Latest / The Medical Mind / APA Innovation Hour: Asynchronous Approaches to Mental Health Care with Dr. Peter Yellowlees

Transcript

APA Innovation Hour: Asynchronous Approaches to Mental Health Care with Dr. Peter Yellowlees · The Medical Mind

• 0:03
  Hello everyone and welcome to the American Psychiatric Association's Medical Minds, The Innovation R.
• 0:09
  My name is Manu Sharma and I'll be your host for today's episode.
• 0:13
  I'm the Assistant Medical Director for Research and Innovation at the Institute of Living, an Assistant Clinical Professor of Psychiatry at Yale School of Medicine, and the Chair on
• 0:22
  the Committee of Innovation for the American Psychiatric Association.
• 0:26
  The main thesis of The Innovation R is that technology and artificial intelligence
• 0:31
  are poised to transform the way we practice psychiatry over the next decade.
• 0:35
  We want psychiatrists to be educated and prepared for these changes.
• 0:39
  In these episodes, we will talk to innovation leaders from across the world who will give us a preview of how today's technology will change the future of psychiatric practice.
• 0:50
  For our inaugural episode, I'm thrilled to be speaking with Dr.
• 0:53
  Peter Yelilis.
• 0:54
  Dr.
• 0:54
  Yelilis is the CEO of Async Health.
• 0:57
  a telemedicine company he co-founded to implement and promote the practice of asynchronous telepsychiatry.
• 1:04
  He's also a distinguished emeritus professor of psychiatry at UC Davis, where he directed the fellowship program in clinician well-being.
• 1:13
  Previously, he was also the inaugural chief wellness officer at UC Davis Health and held the Alan Stoudemire Endowed Well-being Chair in Psychiatry.
• 1:21
  He was the president of the American Telemedicine Association from 2017 to 2018.
• 1:26
  and he's written over 250 academic papers and book chapters, seven books, and has presented over 180 video editorials for psychiatry medscape.
• 1:38
  Welcome to the podcast, Peter.
• 1:40
  Good, well thank you very much, Manu.
• 1:41
  It's real pleasure to be here.
• 1:43
  Absolutely.
• 1:44
  So today we'll be talking about how asynchronous data collection can improve psychiatric clinical practice, right?
• 1:50
  So can we start by maybe you giving us a little bit of history and introduction of what an asynchronous telepsychiatry or asynchronous data collection is.
• 2:00
  Certainly, mean, this goes back a long, long way.
• 2:03
  And I'm going to use two patients as to illustrate really how my journey has moved in terms of this particular area.
• 2:12
  mean, asynchronous telepsychiatry is actually very simple.
• 2:15
  It's really just a matter of doing an interview with a patient which you record and then sending that recording, essentially a recording of a patient's mental state as well as
• 2:24
  their background history to an expert for them to review in delayed time.
• 2:30
  So often called store-and-forward, but I think nowadays we're talking about asynchronous telepsychiatry.
• 2:37
  And of course, the big difference between the telepsychiatry that most of us do most of the time is that, you know, telepsychiatry is traditionally real time and we're just
• 2:47
  interviewing a patient without any recording going on.
• 2:51
  Now, the first patient I want to mention that really made me interested in asynchronous telepsychiatry was actually the first patient I ever saw
• 2:59
  on video back in 1991 in Australia.
• 3:03
  This was an Aboriginal woman who was a very pleasant lady who had a major depressive disorder.
• 3:11
  And at the end of our interview, she asked me if I'd actually recorded the interview because she told me that she wanted to show the whole recording to her community because
• 3:22
  she'd never been on TV before and she had no concerns about privacy.
• 3:27
  Now I was regretfully had to tell her that I hadn't recorded it because our legal advice at the time was that we shouldn't do that.
• 3:35
  But it struck me immediately that really we were missing a huge opportunity here as psychiatrists.
• 3:42
  We were seeing all these patients, know, meeting with them, observing them on a video environment.
• 3:50
  And yet we weren't actually recording their mental state.
• 3:52
  And later on, we would just simply write the mental state down in the traditional way.
• 3:57
  So for more than 30 years, I've been thinking about how can we best start recording patients and then using those recordings to give a more objective view of patients' mental
• 4:09
  state and potentially to compare patients over time.
• 4:13
  So so forward 18 years, and it's 2009, I've moved from Australia to America.
• 4:19
  And I have a long-standing patient, an elderly female who had bipolar disorder.
• 4:26
  that I'd been seeing for several years.
• 4:28
  She and her husband were retired and they spent six months of the year in Hawaii every year away from California where I was working.
• 4:39
  Now, the husband was a very techie guy.
• 4:41
  This was the early days of the iPhone.
• 4:44
  And when they were in Hawaii, she became depressed.
• 4:49
  And so he decided that he would just take her out onto the lanai of the...
• 4:55
  They were staying at and he'd make he'd get her to do a recording for me of how she felt Which is what he did.
