How Telehealth Can Improve Patient Care and Outcomes

How Telehealth Can Improve Patient Care and Outcomes


We’re ready to go. OK so we’ll start with a little
bit time but will we’ll make it up and hopefully we’ll have
plenty of time for discussion afterwards. So welcome very much to the
session. Very excited about this
particular area. And I’m a senior adviser within
health care delivery and Disparities Research Program at
PCORI and I also serve as a role sort of providing sort of broad
expertise and subject matter expertise in the area of
telehealth. And so I’m very excited to have
the panel that’s joining me today. The focus of this annual meeting
is on evidence to impact and we’re very excited that two of
our presenters today are going to be presenting the results of
their studies that are pretty much hot off the press. So Dr. April Armstrong’s study
was just published in JAMA Open Science about two weeks ago. Is that right? Two weeks ago. And Dr. draw bends of study was
published in psychiatric services in October. Is that right. May. Oh OK. Well almost hot off the press. OK. And so two of the panelists that
are presenting now Dr. Armstrong she’s associate dean for
clinical research at the University of Southern
California. And then Dr. Drew Urban is the
Professor of Psychiatry and Behavioral Sciences and also a
co-director of the Center for Behavioral Research and
Technology and Engineering at the University of Washington. And then in addition though to
thinking about these great results both of them have very
promising results for patients. But in addition to hearing about
that we also want to learn from stakeholders about what is
needed to actually get these put into practice. And so how do we get to that
impact stage and to discuss their thoughts on this matter. We have three people
representing diverse stakeholder perspectives. Dr. Mark Ishaq who was also a co
investigator on Dr. bends of study. He provides the perspective of a
health system administrator. He serves as the CEO of
thresholds which is one of the largest and oldest behavioral
health system providers in the state of Illinois and was also
the setting for Dr. Benz of study. In addition to that we have
providing the viewpoint of practicing clinicians. We have Sylvia Tokyo. She’s a senior legislative
counsel at the American Medical Association with an expertise in
regulatory and legislative issues in telemedicine. And then finally representing
the perspective of state health policymakers. We have Jennifer rec. She’s a project director at the
National Academy for State Health Policy where she helps
advance evidence based policy and using Patient Centered
Outcomes Research. Perfect. There’s our panel right there. So I’m going to give a brief
overview of the our telehealth research and investment in the
quarry. And then I’m going to hand it
off to Dr. Armstrong and then doctor and Zev supported by Mark
Shugg to talk a little bit about the results in his setting and
then we’re going to turn it over to our stakeholders for
reaction. What is going to need what’s
needed to actually put these into into practice so we have a
very large portfolio in the area of telehealth and I think
there’s a lot of good reasons for that. This is really reshaping the
health care delivery system very rapidly. Just this this is a sly that
just gives a basic image of some of the rapid change here. This is in telehealth tell a
mental health visits in the Medicare program but over a 10
year period the number of tele mental health visits and
Medicare grew by 14 fold. I read in 2015 Kaiser Permanente
about half of their visits were virtual visits and they were
anticipating by 2018 which is this year that the number of
virtual visits would actually surpass the number of inpatient
visits. So we have invested over three
hundred million dollars to fund 78 studies in this area and a
lot of that is driven by investigator interests. So a lot of this came from
investigators that are were submitting applications to our
what we call our broad portfolio. What I show up here is sort of
these concentric circles in telehealth and health and
telemedicine. I’m not going to get into the
details of the definitions there but basically one thing that I
did want to point out is that when you think about telehealth
a lot of times thing people think traditionally back to the
consultative model that works with a specialist in a private
patient in in a clinic through interactive video. And the largest part of our
portfolio is actually in health meeting the patient where they
need when they need it. So mobile health accounts for
about two thirds of our portfolio. But that said that’s not to
lessen the importance of the investment that we’ve made in
the area of telemedicine. Some of our larger and longer
studies that cross multiple states are in the area of
telemedicine. And so we’re very excited about
those studies that we’ve funded there. We think that telehealth also
one of the reasons why we’ve gotten so much interest from the
investigator community in this area is it’s a very fruitful
area for Patient Centered Outcomes Research. First of all when we’re talking
about personal health personal care capturing individuals
