Hello and welcome. I'm Dr. John Torres, the Director of Digital Psychiatry at Beth Israel Deaconess Medical Center and technology expert for SMI Advisor. I'm pleased that you're joining us for today's SMI Advisor webinar, Mobile Technology for Treatment and Support of Homeless People Living with Serious Mental Illness. Next slide, please. SMI Advisor, also known as a clinical support system for serious mental illness, is an APA and SAMHSA initiative devoted to helping clinicians implement evidence-based care for those living with serious mental illness. Working with experts from across the SMI clinician community, our interdisciplinary effort has been designed to help you get answers you need to care for your patients. Next slide, please. Today's webinar has been designated for one AMA PRA category, one credit for physicians, one continuing education credit for psychologists, and one continuing education credit for social workers. Credit for participating in today's webinar will be available until January 9th of 2022. Next slide, please. Slides from the presentation today are available in the handout area found in the lower portion of your control panel. Select the link to simply download the PDF. Next slide, please. Feel free to submit your questions throughout the presentation by typing them into the question area found a lower portion of your control panel. We'll reserve about 10 to 15 minutes at the end of the presentation for your Q&A. So again, you can type those questions in any time and I will be moderating those questions. And next slide. Now it's my great pleasure to introduce today's faculty for the webinar, Dr. Niranja Karnik. Dr. Karnik is a professor of psychiatry and behavioral sciences, as well as associate dean for community behavioral health at Rush Medical College in Chicago, Illinois. He is a co-director of the KL2 Career Development Award Program and the director of the Great Lakes Node for the National Drug Abuse Treatment Clinical Trials Network. As you'll see too, he's an expert on using technology, integrative mental health, and community work. So we're very fortunate to have Dr. Karnik joining us. Thank you so much for leading today's webinar. Thank you, John. Thank you for having me. It's a real pleasure and honor to be here. So as a beginning point, I need to disclose that I do receive substantial federal funding for my work. I'm very fortunate to be supported by both the National Institutes of Health and the Substance Abuse Mental Health Services Administration. The specific grant numbers are all listed here. These are the learning objectives for today. First, we're hoping to evaluate the range of digital and mobile technologies that can be used in psychiatric care. Second, I will briefly review the nature of what we call the digital divide that poses challenges for the use of technology in community settings. And third, we'll be analyzing the current research and its limitations for the implementation of mobile technologies to address substance misuse among the individuals with serious mental illness. So this is my outline. In order to address those objectives, we'll be going through a summary of homelessness and youth experiencing homelessness in particular. My own work is especially focused on youth, but I want to speak to the broader population as well. We'll be talking briefly about SMI and substance misuse amongst homeless populations. Then moving on to mobile technology and psychiatric care. I'll review with you studies that my team and I have done largely here in Chicago, which we call the stepping stone studies. And, you know, walk you through some of the results and lessons learned from that. And then I'm going to do a little bit of prognostication about the future and where we might be going in terms of precision psychiatric care using mobile health technologies. So with that, we'll begin. So homelessness, as many of you who are on this webinar probably know, is a major issue that continues to face, you know, the United States, which is the wealthiest country in the world. On a single night in 2020, roughly over half a million people were experiencing homelessness in the United States. Homelessness has increased for the past four years and unsheltered homelessness now supersedes the sheltered homeless population. In other words, there are more people now living on the streets and there are homeless individuals living in shelters. And this should be a very sobering finding for us. Chronic homelessness has increased to 21% of the population of homelessness. And sadly, but unsurprisingly, I think given the dynamics of disequity and disparities in our society, Black, Indigenous, and people of color are disproportionately represented in the homeless population. So on a single night in 2020, focusing on youth experiencing homelessness, approximately 34,000 people under the age of 25 experienced homelessness and are classified as being on their own as an unaccompanied youth. Unaccompanied youth are more often non-white, Hispanic, or Latino, female, or identify themselves other than male or female. So this means that these are amongst the most vulnerable individuals in our society who are represented in this population disproportionately. And 90% of youth experiencing homelessness are between the ages of 18 to 24, with the other 10% being younger. And it's largely the fact that the foster care system picks up a number of the younger children that shifts these numbers in this manner. But there are various challenges in the foster care system that I don't have time to go into today. So amongst the serious mental ill and homelessness, pooled estimates show that about 76% of homeless populations have at least one major psychiatric diagnosis. Diagnosis. Alcohol use disorder is the most common at about 37%. And just for reference, this is about 10 times the rate in the general population. Substance use disorder has a pool prevalence of about 22%. And then this is followed by major depression with about 12 to 13%. And schizophrenia spectrum disorders are present in about the same, about 12% of the population. So this is a little bit of background data on a sample that we did, an epidemiological study we did of youth experiencing homelessness here in Chicago. In this study, we recruited only 18 to 24-year-olds from a shelter here in Chicago. And I mean, sorry, 18 to 21 in this study. And this was largely female and largely African-American. And this reflects the ethnic and racial dynamics in Chicago, as well as