Latest / The Mental Health Evolution / Ep 27: Tiered Care, Technology, and the Future of Mental Health

Transcript

Ep 27: Tiered Care, Technology, and the Future of Mental Health · The Mental Health Evolution

• 0:00
  (Transcribed by TurboScribe. Go Unlimited to remove this message.) This podcast is part of the Frederick Podcast Network.
• 0:03
  Learn more at listenfrederick.com.
• 0:05
  Welcome to Mental Health Evolution, a podcast about what's changing in mental
• 0:10
  health and why it matters.
• 0:12
  I'm your host, Rachel Harrison, inviting you into honest conversations with
• 0:16
  people from all perspectives in the field, clinicians, tech founders,
• 0:20
  investors, insurance companies, and all the folks in between.
• 0:24
  Let's explore what's working, what's not, and what's next.
• 0:34
  Welcome back, everyone, to the Mental Health Evolution podcast, where we
• 0:39
  explore changes happening in the mental health industry and discuss why they
• 0:43
  matter.
• 0:44
  Although we do often have guests, today's episode is a solo discussion.
• 0:49
  I want to share some observations based on a research article that
• 0:54
  I read, as well as some additional articles and questions that have
• 0:58
  been coming up for me as I read more research, talk with
• 1:01
  clinicians, and watch how mental health care is being delivered in real
• 1:05
  time.
• 1:06
  What I think is interesting is there is very clearly a couple
• 1:10
  of different sides of the story here.
• 1:13
  There are people who are thinking about how technology can take over
• 1:19
  mental health, right?
• 1:20
  How do we not even need a person to do this anymore?
• 1:23
  And then there are others who are thinking about how could technology
• 1:30
  even replace anything remotely?
• 1:34
  And so for those people, you know, they're thinking technology should be
• 1:39
  avoided at all costs.
• 1:41
  We don't want that.
• 1:42
  We don't want anything related to that.
• 1:44
  And then there's the middle.
• 1:46
  Then there's how do we integrate these two things?
• 1:49
  How do we look at that?
• 1:50
  And that's kind of what I want to talk to you about
• 1:53
  today is the middle.
• 1:54
  So if you are on either side of this, you may not
• 1:59
  completely agree with a lot of things.
• 2:00
  You may even have questions about what I say, and this is
• 2:03
  more about asking questions.
• 2:05
  I'm not predicting the future.
• 2:07
  I'm not taking a hard stance, but exploring a model that is
• 2:11
  already emerging.
• 2:12
  It's right out there in the research.
• 2:14
  I'll show you the articles that it's in.
• 2:15
  And one that blends technology, human care, and escalation across different
• 2:20
  levels of need and thinking through what that could mean for clients,
• 2:25
  clinicians, and organizations and businesses involved in care.
• 2:30
  So let's get started with some of these articles so that you
• 2:34
  have an idea of where we are launching from.
• 2:37
  This first one is the research article I was talking to you
• 2:40
  about and referencing already in the intro, and it's called The Evolving
• 2:45
  Field of Digital Mental Health.
• 2:47
  And this peer review article outlines how AI and digital tools are
• 2:53
  currently being used across multiple stages of mental health care, from
• 2:58
  prevention to post-treatment monitoring.
• 3:01
  And this is from PMC published this.
• 3:03
  And I am actually going to look at these different stages and
• 3:11
  some of the things that are happening in each stage of treatment.
• 3:14
  So we're kind of using this article as a way to sort
• 3:17
  of structure what our conversation is going to be about today.
• 3:21
  The next article is from the APA, and if you haven't noticed
• 3:24
  a trend, I just want to give out a shout out to
• 3:27
  the APA, the American Psychological Association.
• 3:29
  In so many of our episodes, we are using things that they
• 3:33
  are producing and writing because they are standing out right now as
• 3:38
  a body that is actually providing guidance and publishing guidance on all
• 3:44
  kinds of things.
• 3:45
  But this article in particular says health advisory on AI chatbots and
• 3:51
  wellness apps.
• 3:53
  And this article looks at where AI-based tools may be helpful
• 3:57
  and where limitations, risks and ethical concerns remain.