preferences and autonomy and needs the interface through
telehealth has the great potential to do that and through
potentially cultural tailoring of the interface one can better
reach low health literacy populations limited English
proficiency. Also this is an area where there
is a lot of need for comparative effectiveness research. There are lots of studies out
there there’s lots of systematic reviews there’s been a lot of
research done in this area but not a lot of head to head trials
that say should we use for example this mobile health app
versus that mobile health app that can really help with both
clinician and patient choices. And finally it’s one of the
barriers that people have talked about in terms of using
telehealth has been that there is like a disconnect between
clinicians and the interface or a disconnect between the end
user the patient and the interface. And so potentially by engaging
patients and other stakeholders in the development of this
interface from the beginning and in the implementation we can
better integrate telehealth into the workflow and so it’s more
acceptable to patients at the end of the study. Our portfolio spans a broad
array of conditions. I think this is also the fact
that telehealth is becoming ubiquitous. While there’s a lot of studies
that we have concentrated in the area of telling mental health as
well as in diabetes and management of chronic diseases
we also have some areas where looking at telehealth for
screening we’re looking at telehealth for managing rare
diseases. The we have classified our our
telehealth portfolio by purpose of the of the studies and
similar as I said before in terms of the focus on mobile
health and self-management of chronic disease. A lot of this is really related
to promoting self-management and on the educational piece. But one thing that I did want to
point out too is that another unique aspect of our investment
in telehealth is that a large proportion almost half of our
studies are focused on addressing disparities. So this literature the sort of
digital divide has raised a lot. We have a lot of studies that
are working to try to bridge that in the area of scene
disparities we have addressed racial and ethnic minorities of
the low income persons with disability and also as I
mentioned low literacy among these other priority populations
for that technology platform. A large number of our cities I
do use the mobile help mobile phone or the tablet but we also
have a few studies that are talking about store and forward
technology in April’s screen to talk about that we also are
looking at on outcomes that are important to patients and it’s
interesting when we were looking at systematic reviews this is
something that was called out a lot. A lot of studies have looked at
acceptability but haven’t really looked at the clinical outcomes
that are important to patients or those outcomes that are
important to their health and well-being. And obviously being a quarry
that’s something that’s important to all of our studies. When we did look at the
systematic reviews in this area some recent systematic reviews
we categorized some of the evidence gaps and we feel like
our portfolio is really addressing a lot of these quite
well. So we have some in as I
mentioned a lot of them are in patient stakeholder engagement. About 60 percent are head to
head trials. And while you know we do have
this focus on addressing disparities to a smaller
proportion of looking at culturally tailored
interventions. But there are some important
ones there. So in conclusion I think that
our portfolio really has the potential to impact the field
advance knowledge in this particular area because of the
unique aspects of what we’ve invested in. And I’m going to without further
ado hand this over to Dr. Armstrong. Thank you. Penny and thank you for inviting
me here to speak today. I’m going to talk about how we
could use telehealth to improve specialty care delivery in
chronic skin diseases I have no relevant conflict of interest
specific to this presentation. Well first I’m gonna discuss
some evidence gap and then I will provide a study overview of
the study that was funded by P. Corey and then highlight some of
the study’s findings and then finally explore what’s needed to
change practice. So first evidence gap. If I could buy a show of hands
ask how many of you have found it difficult at times to get to
see your dermatologist O K. So most of the room have yes. In fact the waiting time for a
dermatologist is among the highest among the specialty
providers. And I often have patients will
come to my clinic and they will want a doctor. I’m trying to try to see you but
the wait is three months so I’m here right now. But my rash is gone. I originally called you for and
I say oh well that makes that easy visit. But the true question is how can
we be responsive to our patients. The moment that they really
needed us. My husband’s grandfather my
husband’s pushing 50 now and his grandfather is one hundred and
seven years old and I asked him and I said gosh grandpa you know
how many doctors do you see you must see a lot of doctors. He says No I only see three
doctors. My primary care doctor my
