Good afternoon, everybody. Thank you for joining us on Friday afternoon. I'm Mike Dawes, I'm the host for this session and the moderator. We're going to mostly just do the talks and have question and answer mostly at the end unless there's a very distinct clarifying question, but try to save most of the questions toward the end. Each of our presenters will have about eight minutes to talk, so I'm going to be efficient here. Our first talk is going to be, it's entitled, Availability of Medications for Opiate Use Disorders in U.S. Psychiatric Hospitals. And our presenter is Srinivas Bhuvala, who is an Associate Professor of Psychiatry and Associate Program Director for the Addiction Psychiatry Fellowship at Yale School of Medicine, where he also serves as the Medical Director of the Substance Use and Addiction Treatment Unit, the SATU, at Connecticut Mental Health Center. He focuses on optimizing pharmacotherapies for substance use disorders and integrated treatment for comorbid conditions. He received multiple teaching awards, including APA's Irma Bland Award in 2021, and serves as a consultant for Connecticut's Department of Mental Health and Addiction Services. Dr. Bhuvala is also involved in national organizations such as ABPN, AAAP, and ACGME. With that introduction, let's go to him. Thank you, Dr. Doss. So today we're going to talk about availability of medications for opioid use disorder in U.S. psychiatric settings. So the timing is great, because this paper was just published in JAMA Open Network yesterday. So you could access this paper. I have no disclosures to disclose. At the conclusion of this session, participants should be able to describe the availability of medications for opioid use disorder within psychiatric hospitals, and identify the need to integrate opioid use disorder treatment into inpatient mental health and substance use care. I know I'm preaching to a choir, but just to talk about the opioid use in the U.S. Over 2 million people in the U.S. have opioid use disorder. Many think that this is an underestimate. It's probably around 6 million. This is based on the NSDUH survey. And over 200 people die every day of an opioid overdose. In 2022, an estimated 4% of U.S. adults, 9 million needed OUD treatment. Of these, 25% only received medications for opioid use disorder. And the majority of the substance use treatment is provided in specialty addiction treatment programs. Among the 21 million adults with any mental illness and comorbid substance use disorders, only 59% received treatment for either condition. And again, the majority only received care for mental health and not substance use care. So 1 in 10 with any mental illness and substance use disorder combined receive both treatments. That's less than 10%. If you're talking about CHF or diabetes or hypertension, there would be a national outrage to this, correct? But because it is an addiction, it is allowed to pass by. There's many reasons why this is happening, but I think if you look at why substance use disorder treatment is so undertreated, it falls into two buckets, mainly stigma and lack of provider education. Dr. Petrakis is here. We published this op-ed in 2018 in JAMA Psych, and this was to really have a call for psychiatrists to do more in responding to the opioid epidemic. We, at that time, said we are psychiatrists and mental health providers are uniquely qualified to provide treatment for our patients with substance use disorders. Yet, in the 2000s, 90% of the buprenorphine prescriptions were being prescribed by psychiatrists, and that shifted to only 30% of them being done by the 2015 or 16. And we thought the reason for that is more primary care physicians are taking up prescribing buprenorphine, which is great. And it's a welcome sign. Opioid use disorder treatment is being treated like hypertension, so it's good that primary care physicians are prescribing it. But then this survey, this study, was actually done by the federal government, came in where they looked at prescription fills by clinicians by specialty. And this was an eye-opening study. What it shows that if you look at the number of patients receiving buprenorphine in thousands by specialty, non-psychiatrists have been prescribing all the buprenorphine that has increased over the last few decades, and psychiatrists have really plateaued or even have come down. So we as a field have really given up on opioid prescribing. I'm not talking about addiction psychiatrists. We are in a bubble here. But I'm talking about more sort of general psychiatry giving up on addiction treatment. So there was a series of papers that came out in JAMA recently. And when we're talking about medication for opioid use disorder, we're talking methadone, buprenorphine, and naltrexone. So this paper, Cantor et al.'s paper, came out, which showed that there's low availability of buprenorphine prescribing and methadone prescribing in the outpatient settings. So based on that paper, less than a third of outpatient psychiatric settings are even offering buprenorphine treatment. So the critical step is expanding availability of medications for opioid use disorder in all health care settings, including psychiatric settings where there's a high comorbidity. It's estimated that 60% of our patients with opioid use disorder have some sort of mental health problem. So the aim of this study that we did was to look at the medications for opioid use disorder availability in psychiatric hospitals throughout the United States. And we used the 2022 National Substance Use and Mental Health Services Survey to look at the data. So this survey is done to look at mental health care and substance use care across the United States in all hospitals. It could be