Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Celebrating 75 Years! Learn More >>

Director’s Innovation Speaker Series: The Opioid Crisis: Disrupting the Status Quo with the HEALing Communities Study


JOSHUA A. GORDON: Hello and welcome to the National Institute of Mental Health Directors Innovation Speakers Series. I am Dr. Joshua Gordon, director of the National Institute of Mental Health. It is my pleasure to welcome you all here today to this webinar. I am really pleased to have Dr. Sharon Walsh visiting us from the University of Kentucky to talk to us about the opioid crisis. I will introduce Dr. Walsh in just a minute. But before I do, I just want to go over some of the ground rules.

If you require technical assistance, having any trouble hearing or seeing, please use the Q&A box to communicate with the event production staff and someone will reach out to you to try to help.

We are very much looking forward to Dr. Walsh’s talk and we are going to have time for questions and answers at the end. At any time during the webinar, you can type your question into the Q&A box and I will go through those questions and hopefully get Dr. Walsh to answer them at the end. But you do not have to wait until the end. You can put them in at any time.

As always, this is being recorded. The recording will be made available in the coming weeks on our website as indicated here. If you have friends, colleagues, countrymen, whoever that you would like to have see this and they did not have the chance, please let them know and it will be available at a future date.

With those housekeeping notes in hand, I will now turn our attention to today’s speaker, again, Dr. Sharon Walsh. Sharon is a professor of Behavioral Science, Psychiatry, Pharmacology, and Pharmaceutical Sciences in the Colleges of Medicine and Pharmacy at the University of Kentucky and serves as director of the Center on Drug and Alcohol Research and the Substance Use Disorder Priority Research Area there.

Her clinical research focuses on pharmacological and behavioral issues and opioid use disorder from studies on the efficacy of pharmacotherapies to the abuse liability to the actual pharmacology of the widely used opioid drugs, including analgesics. It is a remarkable breadth of research that really spans everything from molecular pharmacology all the way through to the principles it takes to figure out how to get therapies in the hands of the individuals that need them.

Dr. Walsh has published more than 150 manuscripts and her work has been recognized with numerous awards, including the Betty Ford Award, the Marian Fishman Award, and the Presidential Early Career Award for Scientists and Engineers. Her funding has come from the National Institute of Health really throughout her career as well as other organizations. She regularly lectures nationally and internationally on opioid misuse, opioid use disorder and its treatment and has served as an advisor to National Institute on Drug Abuse, the Veterans Administration, NIH generally, and a number of other organizations, including the FDA.

She currently serves on the Scientific Advisory Council for NIDA. She is presently the principal investigator of the HEALing Communities Study at the University of Kentucky that is supported by NIDA and the Substance Abuse and Mental Health Services Administration.

Dr. Walsh, thank you so much for joining us today. Very much looking forward to your talk and I will be back at the end to ask questions.

SHARON WALSH: Thank you so much, Dr. Gordon, for that lovely introduction and also thank you for the invitation and the help from your team. It is a real pleasure for me to be here today and to have an opportunity to share with this audience about the HEALing Communities Study and I am doing this really on behalf of the whole consortium because this is the type of study that takes a village. There are many of us that have been working very hard on this over the last several years.

The title of my talk is the Opioid Crisis: Disrupting the Status Quo with the HEALing Communities Study. These are my disclosures for the past two years. You can see that I have served as the scientific advisor for a number of pharmaceutical companies, all really aiming at developing treatments for those suffering with substance use disorders. None of those have really anything to do with this particular project that I am going to talk about today though.

The outline for my talk is I am going to provide a very brief current state of the opioid crisis just to make sure that everyone is on the same page. I think it is hard to miss because it is in the news every day.

I am going to talk about the HEALing Communities Study and why it was needed and how it is designed. And then within the design, there are three major elements that make this a very unique project and those are community engagement, strategies to expand, evidence-based practices to prevent opioid overdose deaths and public health communications campaigns. When I share that and share the experience that we have had across the states that are participating, I will wrap up and then there will be time for questions.

This is where we are in 2021. And what you can see is that when we talk about the opioid epidemic, we talk about three different phases of the epidemic with the original one really starting in the 1990s with the over-prescribing and elicit market for pharmaceutical opioids. That went on for a really long time. And then many stakeholders, government agencies, practice agencies all decided that they really needed to change what was happening. CDC put out new guidelines. All the states were adopting prescription drug monitoring programs. We really saw a clamping down of the commonly prescribed opioids as the primary problem sometime around 2010. The commonly-prescribed opioids are in what to me is the light turquoise. Then you see the darker blue starting to rise and that is heroin.

And what happened around the time in 2010 is that heroin started to spread across the country through the cartel. And in places where people could no longer get prescribed opioids but had developed an opioid use disorder, this was a really easy and inexpensive substitute. What we saw was places in this country suddenly had a heroin market that really previously no heroin could be found. That was the case for Lexington, Kentucky when I came here.

