Episode Transcript

Rene:  And Rene, that's me. I'm back here in the studio. We're going to continue a conversation that Kirk had with Dr. Johannes Thul. And here's Kirk from the last episode, E173 Part 1, explaining a little bit about Dr. Thul.

Kirk:  Anyways, so what I discovered is that this guy, this guy belongs to a team. Of other researchers that are trying to set up a national database of medical cannabis users with the hope that they can have a library that they can distill out to people on cannabis. It's a fascinating story, national cannabis study that they're doing and their hope to gather information about medical cannabis from around America and the world. That's basically the story.

Rene: without further ado, here's Kirk, with Dr. Johannes Thul, on Reefer Medness, the podcast. 

Kirk:  First of all, let's start, this is a huge database you're proposing. How far along are you in gathering this information?

Johannes Trul:  Yeah, so at different stages, based on four different sub studies. For the prospective data collection, the national cannabis study was still relatively early on. So we built relatively a pretty robust system that has, you know, gone put, like been put through the paces. With online research these days, it is nontrivial to set it up so it's well protected. Early last year, we were savagely attacked by a swarm of bots and probably humans in the loop that initially it was very easy to determine what is a bot response versus a real participant, but over time they learn and get better. So it gets really tricky. So now we had to essentially re-engineer everything from scratch to build really robust safety security systems that now essentially requires folks provide an ID that gets checked against the facial biometrics to make sure that they're real participants because we pay incentives for survey completion which means that it becomes very interesting for people to scam research studies like that.

Kirk:  Sure.

Johannes Trul:  And with you know large-scale AI available to to anyone and everyone. There are a lot of people who have an interest in getting into research studies like that. So that's kind of the trade-off we have to make. We have to be really diligent about protecting the data quality.

Kirk:  Fair enough, which leads me to the next question in tombstone information. When I decide as a user to say, oh look at this study, I can participate in this, what tombstone information are you asking from the individuals? Are you taking zip codes, ethnicity, social terms of health? What information are you gathering about the individual besides their cannabis use?

Johannes Trul:  Yeah, so at Baseline, we're getting a lot of this type of information. Obviously, we want to know people's background and then a lot of the measures we have. This is maybe something we haven't talked about yet, but we have suggested measures on the website as well. One that we developed ourselves, the cannabis use questionnaire that gets at really detailed cannabis use. We talked about what are the facets that important as in, What are the ingredients of the product? What is the dose that people are using? What is their route of administration? Frequency, quantity.

Kirk:  And when you CQTs, you called them? Is that what its

Johannes Trul:  CUQ cannabis use questionnaire.

Kirk:  Cannabis use questionnaire. Okay.

Very Creative.

Kirk:  Yeah, yeah, yeah. But but but it sounds like you've standardized those questions, so that anybody who's doing cannabis research can go to that standardized list of questions. So that that is consistent. Is that one of the goals?

Johannes Trul:  Yeah, so we're hoping that it will be a service to the field to help put out what we think are the important facets of cannabis use that we recommend folks to assess. Now we've started talking about the changes in the marketplace, right? I don't think this is ever something that's static. So there will be new products on the market that maybe our current iteration of the instrument isn't capturing well. So the approach that we took is that, and so I guess I should start by saying that this tool is available for folks to download off of our website in ready to go versions for RedCap, for Qualtrics, common service systems, where you can just import it into your platform and run with it. You don't have to start programming from scratch, which took us months to do. But... Now, knowing that the marketplace is in flux. The products that are available today might not be the products will be available tomorrow. So this questionnaire, the tool has to evolve with time as well. So the approach that we took is that once folks want to download it, we ask them to register with an email address. So we have essentially a list of folks who are... Have at some point at least downloaded and looked at the tool, whether or not they're using it, I can say, but once we make updates to it, those updates will go out to everyone who we have has registered to download the so

Kirk:  So how deep, how deep do you get them for personal information? Do you get the person's individual names? And I mean, obviously, birth dates and zip codes would be helpful. But do you, do you down, do I put my first and last name into this database?

Johannes Trul:  This is not information that will be kept with the like health information, right? So the data will be kept separate from identifiers and then once, especially once the data will get analyzed or open to use by other folks. When people can request the data we have, those data will only ever be de-identified. So we will never share that kind of identifiers. Sure. And it gets some identifiers, obviously, to make sure that people are real persons and all of those things to begin with. But then other than that, the health information relevant data will be kept separate.