medical hospitals, detox programs, rehabs, psychiatric hospitals. So when they looked at 31,000 hospitals participated and facilities participated in the survey, out of them, 1107, that is 1,107, identified as being psychiatric hospitals. Out of those, 92% answered the questions about medications for opioid use disorder. So our data included looking at a little over 1,000 psychiatric hospitals, out of which 428 are freestanding psychiatric hospitals, over 500 are inpatient psychiatric units in general hospitals, and about 35 state hospitals. So this sample really includes all psychiatric hospitals in the U.S. So what the alarming findings were that less than 50%, 47.99% offer medications for opioid use disorder. Now when we say they offer medications for opioid use disorder, we don't know what they mean by that. It could be that they're just continuing the buprenorphine that is prescribed in the outpatient clinics and they're not initiating the buprenorphine by themselves. So the actual number might be much, much lower than 50%. And we've looked at many different parameters, but I'll just focus on some. If a facility is prescribing alcohol use disorder medications, they are more likely to prescribe opioid use disorder medications. If they are providing medically managed withdrawal, then they are more likely to prescribe buprenorphine. What's alarming is even in those who are providing medication managed withdrawal, 30% of those facilities are not offering buprenorphine and agonist treatment, which means that we are detoxing these patients and discharging them after detox, which again we know that increases the risk of overdose and death. Facilities with more substance use disorders are not likely to prescribe more medications for opioid use disorder. In fact, 22% of the facilities said they do not have any patients with opioid use disorder, which is again alarming because it's probably that they're not screening patients with opioid use disorder. And even in those who screen, those who have high rates of opioid use in their communities are not offering buprenorphine treatment. There's limitations to this study. There's potential bias from the facility self-reporting MOUD provision. Again, they might be saying they are providing buprenorphine, but they could be providing them for detox or they could be just continuing the medications that were started in an outpatient facility when they get admitted to the hospital. It doesn't mean that they're screening all their patients and starting the medications themselves. There could be a social desirability bias. This is a public survey, so they might be saying that they're prescribing it just so that it looks good out there. So there might be some overestimation in the MOUD availability in these settings. Though not all facilities reported MOUD provision, the high response rate suggests that there's not much non-response bias, which is unlikely. We analyzed the data looking at all the parameters, and we don't think that the non-response bias is affecting any of these findings. So to conclude, the study exposes the deficiency in medication for opioid use that are accessed in inpatient psychiatric hospitals, which is a possible avenue to improve MOUD provision, particularly in people with comorbid OUD and mental illness. This is a great place to intervene, to change the course of the illness in inpatient psychiatric settings. These patients may not seek outpatient treatment once they leave the hospital. We also know that, looking at ED settings and other medical settings, that if we start the medications in the emergency room, patients are 78% more likely to connect to care, versus 30% if we do not start buprenorphine at the ED visit. This is post-overdose. So this really provides an opportunity to change the course of illness for our patients with comorbid psychiatric disorder and opioid use disorder. I just wanted to acknowledge my co-authors. Sean Cohen is an assistant professor of medicine at Yale. Tamara Betham is a public health student who helped us with statistical analysis. And Dr. David Filene is a professor in medicine and addiction medicine. So thank you so much. Our next presentation is Substance Use and Shame in Minority and Immigrant Communities. And the presenter is Zane Elrey-Schroed, who is a rising fourth-year medical student with a strong interest in psychiatry, particularly in addiction medicine, forensics, psychotherapy, and interventional psychiatry. Zane values the stories and experiences shared by those he interacts with, considering them his most valuable gifts in his journey. With that, can I thank you. Thank you. So some educational objectives, we'd like to identify and some background of people who are age and imagine that number is probably higher after COVID. Current studies that look at these two topics. And really, our study, we look to examine shame and how it interacts and correlates with substance use, particularly in the subset of immigrants and minorities. So we hypothesized that shame and these negative perceptions that come with it have an impact. in all locations that you'd expect to. included demographics, shame, self-esteem, and substance use history. For shame and self-esteem, we assessed those using the internal. Then we also had an investigator design query to ask about specific substance use related problems and specific substances used and their perceived shame before their substance use started and afterwards. So for our significant findings, probable substance Poor self-esteem was We also found that frequent and high levels of shame were significantly associated with cannabis problems and alcohol problems. And just to give you all a kind of a breakdown, our sample size was very