And then about 2013, all of a sudden, the cartel started doing something very different. And what they started doing was synthesizing fentanyl and fentanyl analogues, which had now just taken off precipitously. You can see that this slide is showing the deaths per 100,000 for the overall population of the US and any opioids is about 25. But you can see that that line is really tracking closely with the synthetic opioids, which is largely fentanyl.

Why was the HEALing Communities Study needed? The goal of this study is to reduce opioid overdose deaths. That is a really great goal to have. But why would you need such a big implementation science project in order to do that? Well, what we saw from that last figure is that the opioid overdose crisis just continues to worsen despite substantial efforts to mitigate it. There were 100,000 lives lost to drug overdose in the last year for which there is a 12-month rolling count and about 70 percent of those are related to opioids with fentanyl and fentanyl analogues accounting for the majority.

We have efficacious interventions to address opioid use disorder. We have had some of them for many decades. But they have inadequate uptake and inadequate delivery to the people that need them due to structural barriers, disjointed care systems, stigma, and discrimination, which play a huge role in preventing people from getting access to care.

Really, the vision for this project was that large-scale community activation was needed in order to expand evidence-based practices and reach the individuals who need them. Along the way, what I like to say is we were trying to change hearts and minds.

This is from a recent publication by Rahul Gupta, who is the ONDCP director. Other people have done something similar to this, looking at that continuum of care, how many people may be affected by opioid use disorder versus those who get treatment. And in this rendition, what you can see is they have a starting number of 276 million people and then of those, close to eight thousand had opioid use disorder. And then it goes on, who meets criteria for the disorder, who received treatment in a specialty facility. And then of those, 288,000 people receive medications for opioid use disorder. These are medications that are approved by the FDA and are highly effective at treating opioid use disorder and helping people go into remission and recovery.

Methadone and buprenorphine are also highly effective at saving lives. This is the most remarkable thing really about these medications. This figure illustrates the death rate for the general population in green and then in red, those people who have opioid use disorder and do not receive any treatment. When you look at the blue bar, what you can see is that those who have received methadone or buprenorphine or what is sometimes called medication-assisted treatment in communities. This death is significantly reduced almost so that it is close to the general population.

What we know though is that less than 15 percent and it is probably much smaller than that because it is very difficult to actually count the number of people that have opioid use disorder. A lot of people – we have hidden populations. We have a lot of people who are incarcerated, who, for example, are not counted in the National Household Survey on Drug Use and Health, which is one of the biggest surveys that we have to estimate prevalence rates. No more than 15 percent. That is probably closer to something like 5 or 8 percent of the people who could benefit and by benefit, I mean go on to live, are actually able to receive it despite the fact that we see somewhere between a 50 and 82 percent reduction in death for those who get medication when they are compared to those who do not.

For me, this is remarkable. How could we have a medication that could cut the risk of death by half or two-thirds and yet we are not able to actually get it to people who need it in this country where we have this enormous health care system?

The HEALing Communities Study was envisioned by leadership at NIH and NIDA. There was a call for grants to be written. For those who write grants, it was a very long call. You had to read it many times to make sure that you got everything in there that was needed. And ultimately after that standard grant competition, four states were funded and they were Kentucky, Massachusetts, New York, and Ohio. From those four states, we are going to work with sixty-seven communities.

The objective of this study is to test the immediate impact of implementing an integrated set of evidence-based interventions across health care, behavioral health, justice, and other community-based settings to prevent and treat opioid misuse and opioid use disorder within highly affected communities. We had some criteria for what actually was a highly affected community.

The ambitious, lofty goal was to reduce opioid overdose deaths by 40 percent. I will say that going into this from my perspective as one of the PIs, my goal was to see if we could have a significant reduction so that we could then parse out what does and does not work about the intervention and make sure that we are able to adapt that and disseminate it to places so that we can start saving lives.

These are the four states. There were sixty-seven communities that were randomized to the intervention. Thirty are rural. Thirty-seven are urban. And the total population that is covered by these communities is about ten and a half million people.

These are the baseline data for the four states, looking at the opioid overdose death rate trends in what would be our baseline years before we launch the intervention. I should have pointed out on that slide at the very beginning where I talked about the three stages, the place where we are now where the numbers are so very high, that rate for the US is about 25 deaths per 100,000 persons. Those are age-adjusted numbers. At baseline, you can see that our death rate in our communities was much higher even then what it is now, which has grown substantially since 2019. You can see that by comparing the different sites, Massachusetts had the highest death rate in their communities. Kentucky and Ohio were quite similar. New York was closer to what is now the national average. And you can also see that they actually were already seeing a downturn in 2019 where there was a significant difference in the relative rate of death between 2018 and 2019 there.