• 4:01
  I really appreciate that they are coming from a clinical standpoint and
• 4:06
  looking at what could be useful and also what we need to
• 4:10
  be aware of.
• 4:11
  And then this third article is the first published chatbot trial from
• 4:18
  Dartmouth.
• 4:19
  And I think we talked about this in another episode, but it
• 4:22
  yielded some positive results.
• 4:25
  So this is looking at a therapy chatbot and showing research shows
• 4:31
  that there are some improvements.
• 4:32
  And as we all know, this field is a field that is
• 4:37
  based on research.
• 4:38
  All clinicians want to do things that are shown to be effective,
• 4:43
  that are evidence-based.
• 4:44
  So I think we have to look at it when we have
• 4:47
  some of these.
• 4:48
  And I think we have to look at it critically.
• 4:50
  We have to ask the question of, OK, great.
• 4:52
  It showed a benefit.
• 4:54
  And are there any holes in this research?
• 4:58
  What do we need to be concerned about?
• 4:59
  So looking at it from all angles.
• 5:01
  And I really encourage you to do that for yourselves.
• 5:04
  The next article is called AI is providing emotional support for employees,
• 5:10
  but is it a valuable tool or a privacy threat?
• 5:14
  And this explores a lot of the workplace use of AI support
• 5:18
  tools and the tension between access, effectiveness and privacy.
• 5:24
  And I think privacy is something that, well, I was having a
• 5:28
  conversation last night with another clinician and they were saying, listen,
• 5:32
  nobody cares about privacy.
• 5:34
  Nobody cares that tech companies sell people's data as a means to
• 5:39
  make more money.
• 5:41
  Nobody cares.
• 5:42
  I don't know.
• 5:42
  Is that true?
• 5:43
  I care.
• 5:44
  I think it's important.
• 5:46
  But I think it is possible that we're pretty desensitized to people
• 5:50
  selling our data.
• 5:52
  We have data breaches where our data is out there in who
• 5:56
  knows where.
• 5:57
  And I think it's become kind of commonplace.
• 6:00
  So maybe people don't care.
• 6:01
  Maybe they do.
• 6:02
  But there are some things that we need to think about when
• 6:05
  it comes to privacy.
• 6:06
  And in the clinical mental health field, privacy has always been a
• 6:11
  number one concern.
• 6:13
  It's our ethics.
• 6:15
  Confidentiality is number one.
• 6:18
  The first thing that a clinician says to a client is that
• 6:22
  everything we talk about is confidential with these exceptions.
• 6:27
  The exceptions are regarding safety, child abuse, things like that, that are
• 6:32
  reportable.
• 6:33
  And that's about it.
• 6:34
  Everything else is, by our ethics, supposed to remain confidential.
• 6:41
  So it's an interesting question.
• 6:44
  Is that shifting?
• 6:46
  Are those ethics shifting?
• 6:48
  Maybe, maybe not.
• 6:49
  The next article is called AI Mental Health Tools Breakthrough or Band
• 6:55
  -Aid.
• 6:56
  Again, examining whether these tools are meaningfully expanding access or are
• 7:02
  they a risk at becoming substitutes for care.
• 7:06
  When these systems are understrained, I think, you know, it's also
• 7:10
  affordability piece.
• 7:12
  If someone, if a child has free access to AI to talk
• 7:18
  about their thoughts and feelings, or let's use an adult, an employee,
• 7:22
  is that going to be where they turn rather than someone who
• 7:27
  is trained and licensed and following a set of ethics?
• 7:30
  I think those are really good questions to ask.
• 7:33
  And lastly, from clinical judgment to machine learning, this article looks at
• 7:39
  how AI is beginning to influence clinical decision-making and what that
• 7:44
  may mean for professional roles.
• 7:46
  And that's from the Society for Psychotherapy.
• 7:49
  So a really interesting perspective of where AI is supporting clinical judgment
• 7:56
  or rethinking how we're doing decision-making with AI.
• 8:00
  So let's dig into some of these articles and some of the
• 8:06
  distinct stages of treatment.