limited and that included about 55 percent of... It also took place in the southeastern part of Michigan, which had a high concentration of Middle Eastern individuals. Some additional findings. Probable substance use disorder was associated with cannabis problems, nicotine problems and poor self-esteem, and then poor self-esteem itself was significantly associated with probable substance use disorder, with the statement of wanting to cut down on substance use in the last 12 months. And also there are specific statements within the ISS, the internalized shame scale, such as replaying painful events over and over in the mind and never feeling quite good about So, just to wrap up or to think about our results, these lower levels of internalized shame is associated. Lower substance use led to decreased shame, although it's probably the first. And this sentiment, if I can, what I'd really like to present here or what you can get from this project is that this sentiment that has been coined as the healthy immigrant effect. us. So some hypotheses or potential thoughts about why this is happening is that for the lower shame among these foreign-born individuals is that perhaps these people who came to the United related to substance use and shame were found. Contrary to our expectations, Caucasians and U.S. born individuals were more likely significant to report poor self-esteem and screen positive. And these findings, they challenged what we thought originally. housing and financial insecurity that they would potentially be the ones who would deal with higher rates of substance use disorder and higher rates of shame and ultimately these further studies would better help us understand So this study, we had a flyer that came with a QR code for our survey, and our survey had the two devices, the ISS as well as the TICS, and then our investigator designed queries, and they were placed in locations such as health clinics, mental health clinics, public libraries, coffee shops, gyms, barber shops, public locations, and then spread by word of mouth to colleagues, people at religious centers, really trying to get a diverse breadth of individuals to take the survey. We actually also performed a literature search on potential locations that you'd expect to find individuals with substance use disorder, also locations where you'd expect to find individuals who didn't have potential substance use disorder. Yeah, so it was obtained over a three-year period, 2021 to 2024, and the study is still ongoing. They're still collecting participants, but this is up to this point. So, that actually goes into more detail in our paper, but we found that first generation immigrants actually had lower rates of shame and substance use, and we found that second generation had higher rates of substance use. That's definitely something to look at. Absolutely, and I think that we go into further in our background in our paper on how cultural backgrounds interpret shame and how that impacts their lives. We didn't get into the specifics of, it was an anonymous survey, so individuals were just asked to complete it just for the sake of maintaining the integrity. So, we had 55 total respondents, and then 33 individuals completed the survey till... No, they were posted on walls, bulletin boards, links were sent out, posted on a social media page. Not for this current study, but for these 33 participants, they were all flyer based. So it was a Qualtrics survey. They either didn't complete 95% of the survey, and that was the biggest. Unfortunately, they were not in different languages. That would have been amazing to have that in our study, although this was, we had limited resources when creating it, but definitely if we were to expand this study, we'd love to have that. These are our references, and thank you so much for your time. Okay, our next presentation is entitled Outpatient Addiction Care Outcomes Between State Opioid Response Grant Funded in Traditionally Insured Patients in a Medicaid Non-Expansion State, and our presenter is Toa Le. He is a psychiatry resident at MGH McLean. Great, thanks everyone for being here today. Thanks to the AAAP for the opportunity to present this work. My name is Thao Le. I'm a psychiatry resident at MGH McLean. Today I'm excited to share with you the results from this study whose title is posted. And this study was conducted during my time at Vanderbilt University as a medical student. It was conducted at Vanderbilt University Medical Center in Nashville, Tennessee, which is a Medicaid non-expansion state, and I'll go into the significance of that in a little bit. And before I start, I'd like to thank my co-authors, also mentors listed here, as well as my current residency program at MGH McLean for allowing me to be here to present this work to you. I have no relevant disclosures. So by the end of this session, I hope that you'll be able to understand the differences in demographic, social, and clinical characteristics between patients receiving MOUD funded by the state opioid response grant and those receiving MOUD through traditional insurance at an outpatient longitudinal addiction and co-occurring disorder clinic that I will call LACC, and identify how these differences might impact care outcomes between these two different patient populations, and then finally appreciate the potential impact of state opioid response grant funding or SOAR grant funding on addiction care outcomes in otherwise uninsured patients in Tennessee. So before we dive in, I'd like to give you a little bit of background on the state opioid response grants or SOAR grants. So these are grants launched by SAMHSA in 2018 in response to the opioid crisis, and the goal of these grants is to increase access