Looking at the baseline data in a different way by examining race and ethnicity, we are able to see something that we have not historically seen before in these decades of this overdose crisis. And what that is is that for non-Hispanic blacks within the study, we see a significant increase from 2018 to 2019 whereby the rate of death for non-Hispanic blacks and males are more affected by females has just increased precipitously. And for the first time, they have a higher death rate than whites. This is really the first time that this has happened since the 1990s. This was really important for us to understand when we are thinking about how do we design an intervention, how do we make sure it gets to the people that really need it.

How do you design a study like this? I will say really quickly, we all wrote our own grants and had to compete based on what our teams put together. And then after they decided to fund the four states, they said okay. We are going to write one protocol, and everyone is going to use that protocol. We are going to try to take the best from all the grants and that is exactly what happened. We had a lot of commonalities, obviously evidence-based practices, things like that. But there were unique things that the different sites brought to it and we were able to make this a really rich design.

It is a multi-site parallel arm cluster, randomized, weightless controlled trial. It is an RCT that evaluates the impact of what we call the Communities that Heal intervention compared to usual care in the weightless communities. What all that means is that nearly half of the communities got randomized to start the intervention first and we call them Wave 1. And the other ones waited as kind of the usual care and then after Wave 1 was finished, we have now randomized Wave 2 so that they also will receive the intervention. The Wave 1 communities implemented the intervention for 30 months, during which time the Wave 2 was waiting. And in month thirty-one, Wave 2 communities began to implement.

This is the overall timeline. And you can see in 2019 where it says startup, that is where we were busily trying to put together a single protocol that would work across all the communities and across the states.

Then we began the intervention in 2020 with the Wave 1 communities. There were thirty-four of those. And then the Wave 2 were in usual care. We did have to revise the design twice because this happened to launch just before the COVID-19 pandemic hit. That caused all kinds of other problems that I would be happy to talk about that later, maybe in the question-and-answer period.

Once the intervention is turned on, you can see that in July of 2021, that is when the comparison period starts. That is where we are collecting all the data to see what the differences between those communities on the intervention versus those that are not. After a community finished the intervention, then we work on sustainment. Right now, this is where we are so that Wave 2 is right in the middle of the intervention, and we are busily working on analyses.

The study hypotheses are fairly straightforward. What we hypothesized was that compared to Wave 2 communities, Wave 2 communities will reduce opioid overdose deaths, will increase naloxone distribution because that is an effective reversal agent to bring someone back to life when they have overdosed, that we would expand utilization medication for opioid dose disorder, but the specific focus on methadone and buprenorphine because that is where the strongest evidence lies, and that we would reduce high-risk opioid prescribing. Even though a lot of progress has been made on opioid prescribing, there still are some problem areas that need attention. That was another area for us to focus on.

I am not able to share the topline results with you today. I am sad about that. But it is because of the data lag and how the data sources come from different data partners. This shows that the naloxone distribution data actually have been analyzed and they are embargoed, and we expect to submit that paper probably next month. The utilization medication will be available in March so next month. And then the high-risk prescribing is underway right now. And then the overdose deaths actually have the longest lag, and they will be available by May and that is our primary hypothesis of course.

How do you actually collect all this data and monitor all these things? It requires a tremendous amount of effort and cooperation. Each of the states had to have a key government official at their state level to help them navigate this. We have 52 administrative data sets that include Medicaid, death records, DEA records, things from law enforcement, so many different sources across four states. And these all require data use agreements. And they all work a little bit differently as well. That is an incredible challenge but also an incredibly rich resource.

We also have some de novo data collection that is really aimed at trying to help us understand some of our implementation science aims. That includes surveys of people in our communities who are kind of key opinion leaders and have the ability to make change.

We also have 1300 qualitative interviews of community stakeholders, which if you are a qualitative researcher, you know that that is a very large number and that is going to be a lot of work.

This is an example of some of the data sources that we are looking at. I am using the Kentucky data just as an example here. If you wanted to look at overdoses, how would you go about doing that? It was remarkable to all of us I think during COVID, how all of a sudden there were COVID dashboards everywhere and you could know exactly how many people were in the hospital in your state or how many people – pediatric patients or – we do not really have that for overdose even though it is such an enormous crisis.

Here are just four different data sources as an example, and the one on the bottom is the overdose deaths. As I said before, in Kentucky, they have a four-month lag between they are completely adjudicated. It is even longer in those states.

For the emergency department in red, you can see that those data – those are from billing. That comes 90 days post the close of the quarter. We have syndromic surveillance, which is kind of a more live system and that is very quick, but it is also an opt-in system so not every hospital has to participate. And then we have emergency medical services, which in Kentucky, it is mandatory that they participate, and the upload is very quick.

But what you can see is that the general trend despite the fact that the absolute numbers are different because they are probably capturing different people, but the general trends are the same. But it is important to contextualize when you are trying to interpret your results.