• 5:02
  So he actually then texted me a five-minute video of her on her lanai Describing classical features of a major depressive disorder as part of her bipolar illness I was very
• 5:15
  impressed by this looked at the recording which in those days wasn't actually HIPAA compliant.
• 5:20
  It just came on the text
• 5:23
  But I was able to go into our EMR and send a series of changed prescriptions to her in Hawaii.
• 5:34
  And she changed her medication and then they followed up actually on text and email with me.
• 5:41
  Again, this was in early days of electronic communication from a healthcare perspective.
• 5:48
  And she recovered beautifully.
• 5:50
  And so this was actually the first asynchronous video consultation I ever did.
• 5:55
  And it worked really well and it was driven by a patient's need.
• 6:00
  And I think that to me is really the biggest argument for continuing on with asynchronous consultations.
• 6:07
  We're really saying now, look, video is data.
• 6:10
  It's not just something that we observe.
• 6:14
  We need to start recording video and using recorded video.
• 6:19
  in an appropriate sort of security environment and with all the patient's rights and privacy issues taken care of.
• 6:28
  But we need to start recording videos of patients so that we can actually do better assessments and also so that we can start using the AI tools that are now available for
• 6:41
  facial recognition, movement recognition, language assessment, and similar.
• 6:47
  Because if we don't start recording patients, we're going to miss out on all this very valuable data and on a better way of actually helping patients.
• 6:57
  So from 2009 onwards, I've been basically fascinated by this approach to asynchronous consultations, making use of video as data and collecting that.
• 7:09
  at UC Davis, we've done a series of consultations.
• 7:13
  and a series of trials supported by over $7 million worth of mainly NIH funding to demonstrate that this is feasible, both in primary care and in nursing homes, that
• 7:25
  patients and providers both like it, and that you get at least as good clinical outcomes as you do doing conventional telepsychiatry consultations.
• 7:35
  So we've really now got a strong evidence base for the use of video as data.
• 7:39
  And I think now in the post-pandemic world where
• 7:43
  Most psychiatrists are now doing at least 30 % of their consultations on video.
• 7:48
  Attitudes have changed professionally.
• 7:51
  I think the next big change in professional attitudes is going to be this issue of let's start actually collecting video as data and then using it with the amazing tools that have
• 8:04
  become available.
• 8:07
  Wonderful.
• 8:07
  So just so that we can visualize this a little bit for our listeners.
• 8:14
  So the way asynchronous technologies or asynchronous telepsychiatry would work is a patient in the comfort of their own home or wherever they're at in a private place would
• 8:26
  record an interview, which the psychiatrist might have access to later on or some summary of it.
• 8:35
  And they would
• 8:37
  use that instead of a live one-on-one interaction with the patient.
• 8:42
  Is that how we imagine it?
• 8:43
  That's exactly right.
• 8:45
  And the interviewer can be either live with the patient, maybe in a primary care clinic or in a nursing home, or can be beamed in on one of many video consultation platforms
• 8:56
  available now, and can be in real time or can be in delayed time.
• 9:02
  And obviously with AI, what we're now doing, of course, is using interviewers who are actually highly realistic avatars.
• 9:11
  so that you can actually have these interviews done at any time without the patient even needing to make an appointment to be interviewed.
• 9:19
  They can literally be sent a link.
• 9:21
  So it dramatically improves patient access.
• 9:24
  So there are a whole series of ways of recording patients and then using those recordings for good clinical purposes.
• 9:33
  I think another way of thinking of it is really it's a very souped up form of a curbside consultation.
• 9:40
  We've been doing curbside consultations as psychiatrists in all our careers.
• 9:46
  I worked in consultation liaison psychiatry for many years and was totally used to physicians grabbing me in the corridor, asking me briefly about a patient.
• 9:55
  I'd give them some opinions and they would then go off and, you know, change the approach to that patient and then maybe talk to me in the coffee room a week later about what what
• 10:06
  was, they to do next?
• 10:09
  And so
• 10:09
  We've been doing curbside consultations forever.
• 10:13
  They're actually normal.
• 10:15
  The beauty of having curbside consultations plus a video of the patient is that we actually have a lot better data to do our consultations on.
• 10:24
  The other area we've been working on regularly is with the training of residents.
• 10:30
  Residents now routinely record patients, whether it's for psychotherapy.
• 10:36
  review and assessments or for intakes in particular.
• 10:40
  And then they show those recordings to their supervisors and their supervisors discuss the patients with them.
• 10:48
  And if they're in an intake clinic, perhaps then go and see the patients as well, either in person or on video.
• 10:54
  But we've been doing, again, asynchronous recordings for psychiatric training for many years and in most training programs in this country.
• 11:06
  So it seems to be very bizarre, quite honestly, that if we're prepared to use asynchronous videos for training of our residents, that we don't use them for normal clinical work.
• 11:19
  Yeah, so it almost sounds like it could be revolutionary to kind of improve access, right?