to MOUD and address stimulant misuse through a continuum of treatment options anywhere from prevention to treatment harm reduction and recovery support services. So within 2018 to 2023, these grants have helped deliver MOUD and treatment services to over 500,000 individuals with opioid use disorder distributed over 6.6 million naloxone kits and provided recovery services to more than 1 million people. And the target populations of these grants are high risk and underserved groups, including uninsured and underinsured populations. And this leads us to Tennessee, which is the Medicaid non-expansion state severely impacted by the opioid epidemic. So Tennessee is one of the 10 states where Medicaid has not yet expanded, as you can see here. I wonder how many of you here are from these states. And because of Medicaid non-expansion, Tennessee's rate of uninsurance of people who are under or uninsured is 9.4% as of 2022. And for comparison, Massachusetts' rate of uninsurance is 2.4% and Texas is the last in the nation with a rate of uninsurance of 16.6%. So Tennessee's is 9.4%. And because of this 9.4% of the populations not having access to health insurance and also because of the rate of OUD in the state, the state of Tennessee has utilized SOAR funding program to improve access to OUD treatments through a hub-and-spoke model. And this consists, for Tennessee, consists of four systems located in areas of greatest need and one of them being the UMC or Vanderbilt University Medical Center here located in middle Tennessee that serves as a hub agency for middle Tennessee. And as a central hub for addiction care. So, for HUB and SPOKE model, as you can see here, for example, if we pick Middle Tennessee, which is this region here, VUMC is an academic, large academic institution that serves as a HUB that provides all of these, like a continuum, comprehensive range of addiction care services that will then collaborate with these SPOKE facilities that are included here. And the goal of that is so that patients who come for OUD care at Vanderbilt University Medical Center can then be referred to these SPOKE facilities, and the SPOKE facilities can also refer patients back to VUMC and provide support in terms of education, resources, and serves as an addiction specialty referral resource for the SPOKE agencies. And in terms of Medicaid non-expansion states, the definition of that is there is Medicaid expansion that the federal government offers to individual states so that they can use that money to then cover the coverage gap that is left in the ACA, as part of the ACA. And because of Medicaid non-expansion, there is this gap of people who are doing well enough that is not, whose income is above the poverty line, and therefore cannot receive Medicaid, but they are not doing well enough to, obviously, afford buying insurance on their own. So as I mentioned, VUMC serves as a central hub for addiction care in Middle Tennessee that provides access to MOUD, clinical treatment, recovery support, and care coordination for under and uninsured persons with OUD, via a range of services, including many more. And the two that I will highlight here is the Bridge Clinic, which serves as a low-barrier, short-term clinic that provides integrated behavioral health, primary care for hospitalization with OUD post-discharge, and these include insured patients, uninsured patients, and these insured patients are then enrolled into SOR funding so that we can use SOR funding or SOAR funding to cover their care while they're at the Bridge Clinic and at Vanderbilt. And then many of these patients were then transitioned to the Vanderbilt Recovery Clinic, which is the LACC that I mentioned, a co-located longitudinal addiction and co-occurring disorder clinic that serves both SOR funded and traditionally insured patients who have OUD and co-occurring disorders. So this brings us to this study, and the objective of this study is to compare addiction care outcomes between SOR grant funded and traditionally insured patients receiving care at this LACC, and the larger objective of this study is to inform health policies and resource allocation strategies in states with high uninsurance rates like Tennessee. And this is a retrospective cohort study using electronic medical record data from patients who were identified to have received MOUD at this specific LACC between 2019 and 2024. The participants in this study were 606 patients with 472 were in the traditionally insured group and 134 were in the SOR grant funded group. And the data that we collected from these patients included the demographics, social and clinical characteristics, and the outcomes that we would like to target here are grouped into four categories. One is treatment engagement, MOUD adherence, substance use outcomes, and acute care utilization outcomes. So in terms of demographics, social and clinical characteristics, we see many significant differences between the two study groups. As you can see here highlighted in the yellow are the significant differences. The SOR grant funded individuals were more likely to be male, had higher rates of comorbid conditions including tobacco use, cannabis use disorder, stimulant use disorder, PTSD, and HIV infection. While having higher rates of comorbid conditions, these individuals are identified to have lower rates of medications for said comorbid conditions. And these included medications for tobacco cessation, medications for AUD, so alcohol use disorder, stimulants, benzodiazepine, mirtazapine, SSRIs, SNRIs, antipsychotics, and alpha 2 agonists. So I will explain that in the next slide. So because of the demographic and social and clinical significant differences between these two groups, we use a