Kirk:  Well, I find it fascinating because as I was reading about your survey, and again, I'm sorry I'm keeping you longer than I said I would, but I hope you have the time. As I read through this stuff, I was saying, is there any other medical condition, pharmaceutical product that is trying to capture this? Like, does the MS Society in America have a big depository of information? Is this something new that you're doing here to grab national? And is it just focused on national or is it international? Like it just seems, it seems robust and new to me. I haven't seen this elsewhere.

Johannes Trul:  Yeah. So currently it's focused on the U S potentially we could expand to other places. Obviously English speaking countries would be relatively easy to do. No products might be slightly different. I mean, Canada would be an obvious choice because of national legalization of recreational cannabis, pretty robust medical market. Right. Um, I think I would argue we're probably the first in the US, at least, to assess cannabis use exposure at the level of nuance that we do and at the scale, especially for medicinal cannabis. The goal is to get 10,000 patients into this thing.

Kirk:  But you're identifying recreational users as well, are you not? Like if I choose to enter my name in this, can I do it as a recreational user or just as a medical user?

Johannes Trul:  Yeah, so, I mean, we're essentially taking people by their word that they're using cannabis to deal with a medical condition problem that they have. Now, the definition of what medical or medicinal use is, there's no one type of way to define it. Back in the day, maybe you could have taken, whether or not somebody is registered through a medicinal cannabis use like state program, they've talked to a provider, they've gotten what might be called quote unquote, a medical cannabis card. But now, in a state that is recreationally legal, you can just walk into dispensary buy a product, you know, use it because you have a headache or what have you, you never talked to provider, I would argue that person is a medicinal cannabis user. But they wouldn't ever show up in a state registered list of patients, right? I mean, you could walk into a CBD shop that is outside of the state system altogether and use CBD because you think it helps you with anxiety or with sleep or what have you. So that definition has gotten very tricky. And, and we do see in States that go from, usually the transition of States is that Cannabis is not legal at all, then they have a medical program and then eventually they go to full recreational adult use legalization. That's kind of been the trajectory in most States. And once we see the transition from just a medical to a recreational use, then you see that the recreational market takes a big bite out of the medicinal market that people just don't get a card anymore. They just go to the dispensary and get the product. Whether or not, you know, the States are very different in how stringent they are in allowing for medicinal use, like qualification. So there's no there's not standard across the country. And so you know whether or Not people just want to get cannabis and they qualify because of some kind of condition. And it's really tricky to say right, not everybody will meet the same criteria. But what we see is that once states come online with recreational use, then medicinal use tends to go down.

Kirk:  One question on the data, and then I'll sum this up. The data was positive. That is when you're asking people to submit their records, right? I know some States are collecting information on medical cannabis. So are you getting good response from other practitioners that are giving you their information?

Johannes Trul:  Yeah, it's a big tent approach we're doing here is to reach out to folks and really anyone who has relevant data sets to contribute. There is a process that we have set up with our IRB here at Hopkins to govern essentially, you know, data use agreements and then transfer agreements that folks can collaborate with us and, and deposit the data into our repository. Because a lot of the time... You know, if you live in the kind of soft money research world, you run a study, you collect data, you maybe get your primary paper out. And then a lot of the time you just have to move on to the next thing. And then there's still a data set that has a lot of value potentially, especially when you can potentially combine it with other data sets that then out of a sudden allows for analysis with much larger sample sizes, right? Where you might be getting to a meaningful number of folks with conditions that we wouldn't see in like a one-off survey or we'd only see like a handful of folks for specific conditions that are not as common as chronic pain, anxiety, PTSD, insomnia, like the big ones that you see in all of the surveys, but where cannabis maybe has potentially a lot of promise. And so by combining data sets and harmonizing in a repository like this. We could potentially make a meaningful contribution to improving the data quality and the sample size that is available to get after these questions.

Kirk:  Completely agree with you, sir, completely agree with you. Are you getting a good response? Are people responding to you? Are they saying, yeah, we'll share this stuff with you? Is it happening?

Johannes Trul:  Yeah, yeah.

