Department of Commerce to investigate the new and upcoming aviation field for any civil aviation accident. When the Civil Aeronautics Board's Bureau of Aviation Safety was created in 1940, they resumed that responsibility. In 1967, however, Congress felt that they needed to look beyond aviation and look into other transportation modes. Thus, the NTSB was created by statute, and initially we were housed under the Department of Transportation. Makes sense, but over time they recognized that it's a little difficult for our agency to provide recommendations to our parent or to our sister agencies, such as FAA or FRA, when we were housed under the DOT. So in 1974, we became an independent agency. We report directly to Congress, and we get our funding directly through Congress. As mentioned, we were created by statute, and so there's both statutory and regulatory rules that guide our mission, which is essentially to promote safety in transportation, as well as to conduct accident investigations to find lessons learned from those transportation accidents and send them forward in the idea of trying to decrease the number of accidents. So as stated, we are charged with doing all the civil aviation accidents in modes such as rail, pipeline, and marine. There's more specific why we might get involved, has more to do with how many fatalities there are, how serious industries, or how substantial the property damage is. For highway accidents, a little different. It's more selective. We base it more on, are we interested in school bus safety? So we might investigate more school bus accidents to make recommendations, and we do those accident investigations in coordination with the states. So the objective of our investigation is we try to find the probable cause. So we gather facts, and we look at the operational evidence to pull together what we feel is the cause of the accident. In doing so, we identify safety issues. Some of them may or may not be part of the probable cause, as you'll see from our examples, and then we make recommendations aimed at preventing future accidents. I'll also cover safety studies, which the agency also does perform in transportation research. So for the medical part, one, and of course stepping back, you know, yes, you have the engineering and mechanical issues, we have the operational issues, and we have human performance that we look at for each accident. From the human performance side for the medical, we specifically are looking at, did the medical condition, its treatment, or its use, lead to impairment or incapacitation, which contributed to the cause of the accident. Now, I'm sure many of you in your daily lives deal with both impairment and disability. For us, impairment looks at those, a loss or decrement in terms of the physical, physiological, psychological, and anatomy, structural, that may impair performance. And for us, that also does include things that probably wouldn't enter your mind for disability, and those include judgment calls, attention, alertness, things like that. So for us, the impairment is a spectrum. So at one end, we have peak performance. At the other end, we have, extreme example, we have incapacitation. And everything else, we sort of evaluate in terms of its impairment. Now, that doesn't necessarily mean that you have to go along from impairment into incapacitation, though you would in something like a carbon monoxide poisoning, where the more exposure you have, the more impaired you become. But it could involve some sudden incapacitation that occurs in flight or in operation of whatever vehicle. For this presentation, I'm going to focus on impairment due to drugs and not including alcohol. So to address these things, we have three questions that we look at for evaluating medical impairment. Did increased risk for medical impairment exist? Did some kind of medical impairment occur? And finally, did the medical impairment contribute to the accident? Remember, we're not looking at blaming anyone. We're looking at trying to find the underlying cause of the accident. And for any of these questions, when we encounter them, we might come up with no, or we don't have enough information to actually make the determination. So let's start with our first question. Did the increased risk of medical impairment exist? And for that, we look at a lot of records, so like the medical evidence, and we want to find a potentially impairing condition, substance, or exposure. So this might not be at a level, but we want to know, is there something that says it was there? So for that, since we're focusing on toxicology, I look at those sources. Stepping back, all those statutes and regulations, we have subpoena authority to gather whatever evidence we need for the investigation, which includes medical records, as well as specimens for testing. We also have, under HIPAA, we are a public health authority. So for those of you, as far as you will not need to obtain any kind of authorization from the individual to provide medical records to us. So a big part in the next few slides, we're going to talk about toxicology testing. But as said, we have a lot of sources that we need to look for. So we will look for your medical certification files, if it's a FMCSA case or FAA and others. We will try to gather employee medical files, occupational medical files, if it's person, if the operator's affiliated with a large enough company that has a medical department. We'll get post-accident, pre-accident pharmacy records, as well as we do interview a surviving operator or their spouses to find out information related to any kind of medication use. So toxicology testing. Now again, I'm sure most of you in the audience are familiar with DOT testing, particularly pre-employment, suspicion. If time permits, I'll