Leveraging that data because we were working on them and we had them on dashboards and we were using them to help our communities choose what they should be doing in their local areas, we were monitoring things. This is a great example, I think, of what we would like to think public health should do. We were monitoring the overdose runs for emergency medical services. This is just the ambulances, picking people up in the top figure. And in the bottom, going to pick people up who say there is no way I am going to the hospital, so they are refusing. That is a group that you do not usually get to measure.

But what we were able to do was to use an interrupted time series analysis and find that there was really a very significant jump in overdoses that happened right when the state of emergency for COVID was declared. And of course, what else was happening was the jails were letting everybody out. The hospitals were not letting people in if it was not crucial. We had challenges with ongoing treatment for people that were in treatment. We responded. And the way that we responded was that we decided to fast track distribution of naloxone a little earlier than we had expected in Wave 1 with a particular focus on the jails because all those people were going to be let out without a treatment plan and they are very high risk of overdose. And then we shared the paper. We published it. It was one of the first papers to show a significant rise in overdose after the public health declaration.

The actual intervention has three components, as I mentioned. I am going to touch on each of those. I am going to try to give some examples. At the same time, I am going to move quickly. I am going to pull examples from the different states that are participating. There is the community engagement. And really, what that focuses on is each of the communities had to have a coalition of community members that were dedicated and willing to work on this project. We met the most amazing people through this process. There were so many people that are so willing to give their time and they spent so much time on this study. It was really amazing.

The second component is the Opioid Overdose Reduction Continuum of Care or ORCCA. We like acronyms. This is just the evidence-based practices. What do we know works so that we know what we are going to deploy.

And then the third thing was a very sophisticated, I think – this is not my area of science but public health communications campaign that would tie into the interventions that were being implemented really to reduce stigma around those interventions and to drive demand for those.

These are the phases of the community engagement. There are seven phases that each community coalition had to go through with the study team. And you can see that there – from the titles of them, preparation, getting started, getting organized, community profiles and data dashboards, community action planning. Essentially really like the beginning phases are trying to get them organized, provide them with a landscape analysis, identify leaders and champions from their communities that can really affect change, start to talk about who could be champions for data or communications, things like that. If you have somebody that works for the local newspaper that is on the coalition, that is a great person to have as your communications champion because they all have all kinds of knowledge that we would not have access to.

And then the real meat of it gets to the point where the coalition now really understands their data. They are looking at their dashboards and seeing what is happening in their community on all these different measures that I just described. And then they can start to plan what it is that they want to deploy. That of course is around the evidence base practices. This is a lot of work. I just really have so much respect and gratitude for everybody in our communities who did this. We are just finishing this with our Wave 2 communities too and they are just as enthusiastic.

But there were all kinds of tools. There was a lot of education. There was a lot of back and forth. There was a lot of trying to come to mutual understandings about what is evidence-based care in the field of substance use disorder. There are all kinds of “care” and I am using air quotes if you are not watching that is not actually evidence base. But there are all kinds of things that are happening in communities that people can find themselves in a program receiving that there really is not any evidence to suggest that it is going to be effective for them. We needed to do some education around that as well.

And then this is just an example of a screenshot from a data dashboard. You have these people that live in a community. They have access to all these measures. I said that we had like fifty some measures. They could pull them up. They could play with them. They could figure out what the trends were. We could do geospatial mapping and they could look and see where things were happening in specific communities within their county if thy were a county. It was pretty amazing actually.

It really was very effective in helping to have data-driven decision making because people would have something maybe that they wanted to do, but when they took a look at the data, they could say maybe that is not really how we need to spend our resources. Maybe it would be spent better elsewhere because we are not doing so much in this other area.

Then the coalitions get to the action planning where they are actually choosing what kind of strategies, they want the study to help to implement in partnership with the community and community agencies. There are three what we call buckets of evidence-based practices. Opioid overdose prevention education and naloxone distribution. So, this is the overdose reversal medication along with education on how to recognize that someone has overdosed. Effective delivery of medication for opioid use disorder, including agonist/partial agonist medication. That is methadone and buprenorphine and including outreach and delivery to high-risk populations. Safer opioid prescribing and dispensing was the third one. And then within each of these three categories, there were many, many, many different strategies that one could use to make change in the community.

This slide illustrates a paper. I think I just saw that it actually came out today. That just characterizes what our communities in Wave 1 chose to do. They all had a requirement. They had to choose – in each community, they had to choose one thing from each menu and the menu to the medications for opioid use disorder has three sub-menus. That is five. But what you can see here is what they chose for 453 additional things that they wanted to do. They were very, very ambitious. We were worried. Are we going to really be able to do all these things? It has been amazing. With a lot of help and a lot of cooperation and coordination, it is amazing how much has actually gotten done. You can see how those things break down for naloxone distribution in purple, the medications in red and then Safer Prescribing is in orange.