the geographies in Chicago. You can see here, the age at first homelessness on average was about 16, but the standard deviation around this is about three years. So there's a wide range of experience here in terms of when these young people are becoming homeless. On average, they had almost two episodes of homelessness in the last year. And the longest length of homelessness was about 15 months on average, with the current length of homelessness being 10 months. These statistics are hard to get and are not widely replicated in studies of homelessness. But we believe that we need to gather this information in order to better study this population, understand this population's needs and develop some of the interventions I'm going to be talking about. So in this same population, we used the mini international neuropsychiatric interview to ascertain diagnoses. And we found that psychiatric disorders were present in over 80% of the population of this sheltered homeless population. And you can see here that there was a number of mood and affective disorders. Suicidality, the mini is particularly sensitive to suicidality. Most of the suicidality that we found was very low levels of hopelessness and worthlessness that were triggering positives on this. There were a few cases of acute suicidality, which my team picked up and which we acted upon to ensure safety. We were surprised by the relatively low levels of post-traumatic stress disorder in this population. You can see in this sample, only about 12% met criteria for PTSD. And surprisingly, we found a few young people with bulimia in this population, which hadn't been previously reported. And then you can see this is all under DSM-IV diagnoses of substance dependence and substance abuse, as well as alcohol dependence and abuse. But relatively salient numbers there and much higher than in the general population. So where does mobile technology and psychiatric care come into this? So there are several approaches to mobile technology that we can think about. The first I'll start with are cellular phones. So cellular phones are your basic mobile phone. These do not necessarily have internet access, but have basic functions of being able to make calls and send text messages. So some of you might be familiar with these as being those kind of clamshell phones or the small little brick phones that folks have. And a number of people who are homeless actually have cell phones. The issue often is keeping the data plan active. It can be a challenge if you don't have resources. Smartphones, in contrast, have web access, have apps, have a whole suite of features that can be used. In one study in California, Eric Rice and his team found that over 90% of homeless youth experiencing homelessness actually had smartphones. Now, the smartphones didn't necessarily have data plans. They couldn't use them as phones, but they were often using them to connect to public Wi-Fi hotspots and locations. And that study was particularly informative for us in terms of thinking about next steps in our work. And I have to tell you that a lot of the smartphones and cell phones that we see when we work with populations that are experiencing homelessness, these phones have had a rough existence, just like their owners have had. We see phones with cracked screens and barely functional devices. And the users are trying to look through these very damaged devices and trying to make them kind of go that one step further for them. Far less common and far less available, but certainly in the realm of, as we think about mobile technology and psychiatric care, we should start to think about wearables. So these are devices which can pick up data and sense data on us in multiple ways. Those of you who have an Apple Watch, that's a wearable. It can pick up various kinds of data. Fitbits and other devices like that qualify under this category. And from a medical standpoint, we're starting to see a greater deployment of remote monitoring devices. So these are devices that can specifically focus on particular medical issues or medical testing that needs to be done and transmit that information wirelessly to a receiving station or a medical clinic that could be monitoring this individual. So blood pressure and glucose are probably amongst the most commonly done devices. Other areas that we need to think about that don't necessarily strictly fall within mobile technology, but may end up coming into the mobile space via apps or other devices are social media, chat bots, and virtual reality. So I just mentioned those broadly, and we can certainly go into these further during the question period. So I want to review some of the pertinent findings that sit behind the work that we were pursuing with homeless populations. First up, I want to talk about David Moore and his team's work. So David is based here in Chicago at Northwestern University. And David did a very interesting study of 40 adult participants and did two weeks of monitoring of these individuals. Of these 40 subjects, 28 had sufficient data for their analyses. And what they found was that the PHQ-9, which we use to measure depression, negatively correlated with a couple of things. First, circadian movement. So they were measuring this based on people's smartphones or devices. So circadian movement refers to our daily cycle of movement. And unsurprisingly, people who were depressed were less likely to move. They tended to stay in a single location. Second, this negatively correlated also with mobility between favorite locations. Location variance, so the variety of locations that individuals went to, and phone usage as measured along a variety of metrics. So basically what you see in this is that these individuals who have homes, all tend not to leave those spaces, basically, if they are more depressed. And they tend not to reach out to other individuals on their devices. So this is one of the first studies to really look at geolocation and think about it in a deep way. Around the same time as those studies were being done, Dror Ben-Ziv was pursuing a set of studies. And Dror's work is very interesting in that he's trying to specifically reach the SMI population. And he did a three-month augmentation study of acceptance and commitment therapy with mobile intervention and demonstrated feasibility, safety, and some clinical potential. He addressed assessment of persecutory ideation using this platform, an mHealth platform in 62 subjects. And what they found was that individuals who had