Taking Action Summit: Information Messages to Reduce SUD Stigma and Increase Recovery Housing Support
Okay, good morning. My name's Kathy Taylor. Welcome to this session, it is, Information Messages to Reduce SUD Stigma and to Increase Recovery House Support. I'm here to welcome you and the participants that are here in person, as well as those who are on our virtual platform.
And, of course, we are here at the Eastman School of Music. We have two speakers here today, and I will introduce them to you. The first we have is Madison Ashworth. She is a research associate for Fletcher Group.
Madison is a research associate and a PhD candidate in economics at the University of Wyoming. She studies the intersection of health and behavioral economics, designing and testing of policies to determine their effectiveness in changing individual behavior and public health outcomes. Recently, her research has focused on understanding the impact of COVID-19 on individuals in recovery from substance use disorders, evaluating interventions to reduce stigma around recovery housing, and increase political and financial support for recovery housing, as well as evaluating smart recovery programs within recovery housing.
Other Covid-related studies have explored vaccine hesitancy and the effect of social distancing on public health. Joining her we have Robin Thompson. She is the director of research and evaluation with the Fletcher Group. Driven by her desire to enable equal access to care, treatment and healthy environments for all, Dr.
Thompson earned both her doctorate of public health in epidemiology and her master of public health in health behavior at the University of Kentucky College of Public Health. Her training in, I'm sorry, her training and research experiences have primarily focused on the use of mixed methods and epidemi, epidemi...
- Thank you. Surveillance to reduce health disparities. She currently serves as Fletcher Group's Director of Research and Evaluation, leading a team working on various research projects related to the expansion and/or improvement of recovery housing and other recovery support services for individuals with substance abuse disorder. So please welcome Madison and Robin.
(audience clapping) - Hi, thank you, Kathy, for the kind introduction. and thank you all for being here today. We're excited to have more than like two bodies in the room, so thanks for being here. And thanks for selecting our session as well.
We know there's lots of great sessions to attend at the conference, so we appreciate you selecting ours to listen to. And we've been really excited to be here at this conference this week, and have learned a lot from all the great presentations we've watched and conversations we've had. In our presentation, Messages to Reduce SUD Stigma and Increase Recovery Housing Support, we'll discuss the results of a recent randomized control trial we ran. This project involved our Fletcher Group team and researchers from the University of Wyoming Department of Economics.
We just wanna thank our amazing co-authors listed here, Dr. Linda Thunstrom, Ms. Grace Clancy, Mr. Dave Johnson and Governor Ernie Fletcher.
Also, a special thank you to Mr. Jade Hampton. He's someone in long-term recovery and is now a director of a major recovery Kentucky center. And he actually helped develop the main intervention that Maddie will share later in more detail that was tested in the study.
Erica Walker, who's one of our outreach and engagement specialists at Fletcher Group. Jennifer White, our director of public health initiatives. They both helped us with conceptualization and design of the study, so thank you all. And special shout-out to Kathy, who just introduced us, it's her birthday, so thanks.
Happy birthday. All right, so let's dive in. This slide is, everyone here in this room knows this, so I won't spend too much time on this slide. But we all know substance use disorder is a major public health issue and it's enormous impact on individuals with SUD, their families, communities, is why we're all convening this week at the Rochester Stigma Summit.
And per latest estimates, SAMHSA's National Survey on drug use and health, indicated that SUD impacts over 40.3 million Americans ages 12 and over, which is huge. And since the '90s, we've witnessed an increasing rate in the number of American lives lost due to this disease, and that has been talked about in a lot of the presentations this week. And in 2021, over 100,000 lives were lost due to the drug overdose, and that was an increase of 16% from the prior year.
And back in 2017 to 2018, we did see a decrease of the drug overdose deaths by 4%. But since then, the number has been increasing rapidly, especially, with COVID-19 pandemic and the isolation that caused that lack of SUD treatment access in recovery care, like the in-person mutual aid groups. As we all know, SUD is costly, in a recent 2021 study, Peterson et al. estimated that $13 billion are spent in hospital care costs alone for the care of substance use disorder.
And additionally, estimates indicate that the cost of an individual person with SUD ranges from 16 to 34,000 depending on the type of SUD. So, you know, why do we care about that? Because that's the cost of treating the symptoms. And if we were to reallocate those resources for prevention and tackling the root causes, like ACEs and some others, we could do a lot more with those resources, and we'd see, hopefully, a lot less mortality in years to come.
All right, so recovery housing, Maddie and I love this slide and talking about recovery housing 'cause we're here representing Fletcher Group, and that's what we live and breathe and love to talk about. And as most of you are aware, recovery housing is a housing model that relies on abstinence, peer support and community building to help support individuals in building recovery capital to initiate and sustain recovery from substance use disorder. And recovery ecosystems, which have been mentioned by many presenters this week, include the level of recovery support services available in a community.
And in terms of the SUD continuum of care, recovery housing is a very important recovery support service for individuals, as by accessing recovery housing, individuals have the opportunity to also access a variety of recovery support services offered in and outside the home in that larger recovery ecosystem in order to build their recovery capital over time. So the benefits, there are a lot of benefits of recovery housing, and the evidence for recovery housing has been growing. There are several researchers in the field that have contributed to the research base to examine the efficacy of this promising model.
There's still a lot of work that needs to be done, but from the studies that have been conducted, we do know that recovery housing helps to improve abstinence from use. It enables residents to acquire the skills to obtain employment and income. It improves their mental health and reduces recidivism. So that's a common thing that we see across a lot of studies.