discuss some involvement NTSB has with random testing. But for the most part, this talk will talk about post-accident DOT testing. Now, while this talk was saying beyond, it's not that, you have to recognize that DOT testing is meant to, the objective was to prevent illicit drug use in the workplace. That testing was never meant to find the cause of the accident. Thus, we find sometimes it's not sufficient for us to go on, and we have to look for either additional testing or, as previously looked at, that other sources of information. So here's where we get some of our other toxicology results from. Okay, you notice for post-accident, FRA testing is different than DOT. You also have hospital records that we may have. That's usually a clinical screen for drugs of abuse. Law enforcement, often drugs of abuse, but sometimes a more extensive panel than that. Medical legal jurisdictions differ in terms of what kind of post-mortem testing they perform, but we will obtain that. And finally, I want to mention our FAA relationship with the Civil Aerospace Medical Institute, CAMI, and its forensic sciences laboratory. So we have, they, of course, perform toxicology testing on deceased pilots. However, we also have a memorandum of agreement with them to perform testing on surviving pilots, surviving operators, and other deceased operators in the other modes of transportation. And the picture that you see on the right is one of the tax boxes that we would provide to the coroner's office, medical examiner's office, for the collection of blood, urine, and tissue specimens. So I created this chart to sort of compare some of the different agency rules in terms of post-accident toxicology testing for surviving operators. So you notice we have DOT, and of course DOT, as you know, what you, how you define an accident, how you define who would be involved, and what time you might want to have for, to make sure that you get the testing by a certain time, is really very agency dependent. But overall, these are the general principles that DOT and the U.S. Coast Guard also use in terms of those post-accident testing. So you have a urine screen, and if something comes up, you get a confirmatory testing. They look at five major categories, which amounts to about 14 or 15 drugs. Then we have FRA. As stated, because their regulations for post-accident were created prior to the DOTs, they have kept their post-accident toxicology testing, which is a little bit more expansive. It includes more categories, and therefore about 50 substances that they test for. And FRA does a urine screen followed by a confirmatory blood testing for post-accident. The final column is, has to do with our memorandum of agreement we have with CAMI. So for the surviving operators, we would collect urine and blood. I shouldn't say we would collect. We obtain specimens for urine and blood. From either, we subpoena a DOT specimen A, remember this is a leftover, not a split specimen. We would get, perhaps from the hospital admission, blood and urine. If law enforcement had taken some samples, we have. Our one exception for the DOT under that is FRA does not allow us to subpoena their leftover specimens for their surviving operators. We do for their deceased operators. If the person is uninjured, so doesn't go to the hospital, police haven't asked them for a blood test or urine test, it would be voluntary that the person provided that to us. Because our agency cannot request any additional samples be taken for us. Like, hey, we want to make sure we have enough. We can't do that. So we rely on whatever's leftover that the entity is willing to share with us and send to our FAA laboratory. So as expected, we do look, is this the initial screen, which is somewhat fine, sometimes fine and reasonable. Was there confirmation testing? If there was quantification, different labs vary by what their detection limits are and reporting limits are. And of course, we always have those caveats like, hey, is assessment just not unsuitable versus really the drug is not found? Obviously a difference. Similarly, if you noticed on the previous slide, I had the MRO negative can call something negative even if it is present in the sample. And for us, we don't pay attention to that negative. We do look at whatever level is there. Remember, this is a step where we're looking at, was there a potentially impairing substance available? And overall, again, it's more that we don't really consider just because it's present doesn't mean the person was impaired. And also, particularly more for our civilian accidents like with highways, you sort of have to look was it worse that the person was taking medication or the medical condition that they had? So this accident, and we'll start our examples on toxicology, our first part of does impairment, was there a risk of impairment? So this accident happened about nine o'clock at night. So a truck going northbound on the interstate crossed over and struck a southbound bus. The bus was carrying 15 passengers. They were women from a college softball team on their way home from a tournament. The driver was the coach. And four of the passengers died as well as the remaining people either had minor or serious injuries. So in a written statement after the accident, the truck driver said, hey, I was reaching for a soda. And that's how I came to cross all this traffic and hit the bus or really there was no traffic. So he left the roadway here and started across this hundred foot wide median. So I'm going to show you a short animation. And if you'd like, pretend like you're reaching for that soda. Okay, so he sort of leaves the left hand lane when he's heading northbound, sort of a little curvy, crosses over on the shoulder, starts across the hundred foot wide median, bouncing