these, there were correlations with a number of areas, including decreased travel distance, time spent in the vehicle, length of outgoing phone calls, time spent proximal to human speech, and increased time in sitting. And what they did was a method called text message hovering. So this was with a social worker. And basically they had this individual really trying to support this subject and improve a therapeutic alliance. And the description of hovering is basically like they were present, they were available, they were trying to connect with them. And they had 17 subjects that they looked at who had both psychosis and substance use. So I think this line of work that Jor's been pursuing is really innovative and really important to the field and important for us to think about how we're going to operationalize some of the interventions that he's testing in the community setting. That brings me to the digital divide. So these studies in some ways are clinical trials and studies that have been done in relatively fixed settings. And as we think about rolling out to the community, we need to think about the fact that access to technology is not an evenly distributed phenomenon. And I think most of you who are on this webinar probably understand that. When I do this type of talk or address this topic in person, I often ask people in the audience to raise their hand if they have one device with them that connects to the internet. Now, all of you are actually listening to this over the internet, so you obviously have at least one device. Then I ask people to keep their hand up if they have a second device on their person that they could use to connect to the internet. And most people have nearby to them a cell phone or a smartphone that they could use to connect. And so that tends to be a very telling thing. And then I tell them to continue to keep their hand up if they're wearing a third device that could potentially access the internet. And those of you who have Apple watches or other smart devices like that obviously have another device. And one of the things I point out to them is that the digital divide is best exemplified by that. The digital divide is best exemplified by that. Those of us who are on this webinar have a great deal of privilege. We have jobs. We make sufficient money to afford these devices. If we were to lose one of those devices, it would be an unhappy event, but it wouldn't be a tragedy. The other device that you have could probably partially substitute for this. I will confess that I once, about two years ago, locked myself out of the house. And I didn't have my phone on me, but I had my watch on me. And I was able with my watch to actually call a locksmith by kind of creatively using the device and then initiating a phone call and get a locksmith to come to the house and open up the house. Now, that's only because I have multiple devices. The folks that we work with in the community do not have that luxury. And so for them, the single device that they have is often the most important thing that they have. And I would contend that their ability to use those devices is even more important because they're searching for resources often. They're trying to connect to people that are in key ways. And I think that my contention has been for some time that we should be giving the individuals who are homeless and individuals who are vulnerable the best technology because they need that robust technology more than those of us who live in more privileged circumstances, because we have such a range. But studies generally show that they often only have the basic cell phones. And in many states, the state, Medicaid or Social Security will only pay for a limited device, the lowest level of device, which can only make, in Illinois, I think it can only make 200 minutes of phone calls a month is what they can get through the public system. And that's just inadequate. With all these devices in the community, challenges arise, the maintenance of the devices and the cost of service plans. Often the devices are Android-based devices because generally the costs of these devices are lower. In our studies, the youth that we work with who are experiencing homelessness generally want Apple devices. We discovered this after we tried to hand out Android devices and were met with a lot of questions as to whether they could have an iPhone instead. And theft and robbery are a huge risk for this population and having these devices. And in our studies, we go through a great deal of effort and time to educate the subjects that we're working with about the use of the devices because in our studies, we've been handing them new devices with data plans and we don't want them to be harmed as a result of having these devices on their person. So we talked about security and safety. When is it appropriate to use these devices? What happens if you're using them publicly or not? And we try to actually, we spend extra money to get them non-discrete cases. So black or dark cases that are not very flashy so that the devices are not as visible. So that's much of what I wanna touch on with the digital divide. There are several gaps in the literature related to SMI and homelessness. So first I'll say that most studies are relatively small. A lot of studies actually tend to generally look at house participants. They tend not to look at unhoused participants or individuals on the streets. Studies tend not to plan interventions specifically for the homeless populations, especially when they're dealing with this. Benzie's work comes closest because of its focus on the SMI population and in some ways touches on some of this. And one thing I wanna emphasize to everyone who's listening today is that I believe that when you lack a physical home, a virtual home may be more important. So I'll give you an example from our work. The young people that we work with, they keep in contact with adults and friends and peers, key people in their lives via connection to the virtual space. So whether that is on Facebook or on Twitter, they're often in those using WhatsApp or Snapchat. Those are the places where they keep connection to others. And in some sense, because they don't have a physical address, they often receive a lot of their communications and a lot of their news and a lot of information in that virtual space. And that's why I believe that they need the best technology when they're out there on the streets because this connection is so important. There is some belief in the