These strategies that were chosen were implemented in partnership with 278 different health care agencies, 231 behavioral health agencies, 163 criminal justice agencies, and many others. We work with service programs and public health departments, providers of all sorts, neighborhood organizations, recovery centers.

I cannot spend much time on this. It is easy to say we are going to pick the strategy and then we are going to make something happen. How do you actually make that happen?

Fortunately, there is this whole other field of science that I have really grown a great appreciation for called implementation science. That whole area of science is about how do you actually affect change. How do you say this is what we want to do? How do we get an organization to actually do it?

We modified for the HEALing Communities Study the RE-AIM and PRISM framework that looks at – for RE-AIM, it is reach, effectiveness, adoption, implementation, and maintenance. And then for PRISM, you are really looking at contextual factors that will impact your success or failure. External context, for instance, could be policies – policy workgroup. It could be guidelines. It could be regulations. The internal context is like who are the people that are working within the organization that you are trying to change, and do they want to champion this, or do they have stigmatizing beliefs themselves that are going to interfere with this. It has been very interesting, I have to say, having the opportunity to work on an implementation science project like this.

Altogether for naloxone distribution and education, there were 317 strategies. This just shows that there are all different ways that you can do it. You can have people do it passively. You can have people do it actively. You can provide naloxone in hotspots. You can do outreach in communities. You can go to places where you know high-risk individuals are going to be and certainly focusing on criminal justice and some of the other organizations where we know it is a good touch point for people who are at risk.

I am just going to share a couple of examples. This one is out of New York. New York used geo-targeting. They were leveraging their data to assess where they could place NaloxBoxes in order to save lives. A NaloxBox contains six to eight doses of naloxone. It is on the wall, kind of like a defibrillator. It provides 24/7 access and has instructions. In New York in Wave 1, they placed more than six hundred naloxone boxes and as I said, they used data-driven decisions to find their locations, including OD maps and EMS runs and 911 overdose calls. And the locations were places like commuter train stations, motels, restaurants. You can see the breakdown from their eight Wave 1 communities there.

In Kentucky, we did something slightly different and part of it was that when we wrote our original grant just like the other teams, we had a team, and we were going to keep that team together even with the protocol being changed and we were very fortunate because we had a strong pharmacy team. We had a number of pharmacists that were working with us that knew how to do academic detailing, and really were already doing a lot of interventions with pharmacists. We set up a hub with many spokes models so that we were actually facilitating partner agencies that were serving high-risk populations. We worked with about 170 agencies. We were able to get them to sign a standing order agreement, which allows the distribution of naloxone under the law across health care, behavioral health, and criminal justice. And then we also hired team members called prevention specialists in each of our communities who worked to directly provide OEND to community members of public events. Here is a picture of one of them.

During the Wave 1, we distributed and collected reach data on individuals who received over 41,000 naloxone units with partner agencies and our prevention specialists distributed another 11,000. You can see in the figure how it is broken down by agency type. And you can also see that we were very committed to equitable reach. These are our reach data. You can see that it is about half and half female and male. Our racial and ethnic minority numbers may look pretty low, but we have low representation in Kentucky overall like less than 10 percent. We certainly can always do better. And we had about a third of the people who were under thirty-five of age who were at high risk.

These are examples of some of our distribution campaigns. In the upper righthand corner, you see somebody – he is pointing to his own billboard. We use local people that were known. We had police chiefs and mayors and all kinds of people. In the upper left, you can see an example of one that is in another language. This is in Cape Verdean Creole. That is out of Massachusetts. We had a lot of Spanish materials. We had materials in all kinds of different other languages as well.

You can see partnerships with Kroger where we were able to get the communications materials onto their prescription bags and we did a lot of social media.

For MOUD, medications for opioid use disorder, we had 309 unique settings. And the three categories where we wanted to expand access to MOUD – this could be like some place that is not doing it but is willing to or some place that is only seeing 30 people but they could see 60 people.

We wanted to improve linkage to medications for opioid use disorders so people who are not in care, people that are in criminal justice settings because there is not a lot of care being given in jails and in prisons at this point or specialty courts, probation, parole, alternative sentencing workers that we could actively link them to start their treatment of MOUD.

And then the final one was to increase retention in treatment by addressing social determinants of health that prevent people from being able to stay in treatment.

This is a really clever example, I think, from Gloucester, Massachusetts where they developed a provider on the pier, and they were doing direct care to the fishing community in Gloucester. They had a PA from a local hospital that would come with a navigator and supplies. They opened a new clinic in the harbor master’s office. It was really to reach a high-risk population that was transient and largely immigrant dock workers that were affected by opioid use disorder. You can see how the collaboration worked between the HEALing Communities Study and then all these local organizations to actually make this new initiative happen.

This is another example from our Ohio partners in Guernsey County. I said just a minute ago that there are lots of people that are in the criminal justice system that have substance use disorders, but they are not getting treatment for it. There are lots of people who are in the criminal justice system just because they have a substance use disorder and maybe they really should just be getting treatment instead of being put into jail. But people who are waiting trial or who are incarcerated are very unlikely to get care for their substance use disorder.