And a major key, of course, and special sauce, which I really liked that peer support was termed in that way, and we'll take that with us after this conference, but a major key, of course, is peer support within recovery housing as that's a centerpiece of the recovery housing model. So beyond all the benefits, it's also very cost effective. And looking at the evidence from the Recovery Kentucky model, which we used in our study, found that in 2020, there were $5 million in tax savings to Kentucky residents all accruing from avoided costs to society. And, additionally, results from our secondary analysis we conducted from a Covid impact study in 2020 found that, on average, recovery houses operate with an annual budget of 250,000, which is substantially less than the traditional treatment model.
And, in a lot of cases, recovery housing can be equally or more effective. And you can tell that an economist, Maddie, had a heavy hand in the development of these slides. All right, so stigma, why we're here this week all convening, it's a major issue with establishing recovery housing. The stigma is often recognized as NIMBY in the context of recovery housing, and that also stands for Not in my Backyard.
And individuals that are not aware of what a recovery house does for individuals often feel that a recovery house and the individuals within may bring trouble to a neighborhood. They just don't understand that a recovery house supports the recovery ecosystem of their community and strengthens that community overall. SUD stigma has also shown to have detrimental effects to the residents within a recovery house in terms of accessing and completing their treatment. In pretty much every phase of an individual's recovery, stigma often exists, as we see and have heard a lot this week.
And studies have also examined the efficacy of numerous interventions at reducing stigma, both are related to severe mental illness and SUD. But we need to consider severe mental illness and SUD stigma separately due to the substantial difference in underlying beliefs. For example, those with a SUD are considered more blameworthy and dangerous. All this to say, is we need to be considering what types of policies and preventions will be targeting these beliefs and change misperceptions of those with SUD to reduce stigma and, consequently, enable access to treatment and care that they need, and enable them to be their best selves.
So what sort of interventions have been tested so far? So far, two major types of interventions are in the literature that have been tested to reduce stigma, we have contact interventions and we have educational interventions. Contact interventions can be described as any type of personal contact where an individual shares their lived experience, and that's been found to be very effective at reducing SUD stigma and changing policy preferences. And that's actually occurred this week, we've heard a lot of wonderful stories from a lot of really courageous individuals that have stepped into this field and are making a difference.
And educational interventions have had mixed results, but we do find that sharing information about a treatment effectiveness seems to be most impactful. So in our study here today, we will test both a contact intervention and an educational message to determine their potential impacts on reducing SUD stigma. And, additionally, we also assess changes in stigma by behavior, that being political support and financial donations. And so to our knowledge, no studies have linked changes in stigma to actual behavior.
All right, so now I'm going to turn it over to Miss Maddie Ashworth.
- Thank you. All right, well thank you so much, Robin. So I am gonna be talking today about this randomized control trial that we ran, or an RCT for short. So in this RCT, we tested five different treatments to reduce stigma and increase both the political and financial support of recovery housing.
And so we did this through an online survey experiment using Qualtrics. And so as participants took this survey, they gave us a little bit of information about themselves, and then they were randomized into one of these five treatments. So I promise we didn't show them all of these different interventions, just one, so they could be randomized into a control treatment where they just learned a little bit about what recovery housing is and what substance use disorder is, that's gonna be our control treatment, treatment number one, or they could go into one of our four interventions. So that
data message is going to be the kind of educational intervention where we describe the effectiveness of recovery housing, specifically, the Recovery Kentucky program. And then, we also have kind of three contact interventions where we would've provided participants with a story of an individual in recovery, so that's where Jade Hampton came in. He sat down and told us his story of recovery and they could have been randomized into the anonymous written story, the identified written story or the video story. And I promise I'll explain what all of those are in a little bit.
I'm not just gonna leave you with that kind of level of detail. And then, after they were randomized into one of those treatments, they were sent into our outcomes of interest. So the two kind of behavioral outcomes, the financial support and the political support, and then, of course, why we're all here, our validated measures of stigma to see how stigma would change after these interventions. So we ended up getting a nationally representative sample of about 2,700 individuals.
Our sample is representative across income, age and race. As we'll touch on a little bit later, we did oversample rural participants and Oregon participants, so it is not regionally representative. And because of that oversampling, we do have a higher proportion of females. But we'll talk about that later, it doesn't actually threaten too much in our study.
So moving on to what these messages actually actually looked like, what did we tell participants to try and change their behavior and their beliefs? So the first thing, the data heavy message, this was a little infographic that we showed participants, and it showed the outcomes, the 2021 outcomes from the Recovery Kentucky program. Okay, and so it's gonna touch on outcomes related to substance use, mental health, arrest, employment, homelessness, and then, mutual aid group attendance. So all kind of the big things that we know recovery housing is effective at changing.
All right, so participants were gonna be shown this little infographic, given a little background about what the Recovery Kentucky program is, how many people it serves. It's a pretty widespread program, I'm sure a lot of people in this room have heard about it, that serves about 2,200 people across Kentucky, okay? So this education intervention we wanted to include because we've heard a lot of information from our outreach and engagement specialists that work across the country to expand access to recovery housing, that kind of knowledge about what recovery housing is and how effective it can be is kind of a limitation when they're trying to get recovery housing established.
It's not something that the general public might know. So a message like this can kind of help us show how effective it is and kind of what it is, and so it might help people be a little more supportive of recovery housing. And then, we also have these contact interventions. So this is where we sat down with Jade Hampton, who, like I said, he's a director now of some Kentucky facilities, an individual in long-term recovery, and he sat down with us and told us his whole story.
When we sat down with him, it was an absolutely phenomenal experience. He talked to us for about 45 minutes about all he's doing now, his whole story with addiction and his recovery journey. So he gave us a 45-minute story. Around midnight, I learned some video editing skills.
We got that story down to about 3 1/2 minutes. So we didn't make participants watch a whole movie, but we definitely could have, 'cause he was a great guy to listen to. So with these contact interventions, we wanted to keep some things consistent with the data or education message, right? So we wanted to keep the recovery house that he went through consistent with the outcomes we were showing to make sure they're not like, "Oh, Recovery Kentucky's a one-off, not everything's that successful."