generalized linear models with covariate balancing property, propensity score weighting. So basically using this statistical methods to account for group differences and the covariates that we included were age, social vulnerability index, which is an index that represents social drivers of health outcomes, sex, race, rich clinic attendance, common comorbid conditions and medications for these conditions. So here is the unadjusted data. So this is before we applied the statistical method to account for the group differences. So looking at just unadjusted data, we see mixed outcomes among the SOR grant funded patients. So as you can see here, here is how the outcome categories are divided. So here we have treatment engagement, which included frequency, how many clinic encounters that a patient has per month within the first six months of, within the first six months of clinic enrollment, and then attrition, how many days the patient had fallen out of care during that six months period. And then we have retention, how many patients are still have at least one clinical encounter around that six month mark. We also have MOUD adherence, which is, which included medication possession ratio for buprenorphine. And then we also have rate of buprenorphine positive urine test results. So that is the number of positive urine test results over the number of overall urine test results to account for potential differences in clinical encounter frequency. And then we have non-prescribed substance use, also use of rate of substance positive urine test results. And then we also included naloxone prescription. In terms of acute care utilization, we have ED visits within 30 days of clinic enrollment. Within 90 days, we have hospitalization within 30 day and then 90 day, and we also included length of hospitalization for those admitted during the six months period when they first enroll in the clinic. So with that, before statistical adjustments, we have SOAR grant funded patients had higher rates of treatment frequency, higher rate of MOUD adherence, however, they also have higher rate of return to opioid use and higher rate of stimulant co-use. When we apply the statistical adjustments, so adjusting for differences in demographic, social and clinical characteristics, there were no statistically significant differences detected in all of the categories tested and all of the outcomes that we were looking for. However, when we look at length of hospitalization, the SOAR grant funded patients had longer hospital states compared to the traditionally insured groups, regardless of whether or not the statistical adjustments is there, suggesting a very big response rate or effect size. So in conclusion, there are a couple of highlights that I wanted to point out. One is there are significant differences between these two patient populations when it comes to their demographics, social and clinical characteristics. The second highlight is treatment works. We see that uninsured individuals can achieve very similar treatment success when access to care is provided and when all of these demographics, social and clinical differences were accounted for. However, disparities remain as we see in the longer average length of stay among hospitalized patients in the SOAR grant funded group. And that points to more complex health and social challenges that these patients encounter that would require additional support beyond what SOAR grant covers. And that leads me to the last point, which is that funding mechanisms like SOAR grants are necessary but not sufficient to provide comprehensive and effective care needed to address the disparities in addiction care outcomes, particularly in this high uninsured patient population. And here's the references. Thank you for your attention, and I'd love to hear your thoughts and any questions you might have. So, the questions is, if there are any, I guess, hypotheses that I might have about why the length of hospital stay is longer. That's a great question. We actually see that in a different analysis on similar patient populations as well. And I don't have any substantial, like, objective outcomes for you. But I guess my guess is that the patients, or from my encounter with those patients, those patients are more plugged in for care once they get into the hospital. So, they have a high comorbidity, like, high disease burden that once they're hospitalized, they would need, you know, orthopedic surgery, general surgery to take care of other things that might have been prevented had they had access to primary care and other specialty care other than addiction care. Yes, they did. And that's the medications and the comorbid conditions that we adjusted for. We didn't use like a standardized disease severity index, one of those indices. So our next presentation is entitled Predictors of Treatment Admissions for Cannabis Use in the United States and our presenter is Saral Desai. He is a certified clinical research professional and a PGY2 psychiatry resident physician at Tower Health, Drexel University College of Medicine program in Philadelphia. He is passionate about research in psychiatry and actively involved in research topics related to substance use disorders, public health, and health risk behaviors. As a resident, he continues to teach and inspire medical students and residents to join him in his research endeavors. Thank you so much for having me here. Clearly, I forgot to update my bio because I'm PG3 now and I'm really happy about it. So I'm going to keep it short and sweet. We're going to talk about predictors of treatment admission for cannabis use in the US. So, none of the authors or co-authors have any conflict of interest to disclose. Ah, come on. Okay, next