• 8:08
  So across many of the recent articles on digital mental health and
• 8:13
  AI, I do see this theme that breaks into these distinct stages
• 8:17
  of treatment with each different treatment and goals.
• 8:21
  So for example, one of the supporting articles, I outlined the very
• 8:25
  first one up there, the PMC article.
• 8:27
  It talks about several different stages.
• 8:31
  So I'm going to list those stages.
• 8:34
  And then I'd like to kind of break into what each one
• 8:38
  would potentially utilize for technology.
• 8:43
  So the first is pretreatment and screening.
• 8:47
  This is interesting because if you think about different things like
• 8:52
  pretreatment, to me, pretreatment and screening may be different or may be
• 8:56
  the same.
• 8:58
  It sort of indicates that this is somebody who is interested in
• 9:03
  getting treatment.
• 9:04
  They've identified that they want treatment.
• 9:06
  And this stage may be the stage where they are just thinking
• 9:11
  about what can support them until they get to see the clinician.
• 9:15
  What kind of data is helpful for them to gather before the
• 9:19
  clinician sees them?
• 9:21
  And a lot of these things really can be utilizing technology.
• 9:27
  So things like self-referral tools, symptom checkers, triage systems, risk
• 9:34
  screening.
• 9:35
  Think about the apps that do things like that already.
• 9:39
  Think about the benefits for a client, maybe logging their mood even
• 9:43
  before they come into therapy.
• 9:46
  Maybe trying out a few different skills or maybe a mindfulness app
• 9:52
  as they are waiting to get into care.
• 9:55
  These are things such that could be used with technology.
• 9:59
  I think also about there are lots of online ways to do
• 10:04
  evaluations, screening for depression, screening for anxiety.
• 10:08
  All of that could be done before treatment starts.
• 10:12
  Potentially, there are some risks to consider.
• 10:15
  Like, what if somebody has a negative reaction to those screenings?
• 10:19
  What if they think, oh my gosh, this means I'm depressed and
• 10:22
  that makes me even more depressed or suicidal.
• 10:24
  I mean, there are some safety risks and things to consider here
• 10:27
  for sure.
• 10:29
  But making that process more automated, utilizing technology could be really
• 10:35
  useful, both for clinicians and for clients.
• 10:38
  The next step that is outlined here is treatment, meaning the actual
• 10:45
  treatment with a licensed mental health clinician.
• 10:49
  And how does that use technology?
• 10:53
  The biggest piece I see here, there are a couple things.
• 10:58
  There are certainly tools out there about recording sessions and doing notes
• 11:03
  for you that has become pretty popular in the industry.
• 11:09
  There are also things like, what about pharmacology?
• 11:13
  How do we integrate AI with pharmacology?
• 11:16
  How do we do combined treatment approaches?
• 11:19
  How does decision or support decision making tools or support help?
• 11:25
  So I personally still don't love the idea of recording a session.
• 11:30
  I think notes are treatment planning and conceptualization.
• 11:34
  And I think there are also privacy concerns to consider.
• 11:38
  But despite AI writing our note for us or creating a treatment
• 11:43
  plan for us, what if it can help us think through case
• 11:48
  conceptualization?
• 11:49
  What if it can find nuances?
• 11:52
  What if it can track treatment goals in more detail and perhaps
• 11:56
  aid us in a more robust conceptualization?
• 12:00
  This would have to be specifically created with models in mind that
• 12:05
  include like the DSM and other knowledge accessible to make sound clinical
• 12:10
  suggestions.
• 12:12
  And I think the key here would be utilizing any of those
• 12:17
  support systems would have to be evaluated and reviewed by a human
• 12:23
  -trained licensed clinician before moving in any particular direction.
• 12:28
  One of the things I fear the most with AI tools that
• 12:33
  are presented to help clinical ease, which great, clinical ease is great.
• 12:38
  And if we ever lose the part where a clinician reviews the
• 12:44
  notes, reviews the treatment plan, reviews the diagnoses, reviews the
• 12:49
  suggestions, I think we're going to see a lot of problems.
• 12:54
  I think the reality is there might be some of the legwork
• 12:59
  that is done for us, but if we don't review it as
• 13:02
  a clinician ourselves, then we are not and make edits, make adjustments.