So we made sure he came from the Recovery Kentucky program as well. And we made sure that his story that he told us that we included in these contact interventions was consistent kind of with the outcomes that were shown in the education message. So his story is gonna touch on kind of his history with addiction, how he went through the criminal justice system, and then got diverted to recovery housing, but also how recovery housing helped him, what kind of methods within recovery housing were most effective for him. And just like we've heard, like Robin said throughout this week, that peer-support, social model really was that special sauce that helped him, and that recovery housing community was really what was most effective for him.
And then, of course, we wanted to highlight what he's doing now. He's absolutely amazing, directing all these large facilities in Kentucky. He went back and got some higher education and just has been absolutely phenomenal in the state of Kentucky. So we wanted to kind of highlight and brag about Jade a little bit as well.
So if participants got randomized into these interventions, they saw either an identified written story, an anonymous written story, or a video. So we have the video of Jade telling his story, and then, we kind of transcribed that video with a few minor edits for clarity, and either attached his name to it, said, "This is Jade Hampton's story of recovery," for the identified written story, or we just said, "This is an individual in recovery," for that anonymous story, right? And we wanted to kind of vary the social distance between the participant and an individual in recovery, because we know that sometimes it's not always safe, especially, in a rural community, for an individual in recovery to kind of come out and tell their story.
So we wanted to make sure that we could say, "Listen, this anonymous story is just as effective as if you have your name attached to it." So kind of get out there and say, "Listen, any kind of story of individuals in recovery will be effective, even if it's not your video, your face talking in a video, or your name attached to it." So that's kind of why we wanted to include a little bit of that anonymous versus identified story. So those are an overview of our interventions.
Like I said, they just went into one of those five, the control, the written stories, or the data-heavy message, or the video. So then we can move to our outcomes of interest. So like I said, I'm an economist, I don't like hearing about what people believe, I wanna actually put our money where our mouth is, I want people to actually change their behavior. So while I think that those validated stigma scales are super important and we should be measuring those, I think we need to include some actual behavioral changes in our studies.
And so the first one we're gonna look at here is political behavior. So we're gonna look at petitions for and against recovery housing that's actually gonna be sent to their state governor. So participants were going, they knew that these petitions, their name was gonna be attached to them if they chose to sign it, and it was gonna go to their state governor. And so the first one we have here is a petition for/against investment in recovery housing in your state.
So participants got to choose to sign either the petition supporting recovery housing, against recovery housing, or no petition at all. Didn't have to, they could just keep on going through and keep answering all of our hundreds of questions it seemed, I'm sure, at that point. And what we also wanted to do is include a petition for recovery housing in their zip code. So we wanted to get that state level petition, do you support it in your state?
But then we wanted to go further and see if they would actually support recovery housing in their zip code. So those who chose to sign a petition supporting recovery housing in their state, got another question that asked them if they would sign this petition for recovery housing in their zip code, right? I'm from the state of Wyoming, it's a big state. A recovery house on the opposite side of the state, I'll never see those people, I'll never interact with them.
But if they're in my zip code, that could be next door, that could be down the road, right? So it's a little bit getting at that NIMBY question. If all of the people who sign that petition for recovery housing in their state don't go ahead and sign that zip code petition, we probably have some NIMBY going on. We might not have the kind of support that we would want to see in establishing that recovery housing.
So then we also have, again, because I am the economist, I like money, so we had to have a money question in here, so we have the financial support of recovery housing. So in this, we randomly gave half of our participants an extra $5 for completing this survey. So through Qualtrics, they get a standard amount of money that they get to take the surveys, right? And we said, "Hey, you got randomly selected, you're gonna get an extra $5.
Here's this nonprofit, the National Alliance for Recovery Residences, or NARR, that helps us expand the kind of quality of recovery housing across the country. Of those extra $5 you got, do you want to donate some amount of it to NARR? You don't have to, you can say zero. If you say you don't wanna donate any money, you'll just get paid $5 at the end of the survey," right?
And we wanted to include this donation question because I don't know if you've been on Facebook, but it's very easy to click that sign your name to a petition, you walk away, you feel good about yourself, and that's it. But if you're getting $5, you know, you could go to Subway with that, you could get a cup of coffee, you can do a lot with $5, not that much, but a lot. And so we wanted to make sure that people will put their money where their mouth is and actually support recovery housing, and also figure out who we wanna be inviting to our fundraisers when we're trying to establish recovery housing.
So that's our financial support question. So next, we had our validated stigma scales, right? And so we have, there's tons and tons of stuff out there, ways that we can actually measure stigma. There's so many different ways to do it.
But we really wanted to drill down into the social kind of community-wide stigma towards both individuals with SUD, but also towards recovery housing. And so we found this Community Attitudes and Mental Illness scale, the CAMI scale, as I'll be referring to it, and we adapted it for substance use disorder, right? So where it used to say severe mental illness, we included substance use disorder. It talks about recovery housing and SUD treatment, all right?
So this is a 40-question scale that has 10 questions per subscale. So it's gonna have four kind of domains of public stigma, each with 10 questions associated with it. So the first subscale that it's going to have is this community treatment subscale. And this is kind of the one I was most interested in, being the one looking at the recovery housing support and stigma towards recovery housing.
So this subscale is going to get measures of how important people think kind of residential community treatment is, getting treatment directly in a community, rather than kind of on the outskirts of town. And it's also gonna touch on how dangerous having SUD treatment and recovery housing is in your community. So we're gonna kind of get at that dangerousness piece that Robin talked about earlier. The other subscales are benevolence, which talks a little bit about the sympathy for those with SUD, how much we should be kind of caring for them and investing in their treatment.