literature that the digital divide may be overestimated, that homeless populations have high rates of cell phone ownership, but the challenges have not been well articulated. So I think I continue to believe that there is a digital divide that we need to address and that this is even more extreme than some of the communities that we see that have low access rates to the internet and low connection rates. And I will also add to this one thing, which is that the literature around SMI and homelessness and technology is really centered on the urban center. That's where most of the studies and the work has been done. There is precious little information about rural SMI and rural homeless populations. So, and there the digital divide is even greater. So I think we really, really need to think further about this and do much more comprehensive studies in this space. So with that kind of introduction and background, I'm going to now walk you through the studies that we've done here in Chicago that we call the stepping stone studies. So this set of studies, we initially, when we conceived of it, we really wanted to get feedback from folks in the field and in some of our community partner organizations. So we started this by doing a set of focus groups with youth in some of the shelters that we were working in already. And we really wanted to just understand their perceptions of technology, their feelings about it, their input about this. We had initially thought to actually do a sleep intervention because we knew from the literature and our own clinical work that sleep in a shelter is a really difficult thing to get consistent sleep, regular sleep, restful sleep, because in a lot of the group settings that shelters have, it can be very disruptive. And so we thought we would do something around sleep, possibly sleep and trauma. And the result of doing the work that led to this paper was that the youth told us, yes, sleep is a problem. Yes, we've had some difficult times and some trauma in our background, but they actually wanted more about how they really wanted coaching and support on how to have positive and healthy relationships and how to manage in their day-to-day world, because they found that a lot of times they were having conflict with individuals. They were very enthusiastic about the potential of using technology to do some of this intervention work and encouraged us to do this. And we asked them actually whether they would be willing to share data from their phones, whether we could gather data passively from their phones. And interestingly, they asked us why we wanted to do that. And in a number of the focus groups, we explained the rationale was we wanted to develop interventions using that data and using that information. So they actually agreed that this was a worthwhile thing to do and likewise encouraged us. Notably, they told us that they thought that Google and Microsoft and some of these tech giants were already doing this type of thing in terms of gathering their data passively, but without their consent. And we were trying to explicitly consent them for this. So in Stepping Stone 1.0, what we did was we developed a mobile phone-based intervention that had two major components. First, we installed a suite of apps on the phone. And these apps were drawn from David Moore's team. They have what's called the IntelliCare platform, which has a set of cognitive behavioral therapy interventions that are embedded in various apps. So we took an extraction of that and included that as part of the device. We also included an app that was developed by a local service organization called the Young Invincibles here in Chicago. And that had a lot of resources for youth experiencing homelessness. And they could actually search for shelter locations. They could search for beds, things like that. And we put on other vetted mental health apps. And we asked these youth to engage or we gave them the option to engage with a postdoctoral psychologist for about three sessions of therapy in the first month of the study. And those sessions would be scheduled and would be done largely over the device. And we were really just testing whether they would benefit from that or not. That study was about 30 subjects. And then Stepping Stone 2.0, just so you see the differences, and I'll get to a little bit more detail about these studies. Stepping Stone 2.0 was a fully automated platform. So we took the therapist out of it. There was very low utilization of that resource in the first study. So in the second study, we went towards a fully automated suite of apps and wanted to see how they did with that. So these are kind of the two major studies. And the first day was 30 subjects. The second study was 100 subjects. And we gave each of the youth in these studies a smartphone. They were given Nexus 5 phones, which at the time was a pretty advanced device. And we also gave them up to six months of unlimited data and texting and access on their phones. So that latter part, the data plan, was probably the most valuable part of the study for them. So this is what Pocket Helper, which is the part of the suite of apps that we put on there, looks like. You can see here, part of the app does these daily tips where they are given a kind of CBT tip to try, and they are asked to rate that tip based on stars. So from one to five stars. They were also pushed daily surveys. And for every two weeks where they completed 80% of the daily surveys, they would receive a small electronic gift card sent to them. And this was just a way for us to kind of keep tabs on them and their experience. These two things, these two steps that they did daily required all of seconds to do, you know, less than 30 seconds to complete. And they were rated actually relatively positively by the individuals. In fact, what we found in the result of these studies is that the youth actually felt that these components, the tips and the daily surveys were actually amongst the most helpful things because they were brief and they were very limited. They told us that the tips were, and the surveys kind of made them think about them and made them reflect. In some sense, it's that sort of mindfulness moment when you kind of look internally to see how you're feeling. They really didn't find a lot of utility in the apps that took more time or more effort. And they certainly didn't want to do anything that required them to call a person or speak to a person. They were very much averse to that and rated those the most