one. So, at the end of this presentation, the hope is that audience will be able to identify demographic, socioeconomic, and behavioral predictors for treatment admission for cannabis use in the U.S., and also be able to assess the impact of cannabis use pattern, including early use, frequency of use, route of administration, impact of comorbid mental health and substance use disorders, and then impact of referral source on likelihood for treatment admission. Next slide. So, a little bit of background. As we all know, cannabis legalization in U.S. is increasing. More states are now allowing cannabis to be legal. And with that, the use is also increasing. And it might be seen as a progressive movement, but there are also risks associated with that. We all know a patient who shows up with medical marijuana card or uses marijuana as a treatment of anxiety, and sometimes they don't even mention or consider it a substance unless you ask them. So, it's really important to know who are the individuals who are getting treatment for cannabis use disorder rather than just casual use. So, once again, we're going to focus on demographics, socioeconomic, behavioral predictors. We're going to move to next slide. So, in terms of demographic factors, we included age, gender, race. In socioeconomic factors, we included education, employment, living conditions. In behavioral factors, we included early use, frequency, and route of administration. And then we included mental health and comorbid substance use as a combined variable. We can move to next slide. So, in terms of methods, we use treatment episode data set admission part. This is SAMHSA's data. It's a national data, which is every year there's a new data that's published. So, for this, we use 2020 data. TEDS data includes all substance use facilities, both private and public, that receive federal funding. And the unit of analysis is admissions, not individuals. So, if a person had admission for cannabis use two times in a year, then it would be counted as two admissions. And for our study, we included admission for cannabis, where cannabis was reported as a primary substance. And this report was self-report. There was no UDS done. So, in terms of analysis, we use univariate analysis was done using chi-square test for categorical and then t-test for continuous variables. And then, to identify predictors, we used multivariable logistic regression analysis. We can move to next slide. So, some of the key results, there were 1.5 total admissions for that year. And out of that, 9.1 were for cannabis use. Higher rates of admission were seen in male versus female. And then, in individuals aged between 12 to 24 compared to older individuals. In terms of race, African-American individuals had higher prevalence for admission for cannabis use. And in terms of co-occurring substance use, they have included pretty much all substances that are out there. So, it was difficult to include all substances on the slide. So, we have just included the most frequently reported substance, which was alcohol, which was most commonly reported as secondary and tertiary substance. We can move to next slide. So, this is just a visual presentation of what we discussed. If you want to spend a minute and look at it. We can move to next slide. So, let's talk about the main results, predictors of treatment admission. This is multivariable logistic regression analysis. This was a really large table, like two pages long. So, I have decided to just include key results. So, in terms of demographic factors, again, younger age had higher odds. Male gender, African-American race had higher odds. As it makes sense, early use before age of 11 was associated with really high odds of treatment admission. Same with daily use. In terms of route of administration, smoking was associated with really higher odds of treatment admission. They have just included as one category. So, it could be either. Interestingly, in terms of socioeconomic factors, higher education, full-time employment, and independent living, these were the factors were associated with higher odds of treatment admission. Yeah, sure. So, this is interesting because when we are thinking about substance use disorders, there's functional impairment, there's a lot of homelessness, unemployment. So, this results kind of tells us that people who are getting admitted for cannabis use treatment, they are functional, they have a higher education, employment, stable living. Same thing goes for referral sources. Educational referral and employer referral had higher odds compared to self-referral. So, the whole picture seems like these are the individuals who are going to college or school, and then the school is, or college, they are referring them for treatment, or employer are referring them for treatment of cannabis use. You had a question about smoking route? Yeah, it's really high, and our statistician, he run the model a couple of times. We also did area under the ROC curve, and for this model, the value was 80, which means it's a good fit. So, in a way, it's unusually high, but it makes sense. Anything that hits brain quickly, it's going to produce that high. So, it's a survey-based question, so based on administrative data. So, they have categorized smoking, route of administration of smoking. They specified how they were smoking, versus eating it and other categories. So, for smoking, the reference variable was edible route, and for demographic variable, older age above 65 was reference group. For race, Caucasian group was reference, because it's the largest group. And in terms of referrals, self-referral was the reference group. And in terms of living arrangement, homelessness was reference group. So, yeah, this