The other subscale is authoritarianism, which is kind of like the preferred social distance you would like to have from those with SUD. They shouldn't be treated as outcasts of society, things like that. And then, finally, this last domain of the CAMI scale is social restrictiveness, which is kind of the rights that those with SUD have, some of the statements you would really hope that everyone was like strongly disagreeing with them. So that is our CAMI scale.
So this is kind of our main stigma scale of interest that we're gonna be looking at when we're connecting stigma and reductions in stigma with our treatments. So next we have our final stigma scale, and then, I promise I'll get to the interesting results, this is the last method slide. We also included this affect scale measure because, right, we've posited to you that recovery housing is super effective, so we've created this recovery ecosystem, we should have a lot of individuals in recovery in our community now. And so we wanted to get a measure of how people feel about those individuals in recovery in their community.
And so this affect scale measures participant's perceived emotional response if they were to encounter an individual in recovery. So they would get to say, you know, if you encountered an individual in recovery, would you feel more pessimistic or optimistic, supportive or resentful, fearful to confident, right? So 10 opposing emotional sets that they got to rank how they would feel. And so, again, this gets us a little bit at that kind of dangerousness, their perceived danger that those with SUD and those in recovery might kind of pose to them in their residential community, right?
So now we can get into some of the actual fun stuff. So the results, so, of course, because we're at the Rural Stigma Summit, we are gonna be looking at all of the comparisons of rural versus non-rural. We know that rural people and rural communities, generally, respond a little bit differently with public health things, and just, generally, with their beliefs, than maybe non-rural participants are. And so we let people self-identify if they were rural or non-rural, where they lived.
We did validate it with zip codes and everything, and it was pretty consistent. But sometimes that self-identified rural piece gets a little bit more at how people feel. So we have here with the rural, the first thing we wanna take away is that 74% of rural and non-rural participants wanted to support recovery housing in their state. So they signed that petition that was sent to their state governor supporting recovery housing.
So that's one really great like externality of this survey is that we were sending lots of petitions, which, at the time, we were a little upset that we had to sign so many petitions. We had to send them, it took a little bit more time than we were expecting, but we were more than happy to do it. What we also saw is that there's pretty low opposition to recovery housing, right? We only have about three to 4% across rural and non-rural that signed that petition against recovery housing.
Most of the people who didn't sign for were just kind of ambivalent, "I don't wanna sign a petition. I don't know what these crazy researchers are doing. I'm not interested," right? And so that's a pretty low kind of opposition to recovery housing, which is kind of promising for a lot of the work we're doing.
But we do see some evidence that NIMBY exists, right? We don't have 100% of people who signed that petition for recovery housing in their state going further to politically support recovery housing in their zip code, right? That is statistically significantly different from 100%, which I know you don't wanna hear me say, statistical significance. So if I mention a result, let's assume that it's most likely statistically significant.
So we do have a lower number, or a lower kind of rate of NIMBY in rural areas, as compared to non-rural areas, right? With 83% of people in non-rural areas politically supporting recovery housing in their zip code, compared to 90% in rural areas. So that was a big surprise for us. We would've thought maybe that NIMBY was gonna be a little harder, and a little bigger, and harder to overcome in rural areas.
So this kind of baseline level of support was maybe helpful, was a little nice to hear that maybe it's not gonna be so hard. But, of course, what we're not taking into account here, this is the general public's feelings. Who's actually going to the town hall and making a stink? Maybe it's not our survey takers.
So that is something to consider. There's general broad public support, but you might still encounter, you know, you're at your town hall, you still might be yelled at a little bit, right? We also see kind of a flip though with the financial donations, right? It looks like non-rural participants are a little more financially supportive of recovery housing as compared to rural.
It's not a big difference though, but it is kind of looking like a little bit of a trend. So the non-rural participants donated an average of $1.82 to NARR, while the rural individuals donated $1.77 to NARR.
And then, finally, with our two stigma scales here, we see that there are generally positive feelings towards those in recovery and low stigma. So I know that that number means absolutely nothing to any of you right now, but the CAMI scale, when it's ranked and scored, a score of 140 is going to be neutral. So that means they don't care at all about what anyone's doing. If it's above 140, that means that they have lower stigma.
If it's below 140, they have higher stigma. So, generally, what we're seeing there is that there's less stigma. There are, you know, more positive than neutral, which, I mean, I guess sometimes that's all you can hope for, is that they're a little bit above neutral. And then, in the affects scale, that neutral score is going to be 45.
And so, again, we have some positive feelings towards individuals in recovery, but they're not overwhelmingly positive. You know, it's still pretty close to 45 there. Okay, so then we can get into our actual treatment effects. Did anything we do work, any of those messages actually work?
And so what this graph is, is it's going to show us the share of participants who signed a petition supporting recovery housing in their state, across all our treatments, and across rural and non-rural participants, okay? And so what we have here is, you're gonna notice that I talk to you about five treatments, there's only kind of four little bars there for each rural and non-rural. What we found is that the anonymous written story and the identified written story, there were no differences between them. They acted exactly the same for both political behavior, financial behavior, and stigma.
There was no differences across individuals, things like that. So we went ahead and combined those just as one written story so that the graphs look a little bit less crazy. And so what we see here is if we're looking at the non-rural kind of panel right there, we see that in the control, where they got no extra information about recovery housing, 69% of people were willing to sign that petition, politically support recovery housing in their state. And if we showed participants this written story, we got up to 75% of people willing to support recovery housing in their state.
Okay, so a six percentage point increase, it's statistically significant before we correct for some multiple hypotheses and things like that. So it's hinting at a trend that we're seeing that these messages might be effective. All right, and so that's not a huge difference, but a lot of times these public health messages aren't kind of done as a one off. We don't just hand people a flyer and walk away forever, right?