along because it was a depression to get into the median, sort of goes up of the little gun berm and smashes into the truck causing it to fall on its side and it landed up in the wooded area. So now this is the upright bus, but you can see it sort of had it's totally sheared from that and the passengers who died had been thrown from the bus. If you look at the photo on the right, you can see that while we talked about a hundred foot median, this man is going 72 miles an hour. He crosses this median at an angle of nine degrees. So it's actually, he traverses 1,100 feet as he's bumping along this thing and never comes, becomes aware that this accident is occurring. So let's look at our driver. He does have a current commercial driver's license and about nine months prior to the accident, he had an exam and he was valid for two years. And he reported he had no medical conditions and no medications. So we obtained by subpoena medical records and we found some discrepancies. And well, you know, he had high cholesterol, he had gastritis, but he also had ADHD, and recurrent major depressive disorder. He was taking both Trazodone and Sertraline for major depressive disorder. Trazodone can cause dizziness, drowsiness, and vision changes. And both medications have warnings like, hey, you got to get used to this drug before, like before you operate heavy machinery, see if you have any side effects. But I told you we do take into account how long people are using things. And this man was using it for 10 years and the most recent report was he was stable, his condition was stable, and he had no side effects from these medications. All right, and so the Oklahoma Highway Patrol did, you know, examine the cab after the accident and they did find a bag that contained five prescription medications, including the Trazodone and the Sertraline, as well as seven non-prescription medications. They also found this pipe, like, where's Waldo? They found this pipe and when they went to examine it, like their laboratory that sent it for analysis, it contained 5-fluoro-AMB, which is a synthetic cannabinoid. So, as you may know, these synthetic cannabinoids act on the same receptors as tetrahydrocannabinol. And, you know, in addition to causing euphoria, they can cause seizures, they can cause non-responsiveness. Looking at the medical records, we did find that there was a phone call about a year prior to the accident where the wife called up the primary care doctor and said, hey, I'm really worried about my husband. He's taking K2, and he's been having seizure-like activities. Now, there was no follow-up that we found to that telephone conversation or that message. The truck driver had been seeing a professional counselor between August and November of the previous year. And the counselor had noted that he admitted to using K2 and these synthetic drugs, including while he was at work. Three days after the accident, a former supervisor from his previous company contacted our investigators and said, hey, you know what? This guy worked for us. And during that time, I noticed he had decreased performance. He was missing work. He wasn't forgetful in terms of what he was supposed to do. And so I talked to him, and he told me, the driver told him, he was using K2, but not to worry, because K2 is legal, and it would not turn up on a DOT test. So we did some, the Oklahoma did some toxicology testing that was collected close to three hours post-accident. And you could see, we had the leftover samples and the samples from the Oklahoma State Bureau of Investigation, turned up the sertraline and the trazodone. FAA also found some catorolac. The Bureau also did test for, they had a panel of 15 synthetic cannabinoids, which was negative. FAA, however, does not have a panel for synthetic cannabinoids. So once they detected, found the pipe, and those laboratory results came back, the state, again, looked at the toxicology, and hey, did we find this in the blood? And that was negative for that 5-fluoro-AMB. We did send some leftover specimen, NTSB sent some leftover specimen to the University of California, San Francisco, which had both a university research lab, because they were interested in the synthetic cannabinoids. While they found a peak that might have been the 5-fluoro-AMB, unfortunately, because of the signal-to-noise ratio and some other issues required for them to confirm it, they could not confirm. So, here we have an accident where we have a guy who admitted, hey, I was reaching for that soda. This is what happened. Now, we found the evidence was not consistent. So as I said, we're probably going into some of our other second and third steps in terms of trying to tell you the probable cause. They were not consistent with what he said. And we also know that he has a history of K2 use, admitted K2 use, or synthetic cannabinoid use. We have this pipe that was found that had the substance in it. We really cannot explain why he was unresponsive for that 30-second period. And we really had nothing else to explain it. So our cause ended up being incapacitation stemming from his use of the synthetic cannabinoid. This is also example of why we have some challenges to our studies, including a lot of the novel drugs that places don't test for, as well as the other factors we don't really know enough about them. You know, the half-lives of the synthetic cannabinoids, really I run across the whole gamut. So was it, did we test, was it three hours too long for us to have tested? So there's issues related to all those aspects and sort of some of these substances. So this sort of segues, and as I said, we were repeating, in order to give you the end line, I will sort of tell you some of these other things. Obviously in this one we saw that medical impairment had occurred. And