• 11:25
  Where you can have non-psychiatrists, but trained professionals who might be able to conduct an interview at a remote location and then transmit it to a trained psychiatrist
• 11:37
  who might help the clinical decision making, right?
• 11:40
  So there are natural applications in collaborative care models, in rural medicine, and
• 11:48
  And of course with AI, I would love to hear like how would you see the transition between an actual human being to an AI doing these interviews and what has your experience been?
• 11:58
  Sure.
• 11:59
  Well, I think first of all, I mean, in our research at UC Davis over the last decade or so, we've used mainly social workers and nurses to do the interviews.
• 12:08
  And we've worked out semi-structured interviews that they give.
• 12:14
  And the interviews that they do are then sent off to a psychiatrist.
• 12:18
  One of the beauties of this, of course, is it actually makes the psychiatrist work much more efficiently.
• 12:24
  And we've found that most psychiatrists can do an intake assessment now in about half an hour instead of an hour.
• 12:31
  And we've actually published a paper on the cost effectiveness of that.
• 12:36
  So that one of the beauties of asynchronous approaches is that you can actually scale the number of patients that you see.
• 12:43
  Now, to my mind, the biggest problem facing the psychiatric profession right now is actually the workforce one.
• 12:50
  It's the simple fact that we've got too many patients and too few psychiatrists.
• 12:56
  And traditional telepsychiatry doesn't scale.
• 12:59
  You just see the same number of patients, but at a distance instead of immediately in front of you.
• 13:06
  Asynchronous psychiatry can actually allow you to scale the number of patients that you see and you can review and assess and treat more patients using asynchronous approaches.
• 13:17
  Now, having said that,
• 13:19
  these asynchronous approaches, we have to be careful with them.
• 13:22
  I'm not promoting them for therapy, for instance.
• 13:25
  I'm really just saying, hey, we should be using these for assessments and for monitoring and reviewing how patients are doing over time.
• 13:34
  But clearly, I don't think they're appropriate for therapy.
• 13:37
  So they have their place.
• 13:39
  And if I had to predict, I'd say that within the next five to 10 years, probably about at least 10 %
• 13:46
  of all of the consultations that we do in psychiatry will be asynchronous.
• 13:53
  Now that's assuming that reimbursement models and legal changes occur and they're obviously both very difficult to get.
• 14:02
  But I think if we assume that this scaling issue with asynchronous telepsychiatry allows more patients to be seen, that's a very strong argument for changing our current models of
• 14:14
  payment and similar.
• 14:18
  Now you mentioned also the AI side of things.
• 14:20
  I mean, I can talk about that for a very long time if you wish.
• 14:24
  But I mean, what we've found using highly realistic avatars to do the interviews is two main things or three main things.
• 14:35
  First of all, we've seen a lot of young people and for them this is just normal.
• 14:43
  We actually make sure that the avatars always declare themselves as avatars so that it's very clear to the patient they're not talking to a real person because they're so
• 14:52
  realistic.
• 14:54
  But our response has actually been very good so far in terms of the patient experience.
• 15:01
  And one of the beauties of the use of avatars is that you can be absolutely sure what questions have been asked.
• 15:08
  You have perfect fidelity for the interviews and you aren't necessarily sure about that.
• 15:12
  with human interviewers.
• 15:15
  So I think that's a big advantage.
• 15:17
  The other big advantage of avatars is that there's an important DEI issue here.
• 15:23
  You can have avatars of any racial or social background.
• 15:29
  And so you can potentially have patients choosing the type of avatar they'd like to interact with that actually makes them feel most comfortable and that looks and sounds
• 15:39
  most like them.
• 15:40
  And that's a huge deal for psychiatry, as everybody knows.
• 15:47
  So this is a really important issue around not just patient access, because you don't need an appointment, obviously, to see an avatar, you just get sent a link and you can do the
• 15:55
  assessment any time.
• 15:58
  But also because of patient engagement.
• 16:02
  So I think access and engagement are two big issues for our profession to look at.
• 16:06
  And this asynchronous approach is one off.
• 16:10
  a number of answers for those issues.
• 16:16
  Interesting.
• 16:16
  So traditionally what some the data collection of the interview, which might have been done by say a nurse or a social worker or trained mental health professional now can be
• 16:27
  done by a digital avatar that uses artificial intelligence technology to kind of have a realistic conversation with the patient and does that interview for us.
• 16:38
  Now, what happens once the interview is done?
• 16:40
  In your current applications or experience, do we send the raw video to the psychiatrist?
• 16:46
  Do we do some post-processing there?
• 16:48
  And what are some options or ideas that can further enhance psychiatric care as we record and analyze the data that we collect?
• 17:00
  I think that's a great question.
• 17:01
  mean, what we're doing at Async Health is basically adding AI to the videos that we record.
• 17:08
  So once the videos of a patient's answers has been recorded, we then use AI to transcribe it so that we get a perfect written document as to what they've said.