So they might be done in tandem with some things where that 6% might get us a little bit, and then, the further outreach and actions we're doing might get us a little further, right? But what we see in the rural panel is a bit of a different story. So if we're looking first at this control panel, we see the control bar, we see that in rural areas without any additional information, rural people are more likely to politically support recovery housing. A higher share of them sent that petition as compared to non-rural individuals, right?
But what we also see is that we can't do anything to rural people's behavior, political or otherwise. They're just stuck, they're not doing anything. And, in fact, our data message and our video message, it reduced the share of who was willing to sign that petition, all right? So it backfired on us.
A treatment that we thought was relatively effective, maybe a little ineffective in non-rural, might actually backfire and not work in rural areas, all right? And so that's a little surprising, that's not what we wanted to see, right? We did not want to see that our interventions are backfiring in rural areas. And we have a couple thoughts of why this might have happened, right?
So the first being that our education message kind of had a before and after of who's in recovery housing, right? It says what percent of people had an arrest in the month prior to going into recovery housing, and then, we wanted people to look at how low that number was after, and kind of make the assumption, "Look at how good it is," right? That's what that infographic does. But in rural areas, it might be especially kind of salient to them that, "Oh my gosh, look at that really high number moving into my neighborhood."
In rural areas, that's a really big deal, kind of the outsiders coming in. And so some of those rural kind of beliefs and community values might be a little bit more threatened by that initial kind of who's coming in. So a message that might have looked just at the changes and maybe not given that initial rate, might not have had this kind of backfiring. We also know that who's communicating the message in rural areas really matters, right?
They don't wanna hear from me, the academic, who's just on their little ivory tower preaching about how good recovery housing is. They wanna hear from their local family physician, their community members, their family members. Those are the messengers that they trust. And we've seen time and time again with public health messaging that it's way more effective with a local, trusted, community member giving that message than the government, or some online survey, or Facebook, you know?
And so I think who we were delivering that message might not have resonated as much with rural participants, might have backfired a little bit for us. So maybe testing, you know, if it's a local community member or someone else might be a little bit more effective at delivering that data message, okay? So, again, not super comforting here with the rural results, but I promise it's not all bad. We might have just backfired a little bit with that political support.
So now we can also look at how our treatments change stigma, which is what we're all here for, right? That political behavior, that's boring. We wanna know how it changed stigma. So the non-rural participants, we could do a lot for them.
We could reduce their stigma by a pretty substantial amount, not a ton, right? These are public health messages, they're not, you know, too exciting. They're not actually sitting down with the person, you know, they are just reading text on a screen. But we did have significant reductions in stigma when we showed people that written story, either identified or anonymous, and that as measured by the CAMI scale.
So that's gonna specifically be reducing stigma related to the community treatment beliefs and benevolence. And so what we found is that that written story actually did address some of those perceptions about dangerousness. So there were some statements in that community treatment subscale that looked at how much recovery housing poses a danger to individuals in the residential neighborhood, how it downgrades the neighborhood, things like that. And with that written story of Jade talking about how helpful recovery housing was for him, we actually were able to kind of change some of those specific dangerous beliefs.
Similarly, that written story and the video story did significantly increase positive feelings towards individuals in recovery, specifically, feelings around optimism, tranquility and relaxation. So didn't quite get at the fearful ones yet, but, I mean, relaxation and tranquility kind of gets there, so we might have addressed a little bit of the feelings of perceived danger. But, again, if we now look at the rural side, it's less happy, our results seem to be less exciting. We weren't able to do anything for stigma in rural areas with these messages.
They didn't resonate with them. And so a lot of the reason we probably saw no change in political behavior is because we weren't able to actually change any of their beliefs or their stigma. So if we can't design an intervention that reduces stigma, we're definitely not gonna have an intervention that's changing behavior, okay? And you're all saying, "Wait, you were so excited to talk about that financial piece.
Where's that slide?" Nothing we did changed financial behavior at all, so I wasn't gonna waste a slide on it. We couldn't change anything about how much they wanted to donate. So nothing changed there, rural, non-rural, the effect was zero.
So our treatments did not do anything. Don't show that written story before your fundraiser, it's not gonna help you. Okay, so who's more likely to sign a petition or politically support recovery housing in their state? So this is a graph that comes from a probit regression where we regress the likelihood of willingness to sign a petition on all these participant characteristics.
So that's just a fancy way of saying we could see who's more likely to do what. And so the first one we're gonna notice here is this low-income earners. So low-income earners in our study are defined as those who are making less than $25,000 a year. And so they are about five to 7% less likely to sign a petition supporting recovery housing in their state, as compared to middle income earners, okay?
And you'll notice here that a lot of these characteristics about who people are, some things like gender, age, things like that, those are not significant effects, right? If that 95% confidence interval is crossing that red line, that means it's a null effect. Nothing happened with those characteristics, we can't say anything about it. And so when we control for things like stigma, a lot of those demographic characteristics that we, generally, associate with political behavior disappear.
If we can control for those beliefs, those demographics aren't as important. And so we'll come back to the stigma piece in a second, but the two kind of types of people that are more likely to sign a petition are those who have a family member or a friend with an SUD, and those who are familiar with recovery housing, right? Seeing some nods, that makes perfect sense. Of course, if you're familiar with it, you're gonna wanna politically support it, right?
If you're at this summit, you likely would've signed that petition. But if we look here at this CAMI score, so this little guy right here, it looks like my dot is almost on that red line, but not quite. And so what that's gonna tell us is that stigma has to change by a lot before we can actually change political behavior. So I'm gonna say a sentence that's probably not gonna sound like it makes a lot of sense, but a one point reduction in stigma, as measured by that CAMI scale, is associated with a 0.6%
increase in the likelihood of signing that political petition, okay? So I have to change stigma by a lot of points, or on this scale, I have to have a huge reduction in stigma before I'm gonna have any meaningful change in who's gonna sign that petition, okay? And so this kind of gets at that contact and education interventions that we were talking about, they're really helpful in the literature that we show we've reduced stigma, we clap our hands, we say we did it, and we walk away, right? But a lot of times it's not actually gonna translate into any changes in behavior.