Kirk:  Cool. Yeah, for sure. Cool.

Johannes Trul:  And but we always, you know, we're always interested in more,.

Kirk:  Obviously,.

Johannes Trul:  If folks are interested.

Kirk:  Now did I read someplace that you're accepting international databases for that response so if somebody in Canada had information they could, you're accepting international data repository whereas the national survey right now is strictly America and are you, the US, sorry, and are you also eventually going to all, all that's going to jive and become the library?

Johannes Trul:  Right. So the library is just papers, but the repository, potentially, this is kind of, you know, we have to see a little bit as we go based on what kind of measures are collected in individual data sets. Is there potential to harmonize across data sets? By harmonize, we mean, you know, this study collected insomnia in this way, in this study, another study, maybe in another ways there are potential to. Kind of combine these measures for some studies, some data sets, I think that will be an option for others, maybe not so much. It depends a little bit on what we see in the data that we actually receive. The other thing I will say is that we talked about the cannabis use question that we have on the website. We also have a suggested list of assessments for health outcomes or health aspects around cannabis use that we suggest folks use in studies on medicinal cannabis. We call that the C-A-B-C, so cannabis assessment battery. And that is also on the website for download so people can see, sorry, core assessment battery for cannabis. What are the measures that we recommend for the field to use and this is also a step in. That direction of might make it easier moving forward to harmonize across different studies and across different data sets. If some of those measures are used consistently across studies, obviously it will become much, much easier to then build larger and larger data sets to get at where the signals are on cannabis use and health.

Kirk:  And when do you feel it's going to be accessible? How, like, where are you and the timelines that you're going to say, hey, we've got the information. It's now accessible. And you did say it was a year long study. So is this something that will make, I guess, funding determines this, but is it going to an ongoing thing and progress as long as funding's there?

Johannes Trul:  Yeah, that's certainly our goal, is to keep this thing alive for as long as we can.

Kirk:  It's a huge, it's a huge project.

Johannes Trul:  Add more data in and, yeah, build something that can be really useful to folks in the space and to the field.

Kirk:  Well and as we said right up front, cannabis is such a misunderstood substance and I think from what I've heard from other things I've read and Trevor and I fill in the blanks of stuff but I think it's really quite cool. Is there any question I didn't ask you? Is there information that you would like to share that I didn t get into?

Johannes Trul:  I think we've all covered it. I would just say we're a very collaborative group. And if folks have an interest in working with us in contributing to the repository, in using the measures that we have developed, just an open invitation for folks to reach out and maybe potentially even the platform that we've built, if you know, someone says like, I want to run a study on a specific patient population, and could funnel them through our platform, collect the data, data will go into a repository anyway. That's always an option as well. We maybe don't all have to reinvent the wheel for new studies. If this platform meets your needs, we're also super happy collaborating.

Trevor:  Alright, so Kirk, I'd like to summarize stuff to see if I'm understanding what I'm listening to. So here's how I understand this three part, we'll call it database. So the first chunk is the National Cannabis Survey. So it's an actual firsthand asking people, hopefully before they've started using cannabis for their medical condition, asking them a bunch of questions. Your age, sex, whatever, what kind of cannabis you're using, and then hopefully after they've used it for a while for whatever condition, you know, did it work. So that's kind of the national cannabis survey. Part two is a database repository. So this is info, secondhand info from other studies, maybe insurance companies, maybe practitioner notes, but a bunch of data that's been de-identified so you don't know who the patients were and dumped into a big data pile that can be queried. And then their work on things like data collection, standardized data collection tools and ways to integrate all these various bits of data together. And then the third thing, our research library have actually published cannabis papers and I was just flipping through that this morning and I think there's a hundred years of podcast ideas and research just in there. And it was 3000 plus papers just in that third part. So yeah, this is big and interesting and exciting. And do you think I sort of got the summary half-ass right?