as you also would see, we had to go well beyond just a record because had we just gone on the 5-fluoro-AMB, we'd be like, nothing there, right? So we do look at all the operational evidence, and that could include maintenance records, weather, training records for the person, as well as, of course, the engineering aspects of it. And overall, when we're looking at sort of the skill and experience and how much training and how good someone is at it, we do also reflect on, again, our skill and experience standard. So on your left you have a performance error, and on our right we have an experienced operator, Casey Jones. So if something would happen, like this train falling out of this building, you would say, hey, you know what? I have a higher level of suspicion that there was some kind of impairment because I know this guy is really good. However, if you had a novice operator, that might be what you expect from your peak performance. So we do take that into account when we're trying to discern where impairment may lie. So that brings us to the next case. This happened in Schoharie, New York, at about two o'clock in the afternoon. So 17 people had chartered a limousine. They were heading over to Cooperstown to celebrate one of the passengers' birthdays. And the limousine, you know, this is like it's going to go to Cooperstown. You're going through nice rolling terrain, different grades, different curves, very scenic. So the driver made a right turn onto New York 30, which descended for about two miles where you hit a T-intersection with New York 30A. And that T-intersection was controlled by a stop sign. And the posted speed limit was there was 55 miles an hour. Well, during this descent, the limo started picking up speed. And the brakes completely failed on the limousine. And when they got to the T-intersection, there was a car waiting there. The driver swerved around the car to avoid those two people sitting in the car and drove past that through the New York 30A intersection into a parking lot. It struck a SUV, which propelled, the SUV was propelled into two pedestrians who were standing nearby, killing them as well. So in all, you can see there's very little survivable space in this limousine. We had all 17 passengers were killed, the driver was killed, and those two pedestrians. So let's look at our driver. So he does have a current commercial driver's license. However, it does not have a passenger endorsement. And as a matter of fact, a month prior to this accident, he had been on a roadside inspection, had been stopped and cited for not having that and was told he could not drive anymore. However, he was still driving. And his wife reported that he had actually been driving for this limousine company for about two years on an on-call basis, often on weekends. And prior to that, he had had many years' experience as a truck driver. He also, from the company records, we knew that he had experience, not that this, they had a fleet, but he had been in this vehicle before as well. So he had a current CDL medical exam as well. It was still valid. So that was about a year prior to the accident as well. And on that, he reported that he had a hip replacement as well, but no other medical conditions and medications. So once again, we subpoenaed medical records and we found some discrepancies. So yes, he had a hip replacement. He also had high cholesterol and gastric reflux, but he also had ADHD and bipolar disorder. And the medications that he had prescribed to him included aripiprazole, known as Abilify, which is an antipsychotic used for schizophrenia, bipolar disorder and depression, the antidepressant bupropion, and the antiepileptic oxcarbazepine, which is also off-label uses for mood stabilization and bipolar disorder. We had obtained post-accident testing. Tissue samplers were sent from the medical examiner's office. Unfortunately, no blood was available, which, of course, for most drugs, you really can't equate a tissue level to what was a therapeutic blood level. However, in all his tissues, what he had was delta-9-tetrahydronabinol, as well as the active metabolite 11-hydroxy, as well as inactive metabolite carboxy-THC. So THC, okay, you'll end up having some impairment in potentially motor performance and attention. We also had his bupropion was found, as well as oxcarbazepine. And the fact turned funky when I came here, but it really doesn't matter, because what bottom line is, he is not listed in the probable cause, or his drug use is not listed in that. So what we looked at is, while he had this on it, and we really can't say, was he high when he was driving, what we can say is he did, he had brakes that failed, and there's not many of you going 110 miles an hour, which was the speed he did when he crashed, could actually report how much could you do. And he actually did try to swerve to avoid that. So we recognize that he did actually have enough intentional movements that we did not feel that he was impaired under these circumstances. So he was not part of the probable cause. And as mentioned, we do have challenges with post-mortem specimens. If we do not have blood available, so that difference. We also have issues for post-mortem redistribution, potentially, as well as cavity blood, you might have some mixing from other body fluids. We also try to take into account, is there any pre-accident sources from on-scene contamination? Post-mortem production of alcohol, which is particularly common in some of our aviation accidents, where it takes a high level of trauma with maybe a fairly good period of time before we can recover the body. For post-crash fires, we will look at the carbon monoxide and cyanide levels, and of course, if someone survives initially, to see if there's any resuscitation drugs that