And so it's not gonna change the amount of people that are showing up to support recovery housing at these town hall meetings. It's not gonna increase the amount of people who are handing out the flyers. They're gonna sit back in their house and they're gonna feel good about it. But sometimes we need to think a little bit beyond about what people are actually doing.
And so we need to consider the actual behavioral implications of our interventions. So now going back to who's more likely to donate, this slide, I want you to take a picture. If you run any kind of fundraisers, this is who you're inviting, and you can see who's gonna probably fork over more money, okay? So don't invite young people.
Generally speaking, the 18 to 34, 35 to 54, they're all gonna donate less money, compared to those who are 55 plus. As someone who is in that 18 to 34, I could guarantee you that I'm probably not gonna donate much to your fundraiser. I'm gonna go buy dinner, okay? So we don't wanna invite the young people, but we really want to invite those who identify as liberal, compared to moderate, those who, generally, trust people, which I know, how do you decide that when you're doing your screening criteria?
But who we also really, really wanna invite is our social workers and counselors. They are gonna donate, on average, a dollar more than those who are not social workers and counselors, which is 20% of our $5, right? So we really wanna invite social workers and counselors if we're doing any kind of fundraising, all right? So thank you to all our social workers and counselors out there, beyond all of the amazing work that you do, thank you so much for being so financially generous, especially, in our sample.
And you'll notice here, again, I'm gonna mention this CAMI thing, because stigma's why we're here. Now you're really saying that dot looks like it's right on the red line, right? But that's because I had to make the dot a size you could see. But a one point reduction in stigma increases the amount you would donate to NARR by one cent, right?
So it's there, it's statistically significantly not zero, but it's such a small amount that if you really think about how much I, again, have to change stigma to get you to pull out your wallet, it's a lot, right? All right, and then, finally, we have our last bit of results, the last graph I'm gonna show you, who is more likely to have higher stigma, okay? And this is going to be stigma as measured by the CAMI scale. So the first thing we're gonna notice here is that anything to the right of our red line, that's gonna be lower stigma, higher number, lower stigma.
If you're on the negative side, you're gonna have higher stigma, okay? And so we're gonna look at the people who have a little bit of higher stigma. And so those who were criminal justice professionals in our study had significantly higher stigma as compared to those who weren't, right? And we defined criminal justice professionals, those were police, department of corrections, judges and lawyers.
So you can decide among those four groups who's driving that result. I'll leave that up to you. But what we also see, those who identify as conservative, a little more likely to have higher stigma. Those who are high income, who earn over $100,000 a year, likely to have a little bit more stigma.
And then, medical professionals are likely to have a bit more stigma. So that kind of goes back to what we've been hearing throughout this summit of kind of where people will run into kind of institutionalized stigma and where we might need to be targeting these interventions. What we also see on the flip side is that those who are female, those who are familiar with recovery housing, with SUD, those who trust people, and, of course, those who have an SUD are less likely to have stigma, okay? All right, so then our key takeaways very quickly, we have high levels of recovery housing support, in general, in the public, right?
They're willing and able to, maybe not go out to the town hall, but they will sign that petition. But what we also know is that messages aren't working the same across reality, so we really do need to be thinking what's gonna work in rural areas, and not just looking at the general literature, looking at, "Oh, this worked in a national sample," really look for what's working in rural areas. We also know that these public health messages had limited ability to change behavior. They even had limited ability to change behavior that was almost costless.
They had to click a button to send that petition, they had to write in, it was windfall money, we gave it to them, it's not like they earned that money. And so if we can't change that really costless behavior, what are we gonna be able to do for kind of bigger behavior? Maybe the public health messages, like I said, they aren't the end all be all, but all of the work that you all are doing, kind of the activism in the community, the advocacy, that's what's maybe gonna be making more of a difference in tandem with these public health messages. But we know that we can reduce stigma with these messages in non-rural areas.
It's not all hopeless, they were not useless, right? But they did have some limitations. And then, of course, the familiarity with recovery housing and those who have a close friend or family member with SUD, those are gonna be the biggest predictors of who's kind of gonna be out there doing the work, politically and financially, supporting recovery housing. So I'm going to hand it back over to Robin for just a little bit more on our limitations and next steps.
- Alrighty, thanks for all that great discussion about the results. Just to mention, every study has limitations, so we just wanna point out a few in our study. So, first of all, in our study, our intervention, involved only a middle-aged, Caucasian man, Jade, and so we were unable to control for the impact that his characteristics may have on the message he delivered. Also, with the intervention, we tested its impact on stigma immediately after as a reaction, and so we really don't know if we went out to survey these individuals, same group, three months later, would the results still hold?
So that is another limitation. And then, also we only assessed stigma towards SUD in general. We didn't ask about an individual with alcohol use disorder, stimulant use disorder, et cetera. And we know the individuals based on their knowledge or their understanding of what a substance use disorder is and maybe their experiences, these things could change their levels of stigma.
And then, lastly, this was kind of talked about already quite a bit, although, generally representative, we had a high proportion of females in our study and they exhibited lower levels of stigma, but this doesn't necessarily translate to not being generalizable, as the significant difference between females and males wasn't such that it would impact that generalizability. And then, implications and next steps. So where do we go from here? So we definitely will be conducting more studies in the future and we hope to see a lot of other researchers out there collaborating with individuals in recovery to develop some of these studies.