Kirk:  No, you nailed it. And that's what I stumbled into when he responded to my original emails. I went in and said, okay, who are you? And wow, look what these guys are doing. The intent is to build a platform of data collection that they can basically have a repository of cannabis information. It's fascinating. And I kept trying to ask myself and I don't think any other database like this exists. Like EMRs, like electronic patient records, are just now starting to come up. I mean, I can remember in the 90s, I had a very progressive family doc, Dr. Lindsay, Peter Lindsay, he has passed since, but Dr. Peter Lindsay was a very progressive family doctor. He liked technology, the man just loved technology, and he moved from Lac La Biche to Grand Prairie, and I went to Grand Prairie to to see if I could get some employment in Grand Prairie. And I met Peter and hadn't seen him for a while. Peter used to be my medical director when I was a paramedic professor, a college professor for paramedics. Peter, so Peter back in the 90s showed me his room, his computer, that was a patient, electronic patient record, like revolutionary, right? That's 30 years ago, Peter was doing this. And then, of course, I moved to Manitoba. Manitoba is a little, I don't want to say backwards, but slower. It doesn't catch up to Alberta in that sense. But EMRs to this day are still limited. I mean, not a lot of places have EMR. I think your clinic has an EMR,

Trevor:  we do and i'll we're trying to get too far yes and I've as my whole career i couldn't wait for them to have well they've had electronic medical records but of course the part that affects me the most is they have electronically generated prescriptions now which should have been the fantastic for all pharmacists everywhere. You know the famous I can't read your writing or what do they say or which doctor signed this should all be over now right. Yes but I didn't realize it introduces new problems, so the new problems being, you know, if you write it down on a piece of paper, you don't pick a completely wrong drug. You know, you might spell it wrong, I might not be able to read it, but you know if someone has high cholesterol, you wrote down a cholesterol medication. Well now that they're picking their drugs from a drop down menu, you now someone might have a cholesterol problem and they picked a birth control pill because it's so easy pick something wrong off a drop-down menu. So that's annoyance number one. And annoyance, number two is, with written prescriptions, you would physically get that from your doctor and walk over to the pharmacy and hand it to me. Every so often, you know, a couple times a month, especially coming from the emergency room to us, patients would lose that piece of paper between the emergency and here, and now it's a big struggle to figure out where the hell it was. That should all be over with electronic medical records, right? Nope. Apparently it is phenomenally difficult to hit the print button that goes to the pharmacy. Several times an hour am I trying to call up doctor's receptionist to see where in the ether this prescription is because the print button wasn't hit properly to go to the pharmacies, it's just lost in the ether. They can find it but it's annoying. Patients come and assume I'm the idiot that I didn't get it. The doctor said he sent it. It's not here. They didn't send it. And even worse is the hospital. The hospital has no idea how to send anything out of the hospital to somewhere else. And of course, when you call a hospital, you go through 14 different people. So you get to the right person who might know where, again, I guess the good news with electronic medical records, they'll eventually find it, but oh my goodness. I'm not trying to be a Luddite, but things went so much better when the doctor wrote it on a piece of paper and handed it to people. I don't care that I couldn't read it. At least I got it. All right, sorry. Rant over electronic medical record, you were saying.

Kirk:  Sorry, I was just having this flashback as you're talking to me about communicating with doctors. And it's always killed me how everyone else in the system must communicate and use proper language and are evaluated by how you communicate. But a doctor can come in and write something and scribble it. And you can't read it. So you phone them at three in the morning saying, what is this? What have you written? And you get in trouble for waking him up. Anyways, that's doctors in their communication. I mean. Yeah, no, electronic records, right? So I've had 30 years of understanding that they're out there, right? And when I worked up north, the federal government, I think there was two nursing stations that were moving into EMRs. The whole point about EMR is that EMR are supposed to be able to make things easier. And you would hope they talk to each other. Like one of my biggest pet peeves in the healthcare system. And this is my rant. Is that we are told everyone needs a family doctor. I don't believe everyone needs a family Doctor. I don't think the most important relationship is between patient and doctor. I think the more important relationship is between citizen and government. So. If I go to Alberta and I get sick, I should be able to walk into any physical doctor's office, but unfortunately, they're private entrepreneurs, but I should be able walk into a healthcare clinic. They should be to call up my EMR and say, oh, here you are, Kirk. Here's your EMR that your doctor and Dauphin wrote out. And I can see that here's your, you know, and carry on with my care. But the problem with EMRs is that they become silos. So my doctor has a wonderful database on me. But if I want to share that when I go to BC in a couple of weeks here and I, if, you know, I fall off my perch in British Columbia, they have to start from fresh and probably have stuff faxed to them. But my point being is that EMR should be, should be talking to each other, but they're not. But here we are. We have cannabis, we have cannabis and this national database. What other medication, what other, what? What other I can't I can even think of a word because is a medication condition has a national database and I was trying to ask that ask him if he knows of anything like does the MS is our national bank of MS patients is there.