In Guernsey County, they worked with Cambridge Municipal Court. And individuals who were held in jail waiting trial were released under supervision with their MOUD assessment and compliance as terms of a bond order. And that allowed them to begin their treatment earlier. And then they also moved people to house arrest two months prior if they were inmates, so that they could initiate medication before they were discharged, and they also trained officers specifically to understand how medications to opioid use disorder work.

This is an example of the social determinants of health. One of the things when you talk to anybody about what is the problem with staying in care. Transportation. Everybody says it is transportation. In Kentucky, one of the things that we have are a lot of rural areas. They might not have a lot of treatment providers.

In the opiate space, we have something called an OTP or an opiate treatment program. And that really means a methadone clinic. This is a federally licensed program. It has a lot of rules, and people have to come every day to get their dose of medication. If they do not, then they will get reset back to something that is more stringent. For people that have a problem with transportation, they are always struggling to just maintain because they can never earn take-home medications because of perfect attendance or things like that.

In Madison County, we had an OTP. That county has only 92,000 people but it is 443 square miles. Our partnership with a recovery organization called Voices of Hope provided rides. The majority of them were to opiate treatment programs but they did also to residential facilities and other things. That was equivalent in Wave 1 to five trips around the world just back and forth to Madison County.

In New York, they were focused on doing housing vouchers with care packages to help with MOUD linkage and retention. They provided over one thousand nights of shelter to over 100 unique individuals. Of those, 91 successfully initiated medication, which is pretty amazing. They were retained on their medication with housing vouchers in case management. Again, this is like a cooperative thing where the HEALing Communities Study team was coordinating with BJA, which is the Bureau of Justice Administration, NACCHO, which is the public health departments. Incredible collaboration.

And then here are just some more campaign materials that is related to MOUD. You can see that we even got Starbucks to agree to participate with us and lots of local people and QR codes so you could find out where you could get treatment on demand.

For the prescription opioid safety, there were 92 unique strategies that were chosen. You can see that they were across all kinds of prescribers because that is where these are coming from and then also dispensing programs like safer opioid dispensing.

These are a couple of examples from our academic detailing team of pharmacists in Kentucky. They developed all these training materials to reach dentists, primary care providers, and pharmacists with key messages about prescription opioid safety. They did one-on-one education sessions. They delivered practice guides. And they had more than five hundred visits to pharmacies to discuss opioid safety, naloxone rules, syringe sales, all the things that really impact this population.

They also worked to prioritize pharmacies for safe disposal units. There was a real dearth of safe disposal units in some of our communities. The HEALing Communities Study paid for it but the pharmacies had to agree that they would sustain them and over a ton of medications have been incinerated so far.

And then these are some examples. The one on the left is a fun one because it was something that ran as a video on a closed loop in a restaurant. It had good leftovers and bad leftovers. Of course, the messaging is please make sure that you get rid of your unused drugs.

I am going to wrap up here. I am just going to share a few more things. I have just plotted overdose deaths over the time that the study has been underway. You can see that first purple line. That is where Wave 1 planning begins. The second bar is when the implementation begins. And then the third one is when the intervention ends. You can see that for us to make a difference in these communities with the intervention, we are battling against a changing background that is largely attributable to the widespread availability of fentanyl leading to the highest rates of overdoses that we have ever seen.

Where are we now? Well, the primary outcomes and secondary outcomes will be available in the very near future. Dr. Gordon – I worked with this team, trying to get my talk switched a little bit later just so that maybe I would be able to share those but we were not able to do that. You will have to stay tuned. I am sure that there will be plenty release notices about the findings.

We are currently working with our Wave 2 communities who waited so patiently for so long for us to get to them. We are doing a lot of sustainability planning for Wave 1 activities. The consortium has published more than 40 manuscripts. We have 130 plus manuscripts in progress. We are really writing at a frantic pace right now. And we are doing some very serious dissemination planning with our federal and state departments.

With that, I want to leave us with a note of hope, that despite the fact that we are seeing these outrageous rates of death and a loss of young people, there are a lot of changes that are happening. Most recently, the federal mat legislation is going to eliminate the specific prescriber requirement for the DEA-X license that was needed if you were going to provide buprenorphine in an office-based setting. That hopefully will lead to expanding the number of people who are willing and able to prescribe.

We had an extension during the pandemic for the first time of telehealth waivers or of telehealth rules that allowed people to have telehealth visits and start treatment through telehealth. That has been extended.

The opioid abatement funds are being directed to all the states now. They hopefully are going to be used for expanding evidence-based services and hopefully the HEALing Communities Study can help to inform the states on what that actually means to have evidence-based services.