• 17:20
  The AI also then will summarize the major points in the history for us.
• 17:30
  and we'll put that into the language that you prefer.
• 17:33
  And then using both, we can automatically make PDFs of this summary and of the patient transcription, which of course is a major part of the actual medical note.
• 17:46
  So we actually do a lot of a documentation for the psychiatrist.
• 17:51
  So when the psychiatrist goes in to look at the video, they can obviously look at the video and see any part and we can speed the video up and slow it down and.
• 17:59
  and show particular points that they might be interested in it.
• 18:02
  But they also have a lot of their documentation actually already done, ready to be transferred to their electronic medical record.
• 18:10
  Now, obviously, we want the psychiatrist to very carefully to review what the patient has said and to ensure that the AI transcription is accurate and that the summary is accurate
• 18:22
  and to...
• 18:23
  add to the summary of their own diagnosis and treatment plan.
• 18:26
  I mean, we're not attempting to diagnose the patient, but we are attempting to write summaries.
• 18:31
  And then, so that obviously speeds up the documentation side for the psychiatrist.
• 18:38
  I mean, if you think about what we actually do in a consultation, there's only really three things we do.
• 18:43
  We collect data and we document that data.
• 18:46
  Then we do some analysis, which is making a diagnosis, and then we do the treatment planning.
• 18:53
  Actually, the things that we need our skills for are the analysis and the treatment planning, and yet we spend most of our time doing the data collection and the
• 19:01
  documentation.
• 19:02
  And so that's what we're using the AI for to make that component, you know, the majority of the time that psychiatrists do on intake and monitoring interviews, to make that more
• 19:15
  efficient so that we save a significant time for the psychiatrists.
• 19:19
  Now, at this point,
• 19:21
  You know, we are looking at, you know, how can we also use voice and facial recognition systems?
• 19:28
  How can we use language analysis?
• 19:31
  How can we analyze the tone of the voice and movement analysis?
• 19:35
  A whole series of other approaches to add data that the psychiatrist will be able to see when they go in to see this initial data collection.
• 19:47
  Some of that's possible fairly soon.
• 19:50
  Some of it will take longer.
• 19:52
  But I guess my personal view is that we have to see this approach to care as being essentially the development of decision support tools.
• 20:02
  I don't see these as being tools that will automatically do all of the work that psychiatrists do.
• 20:09
  I think it's very important that we remain in charge as a profession.
• 20:14
  and that we actually make the diagnosis, we write the treatment plans, we, know, we sure we can take advice on that from various electronic areas.
• 20:22
  But really, I think it's very important that we create an environment whereby we can use our skills in the best way for the analysis and treatment planning and obviously to
• 20:35
  potentially follow up with the patient, either in person or online as well and check any points we're not sure about, et cetera, et cetera.
• 20:44
  So I don't see what we're doing right now as being an algorithmically driven sort of diagnostic approach.
• 20:53
  What we're doing is trying to provide much better data for a psychiatrist to do their job really well.
• 21:01
  Definitely.
• 21:02
  do you are you bullish or optimistic about so there's a lot of research that is ongoing and you see like papers every day that document how we can use speech and language samples
• 21:12
  or use some advanced algorithms to evaluate facial expressions to try and identify some objective markers, which might help us tell a little bit about the severity of depression
• 21:25
  or other things.
• 21:27
  How optimistic or bullish are you?
• 21:29
  that we might actually be able to include all of that as a part of this process.
• 21:35
  I'm extremely confident that will happen.
• 21:38
  I mean, there are already, you know, several companies that I know that have spent a lot of time looking at voice tone, for instance.
• 21:47
  And these companies, if you give them literally 20 to 30 seconds of the patient's voice tone, they can be as accurate in terms of whether the patient has depression or anxiety as
• 22:01
  the PHQ-9 and the GAD-7.
• 22:04
  So these questionnaires, which we give regularly to patients as screening for these disorders, I think in future will be unnecessary.
• 22:16
  So we'll be able to have a patient do their interview and not only will we have the video, we'll have the transcription, we'll have a lot of the documentation written, but we'll
• 22:28
  also have an assessment of their anger.
• 22:31
  their anxiety, their depression, PTSD, from the use of other tools that are analyzing things like their tone and their language.
• 22:42
  I think that's great because that's actually going to make it easier and more accurate for psychiatrists to come to the correct conclusions.
• 22:50
  Excellent.
• 22:50
  So what has the real world experience been of the psychiatrist maybe who have tried this, the patients who've been on the other side accessing the service?
• 23:00
  What have they told you?
• 23:01
  What has been that their anecdotal evidence?
• 23:04
  And if there's any large trials that you might have done that kind of point us to some evidence.
• 23:10
  Sure.
• 23:11
  Well, we've published numerous papers and we've done three clinical trials.
• 23:15
  None of them using AI.
• 23:17
  These are trials where we use nurses and social workers as the interviewers.