cardiologist and surprisingly my dermatologist I say really. You see a dermatologist that
those are one of the three physicians you see. He said yep. And by the way can you look at
this. So. And skin as you know a
fascinating organ. It’s actually the largest organ
in our body if you line up all the blood vessels in your skin
end end it will be around eleven miles. And many of you see a
dermatologist because skin is visible and skin can can show
different pathology. I’m not here to convince you of
the importance of Dermatology but rather use that as an
example of showing how specialty care is really important and how
we can improve access to specialty care. So one of the key gaps in terms
of accessing specialist is that it really depends on where you
live. Studies after studies have shown
that if you live in rural areas your access to specialty care is
often limited your outcome not only for dermatologist care but
also even including mortality can be affected by where you
live. We know that regular access to
dermatology care if you don’t have that you’ll have poor
outcomes. These are pictures of patients
with psoriasis. I think while the awareness of
psoriasis was low many years ago these days it’s difficult to
actually probably go through your evening program without
seeing a commercial for psoriasis medications. But the key thing is that a lot
of patients are still untreated or under treated. And part of that is access. So patients not only present
with these with these red plaques but also a third of them
develop psoriatic arthritis and patients with severe psoriasis
also are at higher risk for cardiovascular adverse outcomes. So when we think about the
traditional in-person model we can think about in terms of the
patient. Typically it goes to the primary
care doctor and then depending on that primary care doctors
decision making the patient can be referred to the dermatologist
or the relationship can exist with the PSP now with
traditional asynchronous tele dermatology and what I mean by
asynchronous tele dermatology is that your store and forward tele
dermatology that the images are captured but then they’re sent
to the specialist at a certain time and the specialist looks at
these images at a different time. So in the traditional
asynchronous tele dermatology the patient actually has to find
a piece p a primary care provider who’s interested in
doing telemedicine and then that PSP would take the photos and
send that to a dermatologist. And one would then look at these
photos render their opinion bring those opinions back to the
PSP. Now there’s a lot of problems
with this model. Number one is that the vision
has defined a peacekeeper who is interested in doing telemedicine
and that’s actually more challenging than what is
imagined. The primary care physician also
has to be flexible in terms of looking at dermatologist
recommendations and knowing how to convey that to the patient. The other part that we found is
that patients like to understand the recommendations from the
specialists what if they have what if they have questions
about the recommendations who will then be able to answer
their questions. So in this model the PSP often
plays a middle man the third model is the collaborative
connected health model which is the online model we actually
tested in our study. This model originated from the
idea that if the technology is good enough can we create an
environment where anything that can be done in person could be
done online. So with that we like the patient
to have direct contact with a dermatologist. PSP could have direct contact
with the dermatologists and the PCV can have direct contact with
the patient. And I’ll go go into that model
in a little bit detail later on. So here is a gap in terms of
thinking about the different models to care for our patients
with skin diseases and then finally something I think Penny
has alluded to earlier as a lack of validated instruments to
really measure our outcomes in telehealth research and
especially Intel dermatology I can probably count on both of my
hands. The number of studies that have
used validated instruments things that instruments that are
created on the fly typically has a lot of problems. For example there are studies
that used whether to ask patients whether they have
improved or whether they have worsened such dynamic
instruments are problematic bigger because it depends on
patients recall what if they can’t really recall to a precise
degree how they did. So I’m gonna tell you a little
bit about my study and and then first I’m going to tell you
about the intervention we tried to test. Then I’m going to go over the
study design so the intervention is that such intervention is
such that we have a telemedicine provisional platform. If your psoriasis patient you
could go to your PSP through this online platform go being
going there virtually you can log on. You can submit your images and
you can answer a set of structured questions that’s
intended to elicit how you’re doing and then your PSP can
converse with you on this platform render recommendations
and prescribe medications through that platform. Now the PSP can also communicate
with a dermatologist via that platform can do it by one of two