was the order of that. So, usually there was like socio-demographic factors, and then other predictors. Yeah. No, this, yeah, this is treatment admission to different substance use facilities, which could be IOP, inpatient, different facilities, so that's what it means. Yeah, outpatient would also be considered treatment admission. Can we move to the next slide? So, in terms of conclusion, we identified some of the key factors that are associated with higher odds or risk of treatment admission. And so, in terms of implications, this results could help us tailoring targeted intervention for this specific population as a preventative measure. We identified a lot of comorbid substance use, so coming up with integrated treatment approaches could be another thing that could be implemented. And again, on a policy level or on a public health basis, targeting high-risk group and providing them ease of access could, again, make a difference. Can we move to the next slide? This is just talking about the same thing. Next slide. Yeah, if you have any questions, that's my email. Any other questions? Yeah, go ahead. I'm assuming that 12-year-olds would not be able to buy from dispensaries, but I'm curious as to whether there are higher rates of, you know, admission in patients that buy on the streets versus dispensaries. So I'm curious. No, that's a great question. Unfortunately, this data does not ask that question specifically because this is not just about cannabis. It's about all substances. So for at least this, they just ask primary substance, secondary substance, tertiary substance. And in terms of characteristics of use, they only have characteristics that I mentioned, like when they start using, route of administration, how often they were using, and stuff like that. All right. Thank you. No problem. Thank you for a great use of not-so-great data. The reason I say not-so-great data is that every one of us in the treatment industry, and I'm sure you all in hospitals and such are not exempt either, we got somebody in the billing department who is watching what gets paid for, what doesn't get paid for, and is constantly giving feedback at multiple levels of the system about what's likely to be paid and what's not. Which can distort time trends in who's admitted in the first place. It can distort time trends in what diagnoses we bother to write down. And it can really, really distort time trends in what diagnosis we put first. Because every one of us who's got people saying, if you want this place to stay open and help anybody, you got to help us by doing the paperwork our way. Yeah. And I totally agree with you, that's such a big challenge when it comes to using administrative data. There's also this national inpatient sample that includes, it's administrative data, same story, depends on what diagnosis they're putting, and there could be multiple reasons why they're using specific diagnosis. There could also be error in adding the diagnosis, so that's also another limitation. But same applies to when we are diagnosing patients, for example, major depressive disorder and anxiety disorder versus MDD with anxious distress. Like if you do research on that, then you're going to get into the same problem. So that's one of the major challenges when it comes to using this type of data. So totally agree with you. that there were family relatives suffering from any kind of substance abuse problem. Secondly, what were some of the common mental health disturbances, co-morbid? So again, there are so many factors that we could study when it comes to substance use treatment, substance use disorders. This data is limited by the factors they included. So unfortunately, it was not possible to include that. The other, what was your other question? Yeah, so they did use DSM-5 diagnosis. So if someone shows up for admission, they'll get substance use diagnosis, and they also had DSM diagnosis, like bipolar disorder, schizophrenia. It was hard to fit all the diagnosis in this model as an individual diagnosis. It kind of ended up being really complicated. So we computed a variable that included like mental health disorder as a combined entity and see if that increases the odds and kind of added that to the model because there are already so many variables. What was your decision rule? So we thought that it could be a different study where we are assessing if specific mental health diagnosis is increasing risk of treatment admission for cannabis use. But with all the variables we have included, adding that on top of it, adding like 15 diagnosis was like really tough to do. So that's why we kind of decided to combine that. whether the racial demographics is skewed by where your patient population is, and if that would change if the study was either expanded or included a different racial demographic. It was kind of difficult to see from the back. Your slides, but I was curious. That's a great question, and that's one of the challenges when you're using local data. Fortunately, this is national data, so this is SAMHSA's data, which includes all the treatment facilities that receive public federal funding across the U.S. Yeah, but also federal funding is, to some extent, also kind of skewed. Yes. So I was curious if you went further to kind of look at the demographic, Uh, yeah, I, so, like, to look into what exactly the sources were? Yeah, if I, I mean, correct me if I'm wrong, but I'm thinking that most of the federal funding would be targeted towards lower-income or resource-limited neighborhoods or areas, and that tends to maybe skew the demographic makeup, and I was just curious if, even if you've, this data, or somehow, would it change if it doesn't, if it's not data just from federally