poorly. But this has kind of led us to start thinking about interventions that could be modularized and reduced and done in very brief snippets of time. In Stepping Stone 1, we gathered a lot of passive data. And amongst that data, we gathered geolocation data. So these are just some examples from that study. Three different participants here showing their locations. And here you only see the longitude and the latitude is invisible. What we found amongst the 29 participants that we had usable geolocation data on is that what we concluded was clustering. We used a program called dbScan to pull these clusters together. We defined a cluster as something that was half a kilometer distance and having a minimum of five points within that distance. And that was what defined a cluster. And so the participants had on average about seven clusters. The range was quite large from one cluster to 19 clusters. So we see a lot of variability in this. What we're doing now is looking at the correlation between these clusters and various clinical measures. So do these clusters relate to anxiety or depression or trauma? We're also looking at how they relate to homelessness experience. So do youth who've been homeless more often or for longer periods of time have more clusters than those who've had less homeless experience? So that's part of the analysis that my team is doing right now. But as we looked at some of the data in Chicago, you can see here, some of these clusters follow predictable lines. We saw clusters that aligned with the L here in Chicago or with bus routes. And so you can see some of these follow linear patterns. And the L line ones are particularly interesting to us because we know some of the youth experiencing homelessness and some of the adult homeless population here in Chicago often will ride the L, especially in winter months when it's very cold outside. And they'll just go back and forth from one end of the L to the other and just let the train kind of keep taking them all night and they'll try to sleep on the train. The L is one of the few public systems that runs 24 seven. So they can at times stay on the L undisturbed. And when temperatures really fall in Chicago, the police are advised by the public health department not to disturb these individuals on the train and just let them be there because it's a safer place for them to be than to be out on the streets. So we were interested in that and we're gonna be looking at the temporality around some of this geolocation data and to get an idea of how this might trace patterns that would be actionable. Because one of the things that we are starting to think about is could we align some of our mobile crisis services or mobile mental health services with some of this type of data? That if we see a number of individuals who are kind of in motion on here, it is particularly a bad night here in Chicago and we notice that they're there, could we do some outreach and try to help support these individuals so that they don't come to harm in this situation? Second, we also gathered data about use of Wi-Fi nodes. So one of the things that we know is for the homeless populations that do have devices already and do not have data plans, they will often try to use public Wi-Fi locations. Even though our plans had unlimited data, we theorized that a number of the young people would continue to use Wi-Fi locations to try and conserve the data. There were a couple of young people that got data throttled because they were streaming so much video and music that the carrier actually throttled them. So we knew that they were kind of used to using Wi-Fi locations and would probably continue to do so. So we tried to characterize where these locations were and what types of hotspots they were. So a number of them were public hotspots and hotels. Some of these ones like computer peripherals that kind of came up, these are hotspots that their devices were pinging against, but the youth didn't necessarily connect to them. They used a lot of shopping and retail locations, restaurants, and we saw a few hospitals and clinics Wi-Fi addresses in there. So this to us was telling. A lot of things like Hyatt Hotels and Starbucks, we would see those because those are relatively easy to access public Wi-Fi locations. You can often sit nearby or just outside these locations and pick up enough signal to use those as a means of getting on the internet. Okay. So that brings me to kind of thinking about next steps and where we're going with this line of work and these areas of research and development of interventions. So one thing I wanna kind of highlight for you, some of the technical research that's happening broadly in this space. So our colleagues at NYU, Paul Glimcher and John Routreson, who do a lot of work around treatment of substance use and opioid use disorder. Paul Glimcher is actually a behavioral economist. One of the things that he came up with was this idea of ambiguity tolerance. So they found that individuals who use opioids tend to actually have less tolerance for ambiguity than folks who are stable. And one of the things that they've been doing is trying to look at, could they use a test of ambiguity tolerance to predict relapse? And their preliminary data shows that they can actually do this. And they did this actually on portable devices. So they were actually able to push this out to individuals in the community. And if they started to score, you can see here in panel A in the top figure, the individual enters treatment at time one. And as they're kind of having more difficulty, you can see that they start to have this fluctuation of ambiguity tolerance. And then the individual drops out of treatment. Whereas the person who enters treatment and stays stable tends not to have that degree of fluctuation. And so this has led us to think that this might be the type of tool that we could put onto devices to help support some of the SMI population that have substance use. And could we help them stabilize? And if you start to see this kind of fluctuation coming on their ambiguity tolerance test, if they complete the test, then maybe we could reach out to them and try to proactively get them in before they reach a point of full relapse. So to me, this is particularly exciting. And this is particularly interesting to think about as we think about community-based interventions. Next, I wanna mention a recent work by Dror Ben-Ziv on the gamification of SMI interventions. So Ben-Ziv did a fully online recruitment