And we definitely recommend leveraging strategies from other disciplines, getting out of our silos, and learning about what's been successful in changing beliefs and behavior among individuals in rural areas, so partnering with individuals in rural areas. And also, just the public health in me, we really wanna just encourage a socioecological approach in which that means intervening at every level. And we all know here that we need to do that, and everyone's doing that here. We've heard so many wonderful stories about people, boots on the ground, taking action, talking to each other, sharing those stories.
So yeah, we just recommend that. And, of course, we're gonna share these results so people can get copies and cite them, and build upon that research. And I know we're at time, and just if you're someone who is in recovery and has the comfortability, stability, and desire to share your story, these results do indicate that sharing your story is very impactful, even if anonymous, like a blog post. So we just recommend that based on these results.
And if you all have any feedback, wanna collaborate with us, share information that we're unaware of, we love to talk, collaborate at Fletcher Group, our leadership constantly pushes that. So we love friends and it's been great to meet a lot of you here today at the conference. And lastly, but most definitely not leastly, we just wanna thank anyone who's in recovery that has shared their story with us, people we work with, people we've met at the conference, you all are amazing. And the work you all are doing to actually share that lived experience, support individuals, is key to making a difference.
So thank you very much. (audience clapping) - Okay, ladies, thank you for that, very interesting. This is an opportunity now for any type of questions that you might have.
I know, I'm excited. I love questions, it's my favorite part.
- [Audience Member]
Hi, I was wondering how you made sure that the participants read the entire story or watched the entire video?
- Yes, okay, that's a very good question. So we had a couple attention checks throughout. So when they were presented with that story, we asked them some questions about like, where is this guy from? What is recovery housing?
To make sure they had like kind of read it, but we also included some timing checks on there as well. So, you know, we timed how long it took them to read it. So we know the video is 3 1/2 minutes long, so it's probably not gonna take 'em that long to read it, but it should take them kind of somewhere similar. So we know that some timing checks, some kind of comprehension checks.
That's how we did read it. But that is a very good question, because we did have some problems kind of getting people to read the story. As you would imagine, no one reads the terms and conditions, no one's reading this story, it's long. So we did have quite a bit in the sample that we had to kind of drop from the analysis that weren't reading the story.
And so that is a takeaway that sometimes it can be hard, if it's a long story, to actually get people engaged. But if they are engaged in reading the stories, that's what we know the impact can have. So, great question.
- Okay, real quick, I have a question from one of our virtual viewers here. "And why do you think a social worker supported SUD with donations, yet, had the negative stigma"?
- Right, okay, so let's go back to our stigma. Let's see, higher stigma. Okay, so I will preface this by saying they are on the, like their little dot is on the negative stigma, like higher stigma there, but it is not going to be significant that 95% confidence interval is crossing that red line. So we can't say for sure that social workers have higher stigma, that could be a random effect in our sample.
But what we can say with this financial piece, whoop, there we go, is that they for sure are donating more. So it could be the experience that they have with individuals in recovery working with those with SUD could make them more financially generous. Maybe it's the type of people who are social workers, they're a little more caring, a little more generous maybe, especially, more more caring and generous than economists I'm sure, so.
And probably the lower income bracket in the scale that you have there.
Question again here.
- It's more of a statement. So I'm from the South Carolina Office of Rural Health, and I'm really interested in what it would look like regionally. So we know rural looks, obviously, I've seen not surprising while I'm here, what is rural in South Carolina may be completely different from you guys. So I would be interested in seeing a more regional survey done just to really gauge the changes that occur.
And so we're doing something like this right now in a rural county in South Carolina, and some of the outcomes that you guys have, I'm gonna be comparing to what we see in our county.
- Awesome, I will say we actually do have a little bit of analysis that we can touch on a little bit regionally, not so much of like South Carolina versus Wyoming. Sadly only four people in Wyoming took my survey, that seems rude. But we did break it down by the, and Robin, you might have to remind me what the name of it is, but the...
- It's rural-urban continuum.
- Rural-urban continuum, so very rural, to like metro, to pretty urban, right? And what we found is that the really rural to kind of semi-rural is pretty much consistent. It's the same, it's homogenous, as what is rural here. When we actually break out urban, non-rural, into suburban and urban, the ones who are the most negative are suburban, people in suburbia, so, which, I mean, I don't really know what to do with that.
Maybe the soccer moms in the minivans aren't happy with recovery housing, I'm not sure. But it is that kind of suburban people that are kind of driving some of the lower political support and higher stigma.
I have another question over here.
- Thank you, thank you, both, for the way that you presented difficult data in an easily to understand way that, you know, when I talk to researchers and try to interpret what they're saying, I usually leave more befuddled than I was when I started. So thank you very, very much for that. You know, as you know, I've worked with Fletcher Group for some time, have a lot of respect for the work that you're doing. I have two questions.
The first is that I see that your participant's about 2,200 people roughly?
- 2,700, okay. And so how did you measure or assure that you had removed selection bias in that population?
- Yeah, that's a really good question. So we ran our survey through a company called Qualtrics. And so we basically say we want 3,000 participants and they have all of these sampling methods. They have a panel of people that, they're not professional survey takers, so Qualtrics has been shown to be more reliable than other kind of survey mechanisms, like Mturk and some other ones, I can't remember all the names of them.
But these panels are kind of more regular people and a little less the people who are like in their basement, like clicking through surveys like crazy. So we worked with Qualtrics to make sure we had that kind of representative panel. So it's the best we can do for this kind of research. But because it was kind of incentivized with some money, also to take the survey, they get paid for it, that does help a little bit with our self-selection bias.
They also weren't told what the survey was about when they got to click on it. So they were told, "This is a survey, it'll take 15 minutes of your time, it's about health." So they weren't told it's about substance use disorder and recovery housing. It's very broad.