• 23:21
  And in terms of both patient and provider satisfaction, we had very high rates.
• 23:29
  So with the asynchronous approach, we actually had 84 % of the patients saying that they were comfortable and satisfied with the approach.
• 23:41
  which is a very significant number if you consider what the number of patients are who are satisfied with the average psychiatric interview.
• 23:52
  So now that was slightly less than the synchronous telepsychiatry numbers that we got from that particular clinical trial, but it's still very high.
• 24:02
  We haven't done formal assessments of patient and provider satisfaction with the AI avatars.
• 24:10
  But we're actually just starting those right now.
• 24:13
  And we are obviously assessing every patient that we see as to their satisfaction level.
• 24:20
  What we've found though, because we've used this with a lot of other people, friends and family, students, interns that we are employing, we've found that in fact, there's been
• 24:33
  remarkably little resistance to this approach.
• 24:36
  And that patients, and that these non-patient
• 24:40
  groups have been very comfortable answering questions from avatars.
• 24:45
  So I don't really think there's going to be a major problem, quite honestly, because the main issue that the patients have is actually understanding that the avatars are avatars
• 24:54
  and they're not real humans because they're so realistic.
• 24:57
  And we actually have every every avatar actually introduces themselves as an artificial intelligence avatar to be very clear to the patient that the patient is not talking to a
• 25:06
  real human.
• 25:10
  Now, I might, like as our listeners are thinking about this one, even as I thought about this, I kind of try and go back to some of my psychiatric training where we kind of put a
• 25:23
  lot of importance in the patient-doctor relationship.
• 25:26
  So as you engage in this asynchronous data collection, what feedback have you received around the patient-doctor relationship or the importance of this relationship on treatment
• 25:37
  outcomes perhaps?
• 25:39
  and what has your experience been?
• 25:41
  So we found that in primary care, again, without the AI avatars, with the human interviewers, we found actually that obviously the psychiatrist didn't have a relationship
• 25:56
  with the patient.
• 25:58
  And we were working in a collaborative care environment there where the primary care physician was actually the primary treating person.
• 26:04
  And actually we followed up the primary care docs.
• 26:08
  And first of all, they were very comfortable with what we were doing.
• 26:11
  and liked it and kept referring patients to the study.
• 26:15
  But secondly, they actually implemented just as many decisions that the psychiatrist suggested they make for the asynchronous consultations as well as the synchronous ones.
• 26:29
  So at a practical level, they took as much notice of the psychiatrist, whether the psychiatrist was seeing a patient asynchronously or synchronously.
• 26:38
  I think that's a very important point.
• 26:41
  Now, in terms of the psychiatrist now and how do you integrate this sort of software into a practice?
• 26:48
  I mean, I think the reality is most people who use this software will have it integrated as part of their practice.
• 26:55
  If they want to be the actual avatar interview, so the interviewer looks like the psychiatrist who runs the practice, then that's fine.
• 27:02
  We can do that.
• 27:04
  So the patients can actually be interviewed in the practice of Dr.
• 27:08
  X by Dr.
• 27:09
  X.
• 27:10
  even though that Dr.
• 27:11
  is the avatar.
• 27:13
  And then they can have an appointment with Dr.
• 27:15
  X and see them either online or in person, you know, a few days later or whenever it's relevant so that Dr.
• 27:24
  X can follow up with them and can check any concerns they've got about the interview, add any questions they want to ask, you know, get to know the patient better.
• 27:34
  So I think the best place for these consultations is integrated into a system of care.
• 27:40
  whether it's in a private practice to see patients when they're initially referred and really collect much better data than we typically might ever get from a referral note.
• 27:51
  Or whether it's in a collaborative care environment where you've got a team process going or in places where it's hard to get psychiatrists such as in nursing homes in particular.
• 28:03
  So I think all of these
• 28:06
  sorts of issues have to be taken into account.
• 28:08
  And asynchronous care is not a panacea.
• 28:11
  I'm not suggesting it at all for every interview.
• 28:14
  And as I say, I think probably it'll actually eventually be around somewhere around 10 % of all psychiatric interviews, but that's still a very large number.
• 28:23
  And that still increases substantially the number of patients who can be treated and assisted through this process.
• 28:32
  I agree 100 % and as you were answering the thought that came to my mind was often times we like to unfairly to some extent compare newer technologies with the ideal psychiatric
• 28:45
  interaction.
• 28:47
  But I feel that in that, although we should aspire to that, but at the same time we have to realize in that in reality there's a vast number of individuals who have no access to
• 28:59
  psychiatric care.
• 29:00
  or very limited access to psychiatric care.
• 29:03
  And I feel that if whatever we can do to kind of try and cover the gap between the number of people who need care and number of people who end up getting care, I feel we'll be
• 29:13
  doing very good.
• 29:15
  And as you said, even with 10%, the number will be very high.
• 29:19
  Having said that, do you feel that there might need to be lots of regulatory changes to allow large scale adoption of something like this?