on Google and Facebook for this study and got 315 individuals with self-reported SMI. So they broke down roughly a third, said they had bipolar disorder, about over 40% said major depressive disorder, and about 20% or so had schizophrenia or schizoaffective disorder, the schizophrenia spectrum disorders. Now, the weakness of this study is that these were not clinically validated diagnoses. These are all self-reported diagnoses. Nevertheless, they randomized the sample to a smartphone intervention with daily exercises to address dysfunctional beliefs versus a weightless control. And they saw primary effects in depression, anxiety, recovery, and self-esteem. They saw changes for the better in that for individuals who got the dysfunctional belief intervention versus those who had the weightless control. So despite some difficulties in the study and some criticisms that can reasonably be leveled at this study, I think it's still a very interesting area of work because the more we can gamify and make these interventions engaging for populations, I think the more likely we are to get people to do them. The trick is, how do we embed within the gamification process enough of an evidence-based intervention that it has some clinical effect? That's the hard part with this. We know with most of the mental health apps out there, the retention in terms of use of these apps is very low. And with our homeless population, it's going to be even less. So we really need to be able to get attention. Now, the one thing I can say that we've learned in terms of working with the homeless populations is that in addition to homelessness being a very difficult experience, a rough experience, there's a lot of exposure, there's a lot of physical ailments, skin problems that come. One of the things that we've learned from a psychological perspective is that homelessness is also generally an experience of extreme boredom. And so that's why we saw a lot of our young people who were really using their phones for entertainment. They were trying to distract themselves. They were trying to get out of just being stuck and being in this situation where they keep having to live in that moment. And so gamification, I think, has some appeal for a population which faces boredom and has a challenge with boredom. And if we can break that boredom cycle, these interventions have real potential, I think. Finally, I want to bring up some research that one of my trainees, Veronica Ramirez, is doing. She and her psychology mentor, Stephen Schuller, are based at University of California, Irvine. And Veronica's starting to do work around heart rate variability and opioid use disorder. So what she's trying to do is develop a wearable device that can do some of what Glymtra's doing with the ambiguity test. But can we actually assess people's heart rate variability physiologically and try to assess when people are being triggered to relapse or triggered to use opioids versus not? And could we do this proactively and predictably so that we can try to assess, again, before people fall into a full relapse, can we try to then extend help to them or push assistance to them? The potential here is really quite good if we can actually set these parameters correctly. So Veronica's very much at the front end of this work and developing this line of research. So I want to kind of wrap up. I know we have about 15 minutes and we'll have some questions. But first, I want to kind of emphasize that mobile technology has the potential to extend and reach homeless populations with SMI. I think that it's really important that we think about giving this population the best technology. I get very upset when I hear people saying, let's just give them old devices, recycled devices, that somehow this population deserves less. No, I'm convinced that this population actually needs the best technology. They need technology which is rugged and will handle the difficulties that they're in. And I think that it is such an important tool just for engagement and contact. I know that a lot of the social workers that I work with in the emergency rooms, they really love the fact that we gave homeless youth these smartphones. They felt like it made their job so much easier to know that these young people had a device that was working with a plan that would allow them to reach them. And that's often one of the first steps in this, the simple step, if you will. And everything else we do, the apps, the other interventions, all will extend after we take that first step. Second, I think, as I pointed out with Glimcher's work and Veronica's work, there's real potential, David Moore as well, there's real potential to use technology as a means to advance a precision psychiatry approach, that we can really start to measure some of what the person is experiencing in real time. And we can start to get very creative with these devices. Our devices can now monitor our physiologic state in some ways, our heart rate. The latest Apple Watches, I think, can do oxygenation of our blood. But also, our devices pick up our speed at which we're in motion. They have light sensors. All these things, if we think creatively, can help us better understand what the needs of the individual are. I know that my colleagues at Northwestern and in California have used proxies of light, like when the smartphone is sensing darkness, and when your phone is plugged in. That ends up being a relatively good proxy for when you are likely asleep. And we need to kind of think from a new standpoint in order to come up with these sort of precision psychiatry measures or interventions that might be done. There are challenges with these approaches. I think largely related to, first, I mean, I'll repeat that the digital divide is a major challenge, but with the devices themselves, there is an issue of data reduction. So as a clinician, I really don't wanna see how many steps someone has done over the course of a year, plotted as a daily function. It's not gonna help me. So we need to get kind of creative as to how we use this data. I can tell you just from the 30 subjects that we gathered passive data on, we have so much data. It was actually a real challenge for our team to be able to pull this data and utilize it. The number of geolocation pings we got were happening on the order of minutes. And so think about that over a one month period, how many minutes are you moving? And how much data our phones produce in that regard is remarkable. So we really do need to come up