we take into account. And like before, one of our challenges that we really don't know if the person had built up a tolerance to some of the medications, and particularly for when it's not a professional user, is like, what was worse for them to be on a medication or to be under or not treated at all? So, for our third question is, did it contribute? And again, my other cases sort of point to that. So we have an impairing substance. In the case for Davis, Oklahoma, we showed that, yeah, this not only occurred, we saw the non-responsiveness, but it really was truly the cause of this accident. And as stated, we look for, have to look beyond just our medical records to make that determination. But sometimes, even if that answer is no, is there safety lessons to be learned? And for that, I'll turn to a rail accident that happened about eight in the morning. So what you see as the forward-facing video, well, just image shots from a forward-facing video on a passenger train going about 100 miles an hour on track three. So a little bit of background is two days before this happened, you could see track two was put out of service to have some reconstruction work done. But intermittently, you had to have either track one, three, or four fouled in order to allow for construction activity. So this is eight in the morning. The night foreman failed to disclose or present information to the day foreman about, hey, we still have equipment on track three. So the day foreman released track three for transportation. And at that point, any train was allowed to go at 110 miles per hour. So the engineer comes across and sees this way ahead, as you see from the left, in a distance, he's like, oh, there's something on my track. And so he starts a series of emergency braking as well as horn blasts to try to like, hey, I'm coming. We have inward-facing video that shows him continuing to do that as he goes into a crouch position, and then at the point of impact, he's actually lying on the floor. So unfortunately, both the backhoe operator and the supervisor were killed in this accident, as well as 39 people on the train were injured. Here's our post-exiting toxicology testing. The first is for the surviving engineer, which was drawn about five and a half hours after, a bit beyond their four hours desired time, as well as, you'll notice, we only have FRA testing because he survived, so we would not get leftover specimen. But you see, at five and a half hours, he's positive for THC at 2.2 nanograms per mil. And so for that, we really couldn't say, you know, again, not knowing his use pattern, is this impairment or not? But we looked at his behavior, and truly, he did everything he could for something that really couldn't be stopped. And the train, when it hit those backhoes, were now at 99 miles per hour. For our backhoe operator who died, we have both FRA and leftover specimen testing that were sent to CAMI. Unfortunately, we only have urine and tissues, but they showed that he had both cocaine and its various metabolites, amongst other things. But once again, he had no influence on this accident as well. Finally, we have the supervisor who, we had blood specimens from him, and what we showed from both FRA and FAA is that we have low levels of opioid in his system. Again, no idea in terms of use pattern or his familiarity with it. But once again, looking at the operational aspects of the accident, not really related to this, other than a finding. FRA did test the conductor, assistant conductor, and dispatcher with FRA testing. Those were all negative. And the company itself did testing for the night foreman, the day foreman, and the night watch, the watchman. And those were all negative. So, you look at our cause of this accident, and once again, the drugs are not mentioned. They're not part of this accident. But is there a safety issue? And for us, that answer was yes. So, we recognize that there was a lax safety culture that needed to be addressed, both on drugs and the other aspects of this investigation, and that's where recommendations were made to the company and other entities for that. So, to sort of segue into, like went over the three questions we asked, I also mentioned that we did safety studies. So, we're going to present a general aviation accident, and I recognize this one, unless you fly. Most likely, this is one of the modes you're least likely to get involved with. But it is occupational, because this is a pilot who is flying aerial agricultural applications. And according to the company pilot that witnessed it, he said, yeah, the guy loaded up the herbicide. He took off a little later than usual. The wingtip hit a power pole, which then hit some power lines, which then hit another power pole, causing him to flip and crash, and the pilot was fatally injured. He has over 17,000 hours of flight experience, and he worked for this company for a while. He was familiar with Cessna. So, we're talking about a very experienced operator. Post-accident tox shows that he had diphenhydramine in his system, and even accounting for potentially some post-mortem redistribution, that was definitely a therapeutic levels that were found. So, diphenhydramine, commonly Benadryl, over-the-counter, allergy cold, no problem, or a sleep aid, is one of the most sedating antihistamines around. And so much so that there's been studies with a driver simulator study that found like a dose of Benadryl is equivalent to having a BAC of 0.1. So, pretty extreme, as well as FAA, in their guidance, says you should wait 60 hours after taking this drug before you fly again. So, in our cause for the accident, we did also attribute it to the sedating effects from the medication on both his reaction time and his flying performance. So, as you know, side effects may include, and