Trevor:  No, and this is a very cool thing, and it's not the EMR's fault, it's the fact that health care is provincial, not federal in Canada, but the reason I can segue into that is in Europe, because we get really cool info out of Europe, about whatever, but lately it's been things like, you know, around COVID, because, like, England or the UK has the national NHS national health service and they can pull out you can go anywhere in England and get connected and they pull out chunks of info from the whole well the US is terrified of having anything linked, you know, not even just state to state, town to town. So they are completely against anything being nationalized information-wise and Canada is not a lot better, but back to in North America, no big national databases on stuff. Uh, almost don't exist, but I want to throw in Um, so one of the things with EMRs is the docs originally were supposed to all put in ICD-9s, these fancy codes to say, you know, this person has high cholesterol and this one has diabetes. It's a code to make it easier for a computer to find it and correlate and do something with it. And sometimes they did it and sometimes they didn't. And sometimes the, the EMR would work well and sometimes it didn't, but what until now has been very difficult is the doctor's notes or the practitioner notes the the free hand like typed but typed in there because there it's basically a story right you know i saw Kirk and he had a big purple toe and then he told me about his you know his mother's dementia you.

Kirk:  Your talking SOAP charting.

Trevor:  Exactly and i was going to explain SOAP chartting but i don't do it very much so Kirk just quickly what's a SOAP chart

Kirk:  SOAP charting is just that, subjective information, why the patient came in, what is the patient's complaint, what is a patient's story. The O is objective. So what do I see as a practitioner? What is my assessment? So the person comes in with a belly pain. I do a hands-on physical assessment and that's my O.

Trevor:  Right, their blood pressure, their... And, and, end.

Kirk:  Well, but my observation and my observation is my assessment. So my assessment is and I've said this before, you get a differential diagnosis when you see a doctor. The doctor never tells you what you got. He tells you what you don't have and he goes through the negatives, negatives, negatives until he gets to then it must be this. So you provide your you provide your differential diagnosis, meaning it could be this, it could be this it could be this and then you go with your plan. So SOAP and what is my plan? I plan this so your plan should relate to everything the patient says and your assessment. What sometimes happens that becomes very difficult and this is where I formed a real appreciation for the family docs. Is that sometimes the objective and the subjective don't match. The patient comes in with a subjective story and you do your objective assessment, you go, there's something missing here. And that's where it becomes difficult because sometimes the patients are coming in hunting for meds, hunting for their narcotics, hunting, you know, I've got severe pain, doc. So I, you, there are ways to elicit a pain response from a client And you can do it. I can't say, never say the word, serendipity, sir, I can do this serendipity. Yeah, as I'm assessing and touching a patient, I can throw a little to see how they respond and sometimes you don't get the response that you want but so a SOAP chart so in the federal system.

Trevor:  I'm going to jump in though. The reason we went on this rant and how it really does go back to Dr. Thrul's project is computers had a hard time reading narrative stuff, you know, taking a narrative like that. But now with AI, and he mentioned that, now AI actually can look at narrative stuff like that and make sense of it. And more, you know, it's a, well, you could do that with a human, you know, going and picking up, they could. A human can do a few, maybe a human can a hundred, a human can't do a thousand, or a hundred thousand, or a million narrative reports, AI can. So that's one of the cool things about them using AI on this is if you get a bunch of these narrative reports it can still pull out the useful information and try and categorize it.

Kirk:  Yeah, I found this a fascinating project that they're working on. I could go on and on about medical records, moving from one province to another. The Public Health Department was amazing. I got our immunization records for the family easy. And I went from one public health office to the other. The difficulty is working between doctor clinics, but that's a different story. So looking at this national cannabis study, there's a large team. It's out of the Johns Hopkins University Hospital. It's a study team. What I found interesting is that they seem to be psychologists and psychiatrists opposed to, like they're PhDs. There's not a lot of medical doctors on this team, it seems to me, unless there are some medical doctors buried in the PhD. This is truly an academic project, which makes it unique.