Tomorrow the FDA is going to discuss improving naloxone for over-the-counter use, which is exciting because it has only been available as a prescription drug, which has limited to some extent the way that you can access it, which just creates other barriers. Hopefully, that will expand access.

I want to acknowledge all the consortium team members. This talk is really about all of us. There are 500 of us, I think. We are all close friends and family by now because we have spent so much time together on the phone and on Zoom. Our key government state partners, community advisory boards, community coalitions, and our community members, and of course, NIDA and SAMHSA for all their support of the project. And then that is just the standard disclosure. And then I will close and open it to questions by dedicating this to the memory of Dr. Rebecca Jackson, who was the PI of Ohio and sadly she passed away in October and we miss her.

JOSHUA A. GORDON: Thank you very much, Dr. Walsh. Really a wonderful talk about the scope of the program. It is just mindboggling. Hundreds of scientists, four states, sixty some odd communities. It is really remarkable for us to hear about this project from you and to have the opportunity to ask you some questions and follow up in exchange to your talk. We are going to get to that.

Let me just remind everyone. I see a few more questions coming in. That is great. You can keep your questions coming into the Q&A box as we bring them to Dr. Walsh.

The first one. I am going to take the chair’s prerogative and ask you myself. You talked about a lot of the successes for this large program and some of the challenges. In particular though, I wonder if you might talk to us about what were the challenges within. You talked a lot about the challenges of harmonizing across the different states and different approaches. What were some of the challenges that you saw within Kentucky in terms of – whether you saw them or not in terms of bringing the interventions into action?

SHARON WALSH: That is a great question. I will say that I think some of the challenges that we saw in Kentucky are similar to ones that we are seeing in the other states even though every place is unique. I would say that for us, we had great support from our State Department of Corrections, and they were very keen to work with us. We were able to make some important headway there. But it was incredibly challenging to actually try to get any meaningful medication programs going in the jails. The coalitions wanted them and sometimes the jailers wanted them. They were enthusiastic.

But it is a peculiar situation because the jails were run locally. The jailer is elected. The jails often contract with correctional medical providers. Then you have another player involved. Everybody in the community could say we want to be able to provide buprenorphine. We are going to screen people. We want to treat them. But then if that contract provider says we are not doing that and it is not in our contract, you have to wait until they can get rid of that contract and hire somebody new. I would say that that was one of the areas that was probably most frustrating for us where we spent a lot of energy but did not feel like we got a lot of paybacks for it.

JOSHUA A. GORDON: Thank you very much for that. I am going to turn to some of the questions that the audience has been putting in. One begins with first a thank you for your wonderful talk and then asked if you might expand more about your approach to the dissemination of results at the end of the study and ultimately how those approaches can be expanded to other communities across the country.

SHARON WALSH: Thank you for that question. We are working really closely with our partners at SAMHSA and NIDA of course. But SAMHSA – they are really kind of – they have a long history of dissemination and technical centers. They really know how to do dissemination science. We started by collecting everything that was developed during the study and kind of doing an inventory to see what we thought would be useful and adaptable to distribute. We are really now in kind of the process of coming up with a more formal dissemination plan that will leverage various organizations that already are in that business but then also – I see that somebody had said are abatement funds your term for settlement funds. Yes. That is what I am talking about. It is the same thing.

Those abatement funds or settlement funds are supposed to be used for things that will help people that have been impacted. I think that we will be able to disseminate to the different states and to the local jurisdictions that are going to get money as well. Every state is kind of doing that process differently. But we are hoping to have as much reach as we can.

JOSHUA A. GORDON: Great. Thank you. One of the questions – I am going to expand upon it. The question itself was will you be able to compare efficacy of different types of programs like buprenorphine versus OTP versus other treatments or with or without psychosocial services by comparing across the different sites.

I would add to those treatment approaches also some of the community-based approaches that you mentioned like – it seems like there might have been differences in different states in terms of whether you are helping pharmacies get rid of old prescriptions or doing different kinds of campaigns. Is there an attempt at all to compare across sites to learn from differences that vary from site to site?

SHARON WALSH: We are very interested in that. Honestly, one of the things that we would love to do is use Wave 2 as the example for that where we have a leaner, streamlined intervention because we have learned so much and we are moving faster with communities because they are at a different place as well. We would love to be able to do those kinds of comparisons. Anything that we do though is going to have to be in the context of a much larger intervention. It is not an RCT that is going to isolate whether or not adding behavioral treatment to medication augments that. And we have a lot of those data that are published already. But I do think that we are going to be able to identify things that are community engaged and are happening at the community level that are more effective.

JOSHUA A. GORDON: Got it. Excellent. This question perhaps is more a clarification because I think you might have explained it but maybe some people did not pick up on it. How do you account for the potential differences in outcomes before and after the disruptions from the COVID pandemic?