this involves all sorts of drugs, not just our illicit drugs, but also our prescription over-the-counter ones that are either taken with the intention of sedation, although those may have problems in themselves if they alter the sleep architecture, so you're really not as rested as you may think, as well as the side effects of these medications. We also look at the medications that might interfere with your sleep. Fatigue is often an underlying aspect of, again, from the human performance side that we look at. And truly, there's a lot of drugs that may make it difficult for you to sleep and have a good night's sleep, and therefore, potentially have a risk for daytime sleepiness. And if we do obtain records, ideally, we can tell whether, ideally, maybe these medications were given in the morning. So, we do try to look at that, but sometimes information is not available. So for our safety study, this was released, the report was released in 2020, we looked at drug use trends in aviation for fatally injured pilots for the tested for drugs from FAA. So the 2020 report was a follow-up to a study performed and that released in 2014 that looked at a 21 year period of this. So there was identical methodology in terms of how we figured out the data. However, as far as the laboratory goes, the data is most consistent between 2008 and 2017 because the FAA laboratory had not changed its detection levels and had not changed its panels in that period. So we were looking at the prevalence in these fatally injured pilots for anything over-the- counter, prescription, and illicit drugs. So these are the drugs that CAMI would test for. We use their database as well as ours, and remember we're looking at use. So even if there was an impairing substance, we didn't make a judgment cause whether that had caused the accident or not. So as mentioned, one of the reasons we do fatally injured pilots is because we have a really good database for them because again, we are mandated to look at all those accidents and FAA is sort of mandated to get the tax results on all of those. So you put it together and we have a good data set to work with. Now we only looked at the flying pilot. We excluded if there are other passenger rated pilots or pilots on board. So, you know, for a highway that'd be equivalent to like, we're only going to look at the designated driver, not the passengers. We excluded anything that was an intentional event and obviously if there was no toxicology performed for the availability of specimens. I mean some of our fatal crashes occur in the ocean and bodies are never recovered, for example. So with that said, from the older study that covered the 21-year period, we had 6,700 study pilots and in the newer study we had about a thousand study pilots. When we were looking at the medications, we equated a metabolite to the original drug. Obviously if we found both the parent drug as well as the metabolites, we only counted it once and if we happen to have multiple samples, we blood, urine, liver tissue, we only counted that once as well. So that would have been one positive finding. If the medication was only found in urine, we sort of assume we did not count it because we sort of said, well, it's already left your body. We can't really say, did it have any influence on you? And of course, if we can tell that it was from post-accident resuscitation or and as I said, we did not do any kind of ethanol findings. Post-mortem alcohol production is really hard to tease out. So what did we find? Well, at this point, one, you have an increasing trend over the since 1990 to now or till 2017. The darker blue is the more recent study. The lighter lines are from the older study. But there was always an increasing trend for the finding of at least one drug, which is the top line, as well as for pretty much for the findings of two or more, three or more drugs. We also looked at categories of drugs. So we had a big split between potentially impairing drugs, which would involve like, you know, somnolence, judgment calls, or they had that label about don't operate a vehicle on them, as well as categories related. We were interested in, well, what if they had a potentially impairing condition? You know, what does that drug mean they had that? Like something an anti-arrhythmic or something like that. We did have a category look to see how often we have schedule two through five, and we looked at illicit drugs, including, we would include the schedule two drugs that we obviously felt were either from recreational or abuse uses. So amphetamine, we looked at what level it would have been found at. So our top one is the yes or no, potentially impairing or not, and you can see that also has increased over time, with close to 30% of the pilots having some, and our fatally injured pilots having the detection of some potentially impairing drug in them. And you can see from the other categories that trend also has continued to rise. The blue one is the illicit drugs, which would include marijuana, so some of that might reflect more recent legislation in terms of accessibility to medical or recreational marijuana. I don't expect you to read all this fine print, but we also went through and looked at, divided up the types of medications to, hey, what is it used for? What's its chemical structure? So we categorized them into 20 different categories, and you'll see the top one was cardiovascular drugs. Now if you really could read the fine print and you go further down the line, oh the cardiovascular is about 18%, you'd see that diabetes medications are only at 1%. So that does represent the focus of FAA on their toxicology lab is really looking for either impairing drugs or impairing conditions. So for example, that they might be able to detect metformin, but