• 29:27
  And if so, and if someone
• 29:29
  if some regulators are listening, what would you want to tell them?
• 29:34
  Yeah, I don't think it have to be major regulatory changes.
• 29:37
  I mean, obviously, you know, we do everything in accordance with HIPAA.
• 29:43
  So you need to make sure that whatever environment is being used for asynchronous consultations, it's fully HIPAA compliant.
• 29:50
  So I think that's the major issue.
• 29:54
  Obviously, it'll be interesting to see how the sort of cross state boundaries
• 30:02
  issue gets sorted out in the long run.
• 30:07
  Currently, you know, I think in reality, you'd need to do these consultations within the same state that the physician is licensed in.
• 30:16
  But hopefully, that can be broadened considerably and obviously just as it is with the VA, for instance, and Indian Health Services.
• 30:27
  But I think the
• 30:30
  the challenges are going to be less on that side.
• 30:34
  instance, California is a good example.
• 30:38
  California recognizes asynchronous medicine as being the same as synchronous medicine and actually passed a bill in 2019 defining asynchronous care.
• 30:50
  So they both count the same in terms of regulations and reimbursement in California.
• 30:57
  That's not the case for all states.
• 31:00
  And this is a state by state thing because a lot of healthcare is state driven, obviously, as we both know.
• 31:07
  But, you know, I think those things will gradually change.
• 31:11
  I think the reimbursement issues is going to be interesting.
• 31:14
  mean, for instance, right now, it's perfectly legal to do this sort of consultation within 24 hours of actually seeing the patient and to then, you know,
• 31:25
  charge for the time that it takes a psychiatrist to review the data that has been collected by the avatars, just as if they were phoning somebody to get more information
• 31:36
  about a patient they'd seen.
• 31:37
  So you can charge E codes, I think, for that time currently.
• 31:44
  And obviously, as you well know, and particularly in private practice or in employer-funded systems, there's a lot more flexibility as to how you can charge
• 31:55
  for these particular systems.
• 31:58
  So I think there's a lot of opportunities, but like anything new, know, regulations and reimbursement tend to follow behind innovation.
• 32:06
  And that's going to be the case here, unfortunately.
• 32:14
  Can you think of some other specialties or other industries that have successfully used asynchronous models and they've gone from being something that were at the fringe to
• 32:25
  somewhat of a mainstay in those industries and that we can learn from?
• 32:30
  Completely.
• 32:30
  I mean, think that's a really great question.
• 32:33
  I mean, first of all, in medicine, if you look at the specialties, dermatology, ophthalmology, radiology, cardiology, and pathology are all disciplines that have moved a
• 32:45
  lot of their work into an asynchronous process.
• 32:49
  And where these specialists are now staying at home and are working asynchronously.
• 32:56
  And that's become the standard of care.
• 32:58
  in many of those disciplines.
• 33:03
  If you look at other industries, then of course, mean, can you imagine a major TV broadcaster without the ability to show interviews that happened yesterday or five minutes
• 33:16
  ago?
• 33:17
  I mean, the news is completely dependent on asynchronous videos.
• 33:24
  So the whole of the sort of communications industry has moved.
• 33:29
  The retail industry has done the same, and particularly in terms of advertising and marketing.
• 33:38
  HR is doing the same, and I'm sure many people listening to this have been interviewed on video for jobs, and the whole HR industry is moving into an asynchronous set of processes.
• 33:52
  And of course, the big one is buying things, know, the retail industry.
• 33:57
  You know, if you go to Amazon or an equivalent, now you feel confident about putting your credit card into their site.
• 34:09
  But you now expect that you are not just going to buy something, but you're going to be told where it is in transit.
• 34:17
  You're going to get a photograph of it when it's delivered.
• 34:20
  and you're going to get likely a follow-up survey as to how efficient that whole process was.
• 34:27
  And then within a few days, your actual preferences for future buys will suddenly start changing and you'll be directed towards more of what you've just bought.
• 34:41
  So the retail industry has made dramatic changes in how they actually sell.
• 34:49
  and has moved from a traditional synchronous model of you go into a shop, you talk to somebody and you buy something off the shelf, which is what we used to do, to this new
• 35:00
  hybrid model, basically, where you can buy things either online or in the shop and it all combines together.
• 35:06
  And in healthcare, think we've got to start going in that direction.
• 35:09
  I mean, we need to give more choice for our patients as well as better access.
• 35:17
  And I think to me, the retail industry is actually the best example that we should be following.
• 35:23
  And asynchronous video consultations are just one of many choices that I think we'll be able to give to patients in the future.
• 35:31
  And it's an exciting world.
• 35:32
  think we're going to see dramatic changes, especially with the generation of psychiatrists who've been trained during COVID.
• 35:42
  As you well know, these are psychiatrists now
• 35:46
  who were often trained that the primary way to see patients was on video rather than in person.
• 35:53
  And a substantial number of them want to continue to work like that.