SHARON WALSH: It is a great question. It is a randomized controlled trial. That is the beauty of having a randomized controlled trial is that we have a comparison group and at baseline, those communities that were randomized, Wave 1 versus Wave 2, were done so in a thoughtful manner in order to make certain that we were controlling for factors that we felt would impact the outcome like their baseline overdose rate as an example, rural versus urban. The pandemic happened to everybody whether you were getting the intervention or not.

Now, will the pandemic change the way that the results look? Yes. Do I think that it would impede about ability to actually compare between Wave 1 and Wave 2? I do not because everybody was being affected at the same time, but we were only active with one set of the communities.

JOSHUA A. GORDON: Another question has to do with the way that this might be able to help address issues in equity, which you so compellingly raised by recognizing that more recently rising faster were overdose deaths particularly in black Americans. Thank you for this talk. How will you be addressing any sex differences or differences by racial or ethnic groups or sexual and gender minorities? How do you plan to report out equity in terms of access retention or survival? Maybe you can answer some of those questions.

SHARON WALSH: That was a quick list of several questions. In Wave 1, we were already thinking about the need to ensure that the intervention that we were delivering was equitable and that we were reaching people that were hard to reach.

In Wave 2, were had the opportunity to be even more thoughtful about that. We are using equity planning tools and we are putting that in the hands of the coalition. Before we even roll anything out as far as evidence-based practices, there needs to be a lot of thoughtful discussion around how do you ensure equity when you are choosing your venue or a particular agency, who are they reaching, who is not in the mix. I think that we are being very proactive about it in a way that is more engaging with the coalition too to really get them thinking about it compared to in Wave 1 where we were really kind of doing hotspot analysis and trying to figure out where services were missing and things like that. There were more questions in that list I know.

JOSHUA A. GORDON: I guess one thing to add to that would be any issues around sexual or gender minorities and whether that is going to be addressable in the design.

SHARON WALSH: Yes. But this goes to the issue of when you were working with all these different administrative data sources, you cannot make assumptions about what you will have access to. As an example, a lot of our key measures are Medicaid measures. But the rate of completion for racial and ethnic information in Medicaid records is fairly low compared to other types of records. When we started exploring that, it was very disappointing to us because the rates were different across the states, but they were not – like when you think about how much data you are going to be missing, they were not optimal. Then there are other types of records, death records, where we have everything about the individual.

I think to the extent that we can examine those key differences for the data sources that actually include them in a way that will allow us to do that, we are very interested in that.

JOSHUA A. GORDON: I know that HEAL has other programs specifically targeted in research in American Indians and Alaska Natives, but I wanted to ask the question from one of our attendees. Do you have any information in these studies that you are conducting in the four states on American Indian or Alaska Native participants?

SHARON WALSH: I think that New York was the only participating state that had communities that were Native American, and they were included in Wave 1. I actually do not know about the Wave 2 communities, but I do not remember - we do not in Kentucky and I do not remember Ohio saying that either.

JOSHUA A. GORDON: Thank you for answering that. Thank you. I am going to ask you one more question. Before I do that, I am just going to close things out. I just want to thank you. There were many comments in the Q&A about how wonderful the presentation was. It is really just fantastic for us to hear about this very large-scale implementation effort.

As we talked about in our meeting before this, implementation science has always been a very big and important component of NIMH’s portfolio both for mental health and for our behavioral AIDS work. It is wonderful to see this program moving forward. We are very excited for the results.

And the last question sort of presages the possibility of those results and one of our attendees asked, do you foresee or how do you foresee the results of this study leading to potential widespread policy changes, particularly with regard to how methadone is regulated in this country?

SHARON WALSH: Methadone. I will say that from my perspective, this study already has led to policy changes in the states where it is being done because we have been doing policy scans as part of our work and saying – or we hear from pharmacists, or we hear from jailers, or we hear from EMS, that they do not believe that they are allowed to carry naloxone. And then we ask them, why do you think that? And then we go back and say maybe it is because they are just saying that because it is really stigma and they do not want to do it or maybe the policy is poorly written, and it leaves a little bit of room.

An example in Kentucky, is that fentanyl test strips, which are really important, are considered paraphernalia, the way that our paraphernalia law is written. The public health department could give them out but the person who could then immediately get arrested for it. That policy is getting changed. The wording is getting changed.

We have a long way to go with methadone. It is far too rigid. We have not made any – during the pandemic, we lightened things up a little bit, but some states chose not to do that, and they kept it as stringent as it had always been for the last 40 or 50 years. There is just so much work to do. It is such an effective drug. The delivery model is not effective.

JOSHUA A. GORDON: Thank you again, Dr. Walsh. Thank you again, attendees. I want to thank the ASL interpreters and the whole team from the Division of Extramural Activities that helped run this seminar series and look forward to seeing everyone in the next one. Remember this seminar was taped and will be available online in a few weeks. I apologize to those of you who were unable to get your questions answered. We could not get to all of them but we got to a good chunk of them so that is good. Bye bye for now.