ways. Number one is that if they want
to keep primary responsibility of the patient they can do a
console heatedly much like the arm that I talked about earlier
in terms of traditional telemedicine or two they can
actually transfer care of the patient to the dermatologist
once the patient is in the care of the dermatologist. The patient can also visit the
dermatologist virtually online just as the wedding person and
they can ask questions. They can submit their images
their history as they would so here’s an example of an online
encounter you can have the PCV or the psoriasis patient who
then they have instructions. I’m taking these standardized
images. They can be uploaded onto a
health connected health Web site and then the dermatologist can
evaluate these images make recommendations and prescribe
for the patients. So our study type is a pragmatic
randomized controlled equivalency trial. I want to emphasize this is
equivalency trial. It may seem like a minor point
but in study design it’s actually what we consider a
pretty major equipment trial is designed to test intervention a
is equivalent to intervention B and we intentionally designed it
this way is because we want to be able to understand is it
exactly the same. We have three specific aims. The first aim is to determine
whether this on my model would result in equivalent improvement
in psoriasis disease severity using validated outcome
instruments. Number two we wanted to know if
this online intervention would result in equivalent improvement
in quality of life as compared to in-person care. And then finally we wanted to
know whether this intervention would result in better access to
specialists. So here’s the trial design. We recruited patients with
psoriasis and they were randomized into two arms in one
arm is the in-person arm essentially standard care. They can go to their primary
care doctors or their dermatologists. However they usually carry out
their care and then they are followed over the course of a
year. In the other arm they would have
these interactions online in the way that I had described
earlier. So the study was year long
because this is a chronic disease and we wanted to know
how patients did. Over time this visit frequency
however is determined by the patient and doctors interactions
specifically because what if you have a severe patient that you
needed to see more often than a patient with a mild psoriasis
that you didn’t need to see that often. We want that piece to be
pragmatic just as it would in real life. Now while the visit frequencies
are determined individual apps for the patient we collected the
outcome as structured intervals. This is because we want to make
sure that the data collection was uniform across the two arms
so here’s a baseline patient demographics. I just wanted to point out a few
things about 34 percent of our patients were Hispanic and this
is twice that of the national average. So nationally 17 percent of
patients are Hispanic or Latino in origin and also that these
patients had experienced various types therapies before. So they spanned the disease
spectrum. Importantly the sites war
northern California southern California and Colorado and
hopefully with this multi centered approach that are that
our findings are more generalizable. So first I’m gonna show you one
of the primary outcome which is a change in what’s called the
posse score positive score is nothing more than a granular way
of scoring how the patients are doing. And this is we take their
psoriasis plaques and then patients look at their air thema
the redness thickness and the scaling as well as the body
surface area that’s involved that’s tracked by both arms over
time. So as you can see. So the red bars here represents
the online arm and the blue circles represent the in-person
in-person arm. What we’re looking for is if
there’s differences over time between the red bars and the
blue bars of these vertical bars and there’s a measure in which
we do so on the lower right hand corner you will see that we have
for this study Defining equivalence margin. And what we want to see is that
if the two interventions were equivalent then would the
difference reside within this equivalence margin. In other words if you have
intervention a online intervention in the in-person
intervention. And how do you define what is
same. What is what is sameness or
equivalence matters. So you define that a priori
based on clinically significant differences based on both groups
you define that sameness first and then you do your study and
to see if the two are actually the same if they were the same
they will fall within that equivalence margin and if any
part of that bar goes beyond that means that they’re not the
same. So as you can see in terms of
Psoriasis Severity there were no differences with regards to the
improvement in Psoriasis Severity in both arms. What about body surface area
over time. Here is another measure of
crisis severity. Same thing happens as you can
see here. There are no differences between
the two arms. How about patient global
assessment. This is patients assessment of
how they’re doing in terms of their psoriasis overall what you
will notice is that interestingly while most of the