• 35:57
  They've discovered how effective that is and how good it is for their own lifestyle and the way they can integrate work and their life demands.
• 36:09
  And so these asynchronous approaches, again, give the psychiatrist even more flexibility.
• 36:15
  because you can set aside particular times of the week when you want to do these asynchronous consultations, or you can do them in time that say a patient hasn't showed
• 36:28
  for.
• 36:29
  Whereas, and that's lost time, certainly in private practice for many people.
• 36:35
  So it should be possible for a psychiatrist to have less unpaid time in those sort of environments.
• 36:44
  So it's going to be fascinating and I think we've got to really look carefully at these other industries and see how they've done things well and how we can do things in
• 36:52
  healthcare better.
• 36:53
  Absolutely.
• 36:53
  think that there are two themes that emerge from your answer.
• 36:56
  So one was on the side of the human experience receiving care.
• 37:00
  So as an individual, as I want to see a psychiatrist, what's my experience right now and how can we make that better using this technology?
• 37:08
  But on the other hand, how's the psychiatrist experience as I see my patients day to day, how will my life change if I try and include such technologies in my practice?
• 37:19
  So we'd love to hear like what your experience has been there and what your thoughts are.
• 37:23
  specifically.
• 37:24
  think that's a really important issue.
• 37:26
  mean, as you said in the introduction, I used to be a chief wellness officer.
• 37:30
  And a lot of that time that I spent in that role was actually trying to make life more efficient for physicians, not just the psychiatrist, but all physicians.
• 37:43
  So trying to get rid of what's called pajama time, the time after hours when we're catching up on notes and things like that.
• 37:51
  And I think the beauty of
• 37:53
  these both asynchronous consultations and the use of AI for documentation in particular, is that it should be able to dramatically cut down the amount of time that we spend doing
• 38:03
  documentation.
• 38:05
  And that's a huge advantage to us, not only in that we can actually use that time for other things, whether it's seeing other patients or seeing our kids at school.
• 38:16
  And so I think the process
• 38:20
  of asynchronous consultations is likely to free up time and make our schedules more flexible.
• 38:28
  Again, because they're not dependent on us as psychiatrists having to make timed appointments to see a patient.
• 38:34
  We can do this whenever we've got some free time.
• 38:38
  Just as the same for patients.
• 38:41
  And patients also need better access.
• 38:44
  They don't necessarily want to have to drive to our clinics to sit in our waiting rooms.
• 38:51
  and to effectively spend at least half a day from whatever they're doing in their normal lives to come and see us in person.
• 38:59
  Now, I personally actually think that these asynchronous consultations are best seen in the context of hybrid care, whereby, you know, we continue to see people in person, and I
• 39:09
  think that's always going to be the gold standard.
• 39:12
  But then we also see them electronically, both synchronously and asynchronously.
• 39:18
  And we've been talking mainly about video consultations asynchronously today.
• 39:22
  But of course, know, text and secure email and other approaches to consultations are, you know, equally reasonable as long as they're carried out in a secure environment.
• 39:37
  Definitely.
• 39:39
  I've heard you talk about this in the past where you talk about this multimodal hybrid approach, which will be the future.
• 39:46
  So it might be a combination of in-person visits, video visits, telephone calls, but every interaction is considered an opportunity almost to collect data and collect data, is
• 39:59
  beyond the traditional way we think about things where somebody's telling us about their symptoms, but actually using some objective markers.
• 40:07
  Tell us a little bit more about that and how you imagine that in the future.
• 40:12
  Well, think a lot of people now work in a hybrid way, again, post pandemic.
• 40:19
  really, certainly I started working in this hybrid way back in probably at least 10 years ago in my normal outpatient clinic.
• 40:28
  And I just offered every patient the option of coming to see me in person or on video.
• 40:35
  And I gave them all my email address.
• 40:42
  and obviously the fact that I was available through MyChart, which was the patient portal.
• 40:50
  And I made it clear that I didn't particularly want them to phone me because phoning is incredibly inefficient of time and tends to interrupt everything and you want to avoid
• 41:00
  that if you can.
• 41:02
  So I practiced for, you know, since 2013, essentially in this hybrid manner and
• 41:09
  And what I found over that period of time is that, you know, a lot of the patients still wanted to come and see me in person occasionally.
• 41:17
  A few continued in-person consults all the time, but the great majority shifted to video and were very comfortable just sending me secure email or secure messages in between times
• 41:32
  as necessary.
• 41:36
  Having said that,
• 41:37
  those patients who were seeing me mainly on video, if I wanted to see them in person for a particular reason, and there are sometimes those reasons from a clinical point of view, or
• 41:48
  if they wanted to see me in person, that was totally fine.
• 41:52
  And so it was really a choice that we both had to make.
• 41:56
  And I found that that actually worked extremely well and was much more convenient for the patients in particular.
• 42:06
  but also for me.
• 42:10
  so I think working in that hybrid way, all we're doing really is adding on an extra choice for the patients.