funded? Mm-hmm. What if you had included, like, privately funded health insurance carriers or something like that? Yeah, so, for this data, they did have public and private facilities that receive federal funding, but you're, you're absolutely right. If we include other facilities that do not receive public funding, they might be treating different demographic, and that could change the result, but in the end, it's, there's always going to be, there's no perfect study. In order to get that data, it's going to be really difficult to get that data, and even if you get it, there are going to be other challenges, so it's like, is it better to have that? Absolutely. Is it possible? Well, that's, at least that's on my, in my hand, but in ideal world, that should be something that we should do, have that data that includes all treatment facilities, and that would kind of give us better, you know, more accurate results. Right, and especially if we, I, I think, like, your study is very good, and it would maybe help guide policy changes, so I, I don't know if you'll. as to what you think based on policy. Yeah, in terms of legalization, it's already affecting people. More people are using it. More research is coming out in terms of using it during adolescence. Increased risk of psychosis, schizophrenia. Same thing we see in our patient population. Some of them, their psychosis gets worse. So I mean, based on the results, these other individuals, they're at higher risk of admission. So proactively screening them and providing them resources for treatment, that could be one thing. Also, evaluating if once they get treatment, are they staying, are they improving, or do they need treatment again? Doing analysis, yeah. Yeah, for this data, they only have included admissions. It's not possible to identify individuals, at least based on the public data they release. It's possible they might have a private file, which, yeah. Just a quick question, I was wondering, what other routes of administration were you seeing in the data? I know you mentioned GOMIs, and what other routes were you seeing? And the reason I asked. I have to double check in the code book, but based on what I remember, one was oral, one was injectable, which didn't make sense, and I think those two I remember, but there might be more, but I have to check on that. Yeah, no, I'll definitely get back with my statistician. Why we ever. variables you had there, if you held that constant, would you think that might be possibly confounding the data? Because people were being referred to treatment. model? Yeah, so we used multivariable logistic regression, so we did kind of adjust for other variables. So this result, at least, it indicates that we adjusted for the referral source. In terms of explanation, like you mentioned, it could be highly educated, they're seeking treatment. It could also be that they are not that functionally impaired, and based on referral sources, system and employers, they are kind of maybe asking them to get help in order to continue with their education or to continue with their job, so that could also be the case. odds ratio, but I wonder if this has anything to do with the fact that most people probably selected smoking for their route, and there's just so few responses. I'm just hypothesizing here that there wasn't a lot of responses for, like, the oral route, and so perhaps the vast majority of the admissions were coming from that group, and it's just. Yeah. Yeah. Yeah. I don't know the absolute numbers for either category, but I'm just guessing here that the oral route was pretty small. That might be something. That makes sense. I'll double-check with my statistician. For this matter, I think age, age of an where the psychiatric hospitals are using buprenorphine naltrexone or methadone. privately insured, right, as compared to our So for the study that I was looking at, looking at addiction care outcomes between SOAR grant-funded patients and traditionally insured patients, I think the demographic or really clinical characteristics that stood out most for me is the juxtaposition between higher rates of comorbid conditions, so comorbid substance use and psychiatric conditions, coupled with the lower rates of medications for those conditions in the SOAR grant-funded group that really stood out to me. Were there social determinants that were barriers, though, for the two different groups that you could parse out in your model? Yes, I mean, for the two different groups that I looked at, it's really insurance status is what we were trying to go for, and one is SOAR grant-funded. And the SOAR grant only covers addiction care, not other primary care or other specialty medical care, and traditional insurance. So hopefully, you know, what's the... Yeah, of our demographics, I would say definitely immigration status and race were the most important ones specifically for our study, mainly because our study really looks at how shame works in different types of cultures and how we can provide the most culturally competent specific care for people who are either coming to this country who have a different concept of medicine or seeking psychiatric care or really how we can understand them and target their care before it turns into a substance use disorder. The immigrant status, were there any particular aspects of immigrant status that stood out? Well it was either, we asked if they were born in the U.S. or not born in the U.S. and that was our differentiating factor, that self-reported.

substance use treatment

opioid use disorder

cannabis use

medications availability

psychiatric hospitals

stigma

minority communities

State Opioid Response

treatment outcomes

cultural competence

treatment disparities