with some good data reduction strategies to make this data that we get clinically actionable or usable in our daily work. And this is true not only for the SMI population and the homeless population, but also for all of psychiatry writ large as these tools start to generalize out. I wanna just kind of end on acknowledging my collaborators, my colleagues, my mentees. This work has been evolving over a number of years and I'm fortunate to have some really great people that have worked with me. I wanna especially call out the Knight Ministry and Heartland Health that are both based here in Chicago that do remarkable work as organizations with youth experiencing homelessness and adult homeless populations. We've been really privileged to be partnered with them for now going on a decade. And it really has made our work incredibly important to us and we value that collaboration greatly. So with that, there's the bibliography. You'll be able to see that in the PDF, a number of the studies that I referenced and the papers that I looked at in terms of preparing this. And I will say thank you for your attention today. I'll say, so thank you too for such an interesting presentation, Dr. Karnick. Before we shift to question answer, I wanna take a moment and let everyone know about that SMI Advisor is also accessible via mobile device, very apropos for this talk. You can use the SMI Advisor app to access resources, education, and upcoming events and even complete certain mental health rating scales or submit questions directly to our team of SMI experts. You can download the app now at smiadvisor.org slash app. So we'll go to the next slide. So we have some questions that have come in. If there's time, you can still put your questions in now and we'll work to get to them. But the first question is, you were saying that youth prefer to get Apple phones compared to Android phones. When you've kind of looked at results or kind of people using them, does it make a difference what phone people get in terms of kind of mental health benefit or how to use them? No, I don't think so, John. I think that the issue really has been that a lot of the research tools that have been developed were initially developed for Android devices. And that is largely because of the way that the Apple corporation manages their devices. They have much tighter controls in the Apple universe over our ability to extract data and utilize data. The Android devices, there's a little bit more flexibility there. So that was really the limitation that was driving that. I have to say, I'm an Apple user myself and I find the Apple products to be much, much easier to use than the Android products or Windows products. And that's why I say in the Apple universe, I think the young people prefer Apple because, for lack of a better term, I think in some ways, Apple iPhones are sexier devices. They're designed really well and they look good. And they are more expensive. And that's probably part of it, that if you think about it as a kid on the street and you want to show that you have something cool, it probably looks cooler than the Android devices. But those are broad generalizations. I think that there are a lot of other specific things that govern around this too. Well, I'll say, I mean, sometimes some of our folks who have bought two generation old used iPhones that are pretty good to get like an iPhone 10 and the price tag isn't like as wild, but I hear you, you have to give people good phones. I hear you, there's no point giving people a $20 trashy phone doesn't do anything. That was a wild point. A different question, a little bit broader. So we'll have to kind of think how to chunk this. But in the talk, you mentioned the use of chatbots. What would that look like per se? Or what would kind of an effective chatbot be to help kind of homeless youth with SMI? We actually loaded one chatbot. I think Coco is one of the chatbots that exists in the public space. And we had loaded that on there. I don't think these would be necessarily different for this population. It's about putting the tools that we're developing broadly for mental health, that we create for young people generally in front of these young people. Now, the Young Invincibles app, there's portions of that that could be turned into a chatbot type function. So when they were developing their app, I had actually encouraged them to think about for the youth population here in Chicago. I said to them, we actually have an app and many of you probably live in cities where you have this as well. You know, there's these bike sharing programs, right? And you can go on your phone onto the app and you can see how many bikes are available at a particular corner and how many bikes slots are empty in case you need to park your bike. And to me, it always seemed kind of funny to me that we can do that in the public space, right? But we don't have a means of telling youth experiencing homelessness where there's an open shelter bed, right? They have to call each individual shelter to find that out. So I think if A, if we had an app that had that kind of data, I could imagine that we could create a nice chatbot where you could actually query the chatbot, where's the nearest bed for me tonight, right? And then the chatbot could do the search function in the background and then inform the youth, like here's the nearest bed or here's where you can get a hot meal, here's where you can get a shower, right? Those things could be relatively simply automated. The mental health piece I think has to be something that is built more along the lines of what we're trying to do broadly for the youth population. That makes sense. And this is a question, I guess, on the digital climate thing. How do people ever have a problem keeping the phone charged or battery if they don't kind of have a single home or place? Always. Actually, that's one of the biggest challenges for this population. And when we work with these young people, I can tell you when we see these young people in the clinic or we see them at the shelter, often we're meeting with them in the small room or a private space. Every one of them who comes into the room, the first thing I see them do is scan the room for outlets and they'll take their phone out of their pocket and they'll plug it in. They're constantly doing that. It's a big challenge. And so one of the things that we did with Stephen Shuler when he was here at Northwestern in Chicago, Stephen got actually some funding to basically wire the