• 13:07
  We are not serving our clients very well.
• 13:11
  So those are a few of the pieces potentially with the treatment
• 13:15
  phase.
• 13:16
  Now, also in this article, they talked about post-treatment and monitoring.
• 13:22
  So there's, again, some interesting ideas here, like AI tools used for
• 13:27
  follow-up care, symptom monitoring, risk assessment, relapse prevention, or
• 13:33
  treatment adjustment after formal therapy ends.
• 13:36
  So this would be considering it a phase where formal therapy is
• 13:40
  no longer happening.
• 13:42
  But what if there's a check-in?
• 13:44
  What if there could be an assessment automatically given three months after
• 13:50
  treatment to see where somebody is, to make suggestions for what might
• 13:55
  be supportive if they are struggling after three months, three months after
• 13:59
  treatment, or if they are doing well?
• 14:02
  You know, kind of looking at those pieces could be a way
• 14:06
  that technology could assist.
• 14:09
  The caveat here for me is that any kind of safety access
• 14:13
  is a must-have.
• 14:14
  There must be a way for any technology system to access a
• 14:20
  live person who can help if there's a safety emergency.
• 14:24
  Anytime we are asking technology to ask someone questions about their mental
• 14:30
  health, that safety planning piece, I believe, absolutely needs to be in
• 14:35
  place.
• 14:36
  That is one of the biggest gaps that I see currently.
• 14:40
  And then we have general support and prevention as a place where
• 14:46
  technology is being used.
• 14:47
  This is already pretty commonplace, I would say.
• 14:51
  These would be standalone tools that are designed to maintain well-being,
• 14:56
  reduce stress, or present mental health problems in a non-clinical or
• 15:01
  community population.
• 15:03
  And similar to the pretreatment, these already exist, right?
• 15:07
  They're being used.
• 15:08
  Think of apps that help people rate their mood every day.
• 15:13
  Think of apps that help people with sleep hygiene.
• 15:16
  Think of apps that help people do mindfulness or meditation every day.
• 15:23
  These are things that all support wellness, and they are standalone in
• 15:28
  their just providing a service, a supportiveness to wellness.
• 15:33
  And the last area that the article highlighted was clinical education.
• 15:38
  Using AI tools to train, assess, or upscale mental health professionals,
• 15:43
  clinical students, or educators.
• 15:45
  And I think this is an interesting question to ask ourselves.
• 15:49
  How can we train clinicians using technology?
• 15:52
  Perhaps practice sessions where AI is our client, or perhaps we can
• 15:57
  have a database for clinical education that allows clinicians to access it
• 16:02
  when they need to.
• 16:04
  If I need to ask about a particular symptom of OCD, for
• 16:08
  example, and what the best method is to treat that, is there
• 16:12
  an education tool that technology can give me?
• 16:16
  I can, of course, Google it.
• 16:18
  I could ask ChatGPT about it.
• 16:21
  Do I get accurate enough information?
• 16:24
  I think that would need to be the piece that we are
• 16:29
  really making sure that it has sound clinical education, not just something
• 16:34
  that is searched on Google.
• 16:37
  That would maybe need to be a big difference.
• 16:40
  It kind of makes me think about the idea that if any
• 16:43
  of you are part of like a membership group, for example, we
• 16:46
  have the EMDR circle that is a database and training videos all
• 16:50
  about EMDR to help clinicians when they're having stuck points or questions,
• 16:55
  and to be available 24-7.
• 16:57
  But this might be a step beyond that of taking that kind
• 17:01
  of sound clinical data and having it accessible with a question, right?
• 17:07
  With one question, and then maybe a video comes up for you.
• 17:10
  And it's like, this is a verified clinical appropriate technique that you
• 17:16
  could use in this situation.
• 17:17
  I don't know.
• 17:18
  Interesting to think through.
• 17:19
  And looking at all these stages together, I think it's possible that
• 17:23
  we're moving toward a tier or stepped model of mental health care.
• 17:27
  In fact, I know that we have a guest coming up that's
• 17:30
  going to be talking a little bit about this.