Trevor:  Yeah, no, and I think. That's probably the place to start anyway. You know, these are the researchers, these are guys with the training and the stats and whatnot. But I also want to throw out there. So that was the good part of AI was be, the AI should be able to sort of pull information out of narratives. The bad part, and he mentioned is, their survey has already been attacked by AI bots. So because you can be offered money to do some surveys, well, there's money to be made. So let's unleash a bot storm at the survey already. So they've locked that down, but like everything else, there's just every new technology, there's wonderful things it can do and terrible things. And yeah, the AI pulls out stuff from narrative, great. Attacks the whole damn survey, not so great.

Kirk:  Yeah, one of my one of my mentors years and years ago always said to me, you know, Kirk, this is great. We're problem solving. This is fantastic. You've come up with a solution. But now remember that all solutions come with their own problems. So so you can come up with a wonderful idea, but you have to now critically think about what are the limitations of the solution. I don't know. When I when I when I went into this into this story and spent my three days studying it, I really can't see any disadvantage of it except the fact, you know, what I know about the United States is that no state likes feds. I mean, any novel I ever read, any TV show I ever watch when there's state and federal issues, no one likes each other. So I can only imagine that there's going to be reluctance because this looks like a national project. But it's funded out of, I think it's funded out of Colorado. I don't have that reference in front of me, but I think I heard that because I listened to a couple other podcasts that this team was on, so I could listen to their story from other podcasters' perspectives. And I believe the funding's coming from Colorado, but yet it's John Hopkins, and I think they're in Massachusetts, right? They're in Boston.

Trevor:  On the East Coast anyway, but I think the good part, like you said, you know, we have our provincial silos and Ottawa bare not tell me anything about how to run my health care in Manitoba. I think I think there should hopefully should be less Less reticence, but this isn't run by the federal government So the fact it wasn't being run by government should hopefully make the states a little less less worried about being involved

Kirk:  Yeah. A couple of things I should, you know, ranting about EMRs and stuff, a couple of things I could pull out though that I thought was interesting that I'm going to explore further with other members of his team. We discovered that one of the members of his team is an alumni from UofA University of Alberta and I'm going to investigate some of some of the stories that she did in Canada but one of the things was the cannabis use questionnaire they have standardized they have standardize the questionnaire for cannabis use and they call it the CUQs and what they hope to do is that if any researchers are out there that are doing cannabis-related research. They have a standard set of questions that they want to encourage you to add to your survey so that consistencies are seen. And this is what interested me when I discovered this story was that what these guys are doing is, I think, revolutionary from a perspective of understanding how cannabis affects people in a standard wise. So it's very academic. And it's basically across the spectrum. And we go to our web page, ReeferMed.ca, the links will be there. And it is all participatory. You can sit here and fill out the forms if you want and add your information to their database.

Trevor:  Yeah, very cool and sorry about the rants afterwards, but Dr. Thrul, that was great and we wish the best to your team and this repository of information about cannabis. It's going to be fantastic. We're looking forward to looking at it afterwards and hearing some more research that come out of it. Yeah, keep us in the loop about how it goes because I can't see anything but good things coming from this.

Kirk:  Yeah, we're Reefer Medness - The Podcast. I'm looking for Canadian stories. As I said earlier, I have sent out many requests in the last couple weeks to, I went to AI, Trevor, and I asked AI, I said, give me some Canadian research stories on cannabis, like what's happening with cannabis in Canada. And it, you know, it didn't give me a lot. It gave me about five names.

Trevor:  Well, the good news is our listeners are smarter than AI. So listeners, if you know of some Canadian cannabis research, tell us. We would love to include them in this. Our listeners are smarter that AI. Prove me right.

Kirk:  Yeah, yeah, no, let's get some Canadian content in this podcast, because it's not that we're not looking for it. It's just not coming our way. This is an American centric story. But again, the information is relevant. And I think researchers in Canada, who are doing cannabis research should look at their cannabis use questionnaire. So again, questions become standard. I mean, that's the thing about research is keeping things standard so that they talk to each other, right?

Trevor:  Absolutely. All right. So been another good one. Have a good day everybody.