shelter a little bit better. So we got them a better wifi connection in the shelter, but also I think we were able to put in some extra power ports. So that some of the young people could charge their devices. And sometimes that becomes the point of contention, especially in many of you, I don't know. I suspect that this webinar draws in some of the folks who work in the sector, but some of the shelter space I've been in, it's just like, well, pre-COVID, it was all one big communal room sometimes where these young people would be. And there would be like fights over who got to plug in and charge over whom, right? I mean, such a silly thing to fight over, right? And such an addressable thing for us to help these young people with. Related question, right? First you have to power the phone to get it working, but how do you, you kind of briefly alluded, but how do you build trust with kind of these vulnerable youth, right? You have SMI who are homeless, you said they, some kind of said, well, the government's tracking us already, but how did you and the team kind of get people to trust you just to even use these things overall? Yeah, I mean, we tried to just be very transparent with them, John, and really just be very plain speaking about what we were doing and why we were doing it. And that, I think that went a long way. I mean, now look, we were giving them new devices, we were giving them data plans. There were a lot of incentives that were built into the study. So that also, I think, increased their interest in working with us. And they were at least willing to hear us out in terms of the study we're doing and why we were doing it. But transparency was key. I mean, I think we didn't wanna hide anything from them in terms of what we were doing or why we were doing it. And what we found was actually, there was a certain degree of like snowball effect in terms of recruitment. So once we got like one kid enrolled, they would tell their friends and tell some of the other kids who were on the streets with them that, you go and you sign up with this research team, they're gonna give you a phone and they're gonna do this, but you have to do all these measures. And that helped. So we kind of needed a couple young people to come through the gates initially. And then they were able to kind of convince some of the others that they might wanna explore this. So, but transparency is the key, I think. And then I think it was, someone said they weren't sure they misheard you, but in the two studies that you did, the therapy you realized you didn't need the therapist in the second one, or was it that you added the therapist in? Why, either way? Yeah, no, the first one was the one where we had the therapist. And the second one was where we took the therapist out and just had the automated app system. And really it was just because in the first study, we didn't get a lot of engagement from the young people. They liked our therapist. They actually rated her very well and said that they liked her, but they didn't really utilize the time with her. Even though a number of them actually did complete some sessions. They wanted things that were quick. They wanted things that were brief. They didn't wanna have to be doing yet another appointment. I mean, in some ways, these young people are already having to do appointments with case managers and other folks in the course of trying to access services. There are all these gateways, right? And I think for them in our first study, I think the therapist felt like another gateway to them that they had to kind of jump through. So what we realized and why we moved to the fully automated version in the second study was that we stopped trying to do treatment and instead just tried to do engagement. And we thought like by tiering this, we might be able to get further than trying to push everything at once. That makes sense. And it's just interesting looking at different ways you can kind of utilize this technology. So I wonder if we have time for one more last question and then kind of saying, if you were kind of recommending, if people are working with homeless youth, are there kind of some like toolkits or things you would want them to have on their phone or things that you get excited by if they're using? Great question. I mean, I think I was really excited when these young Invincibles developed their app because I think it just seems so relevant and so much speaking to this population. I get kind of also get excited when I see young people connecting to adults in this space, like reliable adults. And I don't care what medium they do that on, whether that's on Facebook or WhatsApp, as long as they have good adults. And sometimes that's a case manager. Sometimes it's an aunt or an uncle or a grandparent. That actually we know is protective and positive. And so, yeah. Yeah. So I guess we'll actually then advance to the next slide because I realized we're almost out of time. So if anyone has follow-up questions on this topic or any related to evidence-based care for SMI, our clinical experts are now available for online consultations. Any mental health clinician can submit a question or receive a response from one of our SMI experts. As always, the consultations are free and confidential. So SMI Advisor is just one of many SAMHSA initiatives that are designed to help clinicians implement evidence-based care. We'd encourage you to explore the resources available on the Mental Health Addiction Prevention TTCs, as well as the National Center of Excellence for Eating Disorders and Suicide Prevention Resource Center. These initiatives cover a broad range of topics from school-based mental health to the opiate epidemic. To claim credit for partaking in today's webinar, you'll need to have met the requisite attendance threshold for your profession. Verification of attendance may take up to about five minutes. You'll then be able to select Next to advance and complete the program evaluation before claiming your credit. Finally, please join us in 2022. It's fun to say it on January 14th as Dr. Deb Pinell presents Understanding Correctional Mental Health Services. This free webinar again will be January 14th, 2022 from 12 to 1 EST, so 9 a.m. PST. Thank you again to our guest, Dr. Karnick. Thank you for all listening. Until next time, take care.

Dr. Niranjan Karnik

webinar

mobile technology

homeless people

serious mental illnesses

SMI Advisor initiative

challenges

precision psychiatry

Stepping Stone studies

interventions