So that's kind of some of the quality control that Qualtrics does. They have a lot of others. I don't remember all of the like step-by-step bullet points that I'm sure I was supposed to give if I was a good representative, but.
- Okay, no, that really helps, thank you. The second question is, so I work in Maine with recovery housing, several dozen recovery houses throughout the state certified by NARR. And we've approached this issue with stigma a lot and really tried to be thoughtful about our work with the governor's office and their programs. And I was somewhat dismayed by the lack of effect of the interventions, the public health type interventions, that you used.
Not that they weren't thoughtful and creative, they certainly were, have to start somewhere, but there was no massive shift, especially, in stigma and the CAMI scores, and so forth. So that's depressing. On the other hand, we have had some success, we think, not with the scientific rigor to measure that you have, with a couple of interventions that I would be happy to describe. But before I do that, I'd be curious, have you thought of other interventions that you might wanna test the scientific rigor, knowing now what you didn't know then?
- Yes, absolutely. We have probably too many. I think I bother our CEO a lot with like, "Hey, but what if we did this, but what if we did this?" And so one of the big ones we wanna test, especially, in rural areas, is that messenger.
Is it me, "Hey, I'm a researcher at the University of Wyoming." They're not gonna listen to me. Half the people who know research at the University of Wyoming aren't listening. And so, but if we can say, you know, it's your local public health agency talking, it's your friend talking, it's a community member, kind of who's delivering that message, that's what we really wanna test as well.
But we also wanna change kind of who's telling their story, right? Jade, an absolutely wonderful guy, but, you know, he is a white, Caucasian male. If it were a female, an African American male, African American female, Asian man, you know, could be whoever you think would have a substance use disorder, there's this expectation theory. If you expect that person to have a substance use disorder, you're gonna be more receptive to the message.
Your beliefs aren't being challenged by kind of who's telling you their story. But also if you look like the person, you're more likely to be receptive to that kind of message. So if we can kind of target sometimes neighborhoods that we're trying to put a recovery house in, make sure the people who look like them are telling the story, are who they expect, that is a study we're working on right now to kind of disentangle those effects.
- One of the interventions that we found to be very helpful as we put this on our Facebook page and wherever we can, our newsletter, is, as you probably know, one of the NARR requirements is to have a good neighbor policy. And we take that a little step further by saying, "You've gotta do community service, especially, on the block that you live in, not just in the wider community where you reside." And that takes the form of neighborhood cleanups, and helping people clean out basements, and shoveling snow. And if you're walking by and you see a trash can's turned over, you know, you stop and you take care of those things.
And we take pictures, and we put these out, and we think this is a, you know, tremendous response to people, especially, who are directly affected by those who are in the locale. The other thought we have, we haven't done yet, is to create a video, two to three minutes of a town meeting, where you're opening a new recovery house in your community and there's a bunch of townspeople with all the typical tropes, "Not in my backyard, we don't want those people here, my kids won't be safe," you know, all of that. And then, on the other side, you have a very small group of people, the prospective new operator, maybe a couple of prospective residents, and two people who live near a recovery house that's been open for a while.
And so after all the negative things have been said, those two people stand up and say, "Well, you know, we thought the same way that you did before this house opened, and everything has changed, it's completely different. Instead of a drug-infested house, we now have sober people living there. Instead of people that the cops are always being called, they don't show up, the cops don't have to come, and we have all these resources now in our community that we didn't have." That might be an effective intervention if you could weave that into a short story.
- Yes, definitely. I would love to say I'd volunteer our research team here. We would love to work with you to test some interventions like that in this kind of manner. But, definitely, we've seen that that kind of back and forth, frequently asked questions, kind of dispelling some of the misperceptions that they might have has been effective in other contexts, so I bet it would be super effective here.
And I love the community one, I think that's a great one that we should add to our list. Thank you.
- Okay, I have another one here from our virtual platform. "I'm an intensive case manager and counselor from a rural community in the Washington State. And we have a lot of support from behavioral health organizations providing with a great peer-support program and the HARPS, H-A-R-P-S, funds for housing. But we still have a gap with recovery housing..."
Did I skip it? No. "Recovery housing program. What can we do to increase the number of recovery housing in my community?"
- Well, there's probably other representatives from Fletcher Group that could really respond to that. But that is, that's a good question. You know, that's all gonna depend on where you're located, the levels of support there are, and I think Ron could actually answer this, 'cause he's been in the field for a long time working the space. But I think that there are a lot of ways, you know, getting out there talking with Fletcher Group, in particular, we provide technical assistance, our boots-on-the-ground, outreach and engagement specialists through our HRSA-funded Rural Center of Excellence.
And these guys are our outreach, they're our experts for recovery housing development. That's what they do day in, day out. And I do know that it's a very complicated process state by state. Every state it's different, there's different challenges.
Some states, you know, NARR has a presence, there's a NARR affiliate, I think 30 so far. In other states, there isn't yet a NARR affiliate, and recovery housing, in terms of kind of the recognition of that resource, may not be as apparent at the state level. And so I think it just does depend on where you're at. But I think there are a lot of people, specifically, at this conference, that you could potentially reach out to.
You can reach out to us. And yeah, I hope that answered your question.
- Okay, we are at time, but I know there were more questions on our virtual platform and I know there are a few here. I'm sorry we couldn't get to 'em, but maybe you could flag down one of these ladies. But if you do have additional questions, you can submit them to this website, it's the URMedicine_Recovery@urmc.rochester.edu. I wanna thank everybody for their participation.
Some great questions out there. Thank you again. (audience clapping)
Madison Ashworth, MS
Research Associate, Fletcher Group, Inc.
Robin Thompson, DrPH, MPH
Director of Research & Evaluation, Fletcher Group, Inc.