• 42:17
  And one of the extra choices we did at UC Davis was to actually introduce e-consults, which many people now do, whereby a primary care doc simply sends you a note and asks you
• 42:29
  to review an EMR set of notes and send them some opinions about, particularly about medication management.
• 42:36
  And so we started doing a lot of those and they could be done at a particular time of the week that you chose or when you just had some free time.
• 42:44
  And I think these asynchronous video consultations are very like the e-consults in many respects, except that you also get a video of the patient and now we get all of this added
• 42:55
  value to the video that we're using AI for.
• 43:00
  So if there are leaders listening to this episode and they want to try and implement some of this, maybe even pilot the model, are there resources that are available online or
• 43:14
  consultants maybe like they can approach or talk to?
• 43:17
  Absolutely, they can talk to you for sure.
• 43:19
  But like maybe how they can get in touch with you or if there are online resources that are available for.
• 43:25
  Yeah, I mean, they can obviously go to our website and there's a whole lot of resources on there.
• 43:29
  All of the papers and the evidence we've based to back this up are available through the Async Health website.
• 43:38
  We're implementing this in a number of clinics currently, and we're very interested in implementing this more widely.
• 43:46
  And so I think there's a lot of information out there.
• 43:51
  I think this is just
• 43:53
  This is only one way, obviously, of improving our practices.
• 43:56
  are many other approaches.
• 43:58
  And this will obviously change over time.
• 44:00
  But I'd certainly encourage people to contact me in particular for more information.
• 44:06
  And I'd love to follow up with people and see what the options are and see if, in fact, the type of practice they run would be a good practice to have asynchronous video
• 44:18
  consultations integrated within it.
• 44:23
  And so, I mean, there are some practices where this is a very obvious thing to do.
• 44:27
  And I'm thinking particularly of things like student health practices.
• 44:32
  And we're talking to several student health systems at the moment because students have a very busy life.
• 44:38
  They don't necessarily like going to see a psychiatrist, but they also have a lot of psychiatric problems and they're going through a difficult developmental phase of their
• 44:48
  lives.
• 44:50
  And so the potential, for instance, for a student to be able to do an assessment at 2 a.m.
• 44:55
  in their dorm room and have that submitted and then reviewed by a psychiatrist and then a decision made about whatever follow-up, I think is a really fantastic extra service to
• 45:07
  offer.
• 45:08
  So there are certain areas where it's a really obvious thing to do.
• 45:12
  And you've already mentioned collaborative care is another one.
• 45:15
  So I think any collaborative care system could be enhanced by these approaches.
• 45:19
  Private practice is another one whereby you're essentially getting really good information about your new referrals much more than you would normally get from almost any other
• 45:29
  referral source.
• 45:31
  But we're still learning, you know, and I think there's going to be a whole series of other really good use cases.
• 45:38
  This has been an excellent conversation.
• 45:40
  If there are some take-home messages or concluding messages that you want people listening to remember from this conversation, what would that be?
• 45:50
  I think the first take home message is that patients actually really like to be seen on video.
• 45:59
  We haven't had difficulties with recording videos in the hundreds of patients we've seen using this process.
• 46:08
  There's a strong evidence base to support this.
• 46:13
  The people who are most challenged by this approach of recording
• 46:19
  patients are actually psychiatrists and other mental health professionals and they're the people who just as with synchronous telepsychiatry before COVID were the people who were
• 46:30
  most sort of critical of the process and least likely to take it up.
• 46:35
  That's changed now.
• 46:36
  And so I think this is the next step.
• 46:41
  My view is very simple about asynchronous telepsychiatry, which is really what we're talking about.
• 46:48
  I think it's going to become common.
• 46:49
  I think it's going to be a major component of the way we all work within a decade.
• 46:56
  I think that will happen because of patient demand, just as it has happened with synchronous telepsychiatry.
• 47:04
  And I think it's inevitable that this is going to happen because of the changes in other industries and other medical disciplines that have already happened.
• 47:15
  And we've seen how some of our professional systems can be changed by these approaches.
• 47:24
  And I think actually in the long run, this will provide us with much better patient care and will reduce the workforce issue that we have at the moment, the major workforce
• 47:34
  problems with too many patients and too few psychiatrists.
• 47:38
  Absolutely.
• 47:39
  I think the biggest take home message for me from today was that we have access to all these tools and this can be one more tool in our tool chest where we can use this to
• 47:48
  improve not only the lives of our patients, but our lives as psychiatrists as well.
• 47:52
  Peter has been wonderful talking to you.
• 47:54
  Thank you so much.
• 47:56
  And I hope we can continue this conversation at other avenues as well.
• 48:00
  I think this is extremely educational.
• 48:01
  Thank you for all the work you're doing.
• 48:03
  Thank you very much indeed.
• 48:04
  It's an exciting time and I really encourage everybody to sort of review whether this is an option for them.
• 48:14
  Have a good day everyone.