they have a higher level of detection on that. Or some of the newer drugs they haven't really incorporated yet, so say all the gliptins, for example. That's one of the things while perhaps really our pilots are taking more drugs than even those. Detected for drugs and not necessarily impairing. So from that table as well, you'll see from our potentially impairing drugs, sedating antihistamines was the top category at close to 12%. Sedating pain relievers, which would include opioids, with hydrocodone being the most prevalent, was at 5.3. And the winner, of course, is diphenhydramine with about 8% of them, so it was of the pilots having some diphenhydramine in their system. And that would be the most common individual drug as potential and also the most common potentially impairing. Okay, so what do we do with these safety studies? Well, we do make recommendations with them and from the earlier study, as in this one, some of it's to get that information out to pilots and other stakeholders in terms of, hey, do you realize that these drugs can be impairing? So one was from FAA and out of that they had this publication issued for over-the-counter drug medications. So how can I figure out if I'm taking an okay drug? And it pointed out that, if you can see the circle there on their guide, it's like, hey, look at this, and if it says don't operate hazardous machinery, it's not that you can't take it, but you have to allow enough time. And what they use as a general rule of thumb is five dosing periods. So, of course, if it was take this every six hours, you should wait 30 hours before you'd fly your plane. Now, is that exact? Well, obviously not. I just told you the diphenhydramine has a long half-life, around 14 hours, so, and you probably take it every four to six hours, so you wouldn't actually obey that. But they recognize that the average person is not going to know a half-life of a medication, so the five dosing periods seem like a reasonably good rule of thumb for determination, or perhaps even just recognizing that you look for that warning label, and it's like, oh, maybe I take Allegra, because it doesn't have that label on it. What also came from it is, you know, idea of what our study was on fatally injured, and so we really don't have a way of knowing, is that the common pilot population? So that was also a recommendation, is try to study to see how much that the fatally injured pilots, the same drug use was represented in them. But since mostly we deal more with probably commercial drivers, we do have the National Safety Council, and this is beyond drivers, it's more like highway, and every four years they just have their 2021 update. So they have a committee that looks at some of the forensic laboratories, and they survey them in both the US and Canada, and they try to do this with the intention of getting recommendations to make standardization of what drugs are tested for, to make them, you know, defensible in court, and just to have uniform rules. And again, while the column on the right is a little small print, you'll see at least in the DRE category of CNS depressants, you will find that diphenhydrabine does make that list as well. So I looked at my watch, and I do have a little more time, so I'm going to just touch on another aspect that we do in terms of random drug testing. I promised. Okay, so most of you are familiar with the random drug testing. You have each of the agencies have what percentage of their covered employees or safety-sensitive positions have to be tested annually. And this goes up and down essentially based on a positivity of 1%, whether it's below or above, will influence whether the drug screening and the alcohol screening random testing rate has to go up or down. So this is an accident that happened at 5 in the morning. A westbound train was coming out of a yard, and it did not have the right-of-way. It actually had a stop signal, and the eastbound train was come, freight train was coming into the yard. The westbound train, not that it was, wasn't going terribly fast, maybe about 20 miles per hour, but it's a train. It hit the other train, derailing 21 cars on the eastbound train and four on its own, resulting in about five million dollars worth of damage. You know, fortunately, it was like frack sand and some trash, so it was at least not a hazardous spill. So our westbound engineer turns out that his post-accident toxicology testing six hours after the event was at 0.115. So if you do your linear extrapolation, which you can do with alcohol, probably at the time of the accident, he was about at least 0.2. He also was positive for marijuana, and even at the six hours in was at 6.3. Again, marijuana is a little challenging because you really have to know use patterns to know, is this what he's normally at or not? And those are for the metabolites, not for the active compound, because this is FRA testing. So the probable cause, as you might expect, included that it was due to the impairments to alcohol. Now having one little short aside on this, we were a little suspicious of this engineer to begin with because he had raised concerns about, during his interview, about his post-accident testing. So, and this highlights that people are not as accustomed to the FRA being a little bit different. So they collected the urine samples, closed the box, and then said, oh my goodness, it's FRA. We have to do blood. So they reopened the box, collected blood, put it back in the box, and closed the box. But for us, however, again, we're not regulatory. FRA can address those issues with however in terms of procedures. We were looking at the actual sample results. So for us, while it's an interesting aside, and for you I present that it's knowing what tests are required and when is probably very