• 17:32
  But in this kind of model, many people might first encounter support
• 17:36
  through these low-intensity tools, thinking about the basic ones I talked
• 17:41
  about, apps for sleep, stress, meditation, those kind of things.
• 17:46
  And if those aren't sufficient for someone, then they could escalate.
• 17:51
  Maybe it escalates to coaching.
• 17:54
  Maybe it escalates to group therapy.
• 17:56
  Then it escalates to individual therapy, then to specialized care.
• 18:00
  And when necessary, to outpatient or inpatient, or excuse me.
• 18:05
  And when necessary, it escalates to inpatient or higher levels of care.
• 18:11
  I think this model serves to improve access.
• 18:15
  If you think about it, it's a lot of what happens in
• 18:18
  an emergency room.
• 18:20
  There's a triage.
• 18:21
  There are levels of care.
• 18:22
  Getting people to the right care at the right time.
• 18:26
  It could certainly help with provider shortages as well.
• 18:30
  It's already implemented in some places.
• 18:33
  And like I said, we're going to talk to an insurance provider
• 18:36
  that utilizes that, especially for their employee offerings.
• 18:42
  But I think it's interesting what it could mean for different stakeholders
• 18:46
  in the ecosystem here.
• 18:48
  So for clients, there are clear benefits to having earlier access and
• 18:53
  fewer barriers to entry.
• 18:56
  They might not be waiting for an appointment or a person.
• 18:59
  They may have some tools that can help them early on.
• 19:03
  At the same time, there's a risk of bottlenecks if pathways to
• 19:07
  higher levels of care are unclear or delayed or underfunded.
• 19:13
  For clinicians, this raises important questions about professional roles and
• 19:17
  identity.
• 19:18
  Where do generalists fit in to a system where technology may address
• 19:23
  some of the lower-level needs?
• 19:26
  If you're a clinician that specializes in teaching clients how to manage
• 19:31
  stress, can this be done with an app or a chatbot?
• 19:36
  What skills are still necessary for that in-person relationship-based
• 19:41
  therapist?
• 19:42
  And what specialties are needed as the tiered approach happens as higher
• 19:49
  levels of care are needed?
• 19:51
  For businesses and systems, I think technology companies may benefit from being
• 19:56
  embedded earlier in the care process.
• 20:00
  Traditional providers may see changes or declines in referrals if some needs
• 20:04
  are resolved through technology.
• 20:07
  At the same time, specialized services could see an increase in demand,
• 20:12
  potentially even a surge if escalation pathways are well-designed and
• 20:16
  functioning as intended.
• 20:19
  So, overall, this episode aims to just begin to spark the ideas,
• 20:25
  the questions about these emerging models.
• 20:28
  They are all over the research.
• 20:30
  If you dig into these articles, you're going to see this is
• 20:33
  how a lot of tech companies, this is how a lot of
• 20:37
  treatment providers, this is how a lot of university systems are starting
• 20:41
  to conceptualize what care would look like.
• 20:45
  It's reshaping clinical work, potentially referral patterns, and the mental
• 20:50
  health care ecosystem as a whole.
• 20:53
  So, I wanted to present these really as a starting point for
• 20:59
  you to think about in whichever part of this ecosystem you're in.
• 21:04
  And what does this mean for you if you start thinking about
• 21:08
  this differently than maybe the vantage point that you're sitting in today?
• 21:13
  As always, thank you for listening to today's episode.
• 21:17
  And I hope that exploring this developmental model in mental health care
• 21:22
  raises some questions for you.
• 21:24
  And we'd love to hear from you, always.
• 21:27
  So, take a look at those articles.
• 21:30
  Let us know what you think.
• 21:31
  And we will be back next week.
• 21:34
  Thanks for listening.
• 21:35
  Bye for now.
• 21:37
  Thanks for listening to the Mental Health Evolution podcast.
• 21:41
  Join the conversation.
• 21:42
  Find us on social, send us your thoughts, and tell us what
• 21:46
  you're seeing in your corner of the mental health world.
• 21:49
  Until next time, stay curious and keep evolving.