important. But for this case, as I said, we already know he had high levels of alcohol at the time, but we did look at his occupational medical, occupational drug testing records, as well as FRA supplies and drug testing records. And what you notice on the bottom four column of not performed is essentially a 10-year period where he was hired, I believe, in 2004. Anyway, there's a 10-year period where this man never had any random drug testing. Now, if you do a sort of binomial probability function, it's like, it's not totally unreasonable. It's maybe 1 in 20. You know, not the same as buying that Megaball ticket, but still surprising considering the conductor had, in that same 10 period, who has worked at the same yard, had gotten tested three times. We also looked at a report FRA had put out and showed that this this railroad firm company actually had some challenges with their random drug testing, some deficiencies, and in 2017 they actually hadn't met their rate. So, where's some other safety issue? We had our cause being the alcohol, but there were other safety issues identified in this accident, including the random drug testing program, and for that, then, some of our recommendations went to the DOT inspector general to sort of examine, you know, what is with the random testing program, as well as then to get lessons learned that they could share with other railroad companies. So, with that, I am finished, and I think I've left enough time for questions, but I think I've, hopefully, I've been successful in showing you what the three-step process, in terms of us appreciating impairment, in terms of looking at our accident causation or the medical factors involved with that, and as I said, while DOT is wonderful for trying to look at illicit use of drugs in the workplace, for us, it's insufficient in terms of where we have to go with other toxicology testing, as well as other records, in terms of making that determination for impairment, as well as looking at our operational and mechanical issues that might contribute. And, finally, hopefully, nice takeaways, you won't take Benadryl as you drive home tonight, but really, in terms of when you're talking to other, you know, the employees that you work with, that to recognize that both the prescription and non-prescription medications do have beyond the illicit drugs, do have an impact in terms of our accident causation. And I'm going to take questions. Did you need people to come to a mic to ask the question, or could they just ask the question? Okay, are there any questions? You seem to have your hand up, I'm sorry, and I don't mean to make you go to the mic, but. Yeah, I was just curious, in regards to the drug trend testing that you mentioned just a few slides ago, you said the urine testing wasn't included, or I just wanted to clarify. In the last accident? Just a few slides back with the drug trend use over the years, and the different colored slides, or the different trends. Sorry, it's a few more back. I'm sorry. That's all right. This one? A little further back. Tell me when. Oh, from the study? Yeah, the study. Oh, okay. That stuff. Okay, yeah. So remember, you can't add up those other three lower ones to try to add up to the top one. So as you might consider, is this part of your question? Am I answering for you? If it was only found in urine, we didn't include it. Well, we took it, again, we obviously would have had a higher number, because sometimes we only might have a urine specimen, but we took it as, if it had already been processed by your kidneys and is only in your urine, we can't say, if we could not detect it in your blood or in your tissues, it's no longer acting on your body, is how we took that. So it's really, I appreciate your nuance in terms of, we were looking at use, so, but we were looking at the time of the accident. Well, this is FAA, and FAA, we can do blood, urine, and tissue. These are deceased, so we often have tissue available for them. So it's not DOT testing, it's FAA CAMI testing. And yes, that's distinguished from FAA when it's the air traffic control or FAA pilot, commercial pilots. Yes, they would be DOT related, but for a fatal accident, we'll often have the CAMI testing, which includes 1,300 compounds. Does that make more sense? Thank you, Dr. Waters. A question on the diabetes drugs being found, the cardiovascular drugs being found, the mental health drugs being found. How does that compare to general population, though? Because we know that they're there, we know they're being used. Yes, well, according to the report, and again, since I, I think I said the doctor, our chief medical officer, Dr. McKay, and our chief data officer, Dr. Groff, had prepared the report. So in the report, they say, based on their literature, that was comparable to the general population. So you really can't draw conclusions that because there are cardiovascular meds or diabetes meds found in these significant crashes, that they were contributory? Correct. It was never meant to look, it was definitely looking at the prevalence of our tested for drugs, and it was not, even if it was a drug that we knew to be impairing, we did not actually go to see, like, look at the level to say, oh, yes, and it was at a therapeutic level, likely to cause side effects. The problem is plaintiff attorneys see those meds present, and they raise that red flag and say it had to be contributory because it was there, and it's not really any more than the general population, and that's part of the problem. Okay. Any other questions? Well, thank you for your attention. Really appreciate it.

Dr. Michelle Waters

National Transportation Safety Board

NTSB

DOT Drug Testing

Toxicological Evaluation

Accident Investigations

Transportation Accidents

Drug Use Trends

Safety Study