confidence interval is within the pre specified margin it is
actually outside of the margin for one area actually in the
favor of the online group. That means that the patients
overall in the online group actually thought there two races
did just a little bit better than that in terms in the
in-person group. What about improvements in
quality of life. So we measure that through to to
validate the instruments. One is what’s called Skin Dex. It looks at patients quality of
life with regards to their emotional health and as well as
their functional health. And what we find is that as you
can see here again it’s within the equivalence margin that
there were no differences in the quality of life between the two
groups. How about dermatology live
quality index another outcome measure that looks at quality of
life just making sure that using different validated measures
that we still find the same thing as you can see again. There were no measurable
differences between the two groups. In fact the two groups in terms
of the improvement in quality of life was the same. How about access to care to
specialists. We measured two things. Number one how far this patients
had to travel. Number two how long they had to
wait in the waiting room. So the result here is not earth
shattering. On the left hand side is the
in-person group and the right hand side is the online group
because it’s a pragmatic model. Online patients if they want it
to go to the input go have input in-person visits. They could. And we track that and out of the
out of 100 149 patients that were randomized to the online
group over the course of year there were only eight in-person
visits. So that is quite telling. I was surprised about that
because I thought everyone would just rush to someone in person
but as you can see here there were a significant difference as
would be expected between the total distance traveled between
the in-person online group and as well as in office waiting
time between the in-person and online group. We also did qualitative
assessments. We asked patients regarding
their feedback. As you can see many of them had
positive feedback regarding this intervention. I think while some more
apprehensive in the beginning over the course of the year they
did get used to it. So considerations for
implementation are a few and I’ll make my comments brief now
because I know later we’ll have a wonderful panel. We’ll also focus on this. I think technology is central to
the dissemination having a scalable and sustainable
reimbursement model. It’s key. And finally we want to consider
medical legal considerations. That’s one of the key
considerations for the providers. So we want to be sensitive to
that. And no for workforce
considerations. We can clone dermatologists at
least not yet but we want to see how technology can make best use
of their time through various means. So in conclusion the online
model produce equivalent outcomes in regards to disease
severity as well as quality of life and that prevent providers
superior access to care. I think that this is one example
of patients centeredness instead of having patient go to the
location of care. We are trying to bring care to
the patients. And then finally I hope that you
will see this model has the potential of being
transformative and potentially applicable to other areas of
medicine. You can learn all this more at
the Pickering Web site. Our primary outcome outcomes
were published in JAMA network open. And then our second day outcomes
our quality life is accepted by Journal of Investigative
Dermatology a top tier impact factor journal in the
dermatology world and our third aim has also been published in
telemedicine and e-health and we have a fourth manuscript that’s
being submitted. So it takes a village. There were three center three
major centers and so I’m indebted to all these people who
helped carry out this project and my many thanks to quarry and
for penny for spending the last few years with me. And I want to thank you for your
attention thanks very much I’m going to turn it over to Dr.
Bender. Hi everyone. Let the record show that I had a
jacket on those spotlights are super hot up here. And so I took it off but no sign
of disrespect. I wasn’t wearing that tie though
with text. So let’s start with the
disclosure. Quick disclosure on my part my
colleague Mark doesn’t have anything to disclose and we’ll
jump right into it before we talk about serious mental
illness briefly. Let’s put some of the
conversation about mobile health into context. We are all alive to see the
fastest adoption of technology and recorded human history. All right. So all of you I’m sure remember
a time where there were no cell phones and there were no
smartphones. The U.N. estimates that back in
2001 there were approximately 1 billion mobile cellular
subscriptions worldwide. The estimate at the end of 2017
is that the number exceeds 7 billion. That of course doesn’t mean that
every single person on the planet has a mobile phone
companies have more than one license. Some people have more than one
device. But by and large it means that
the vast majority of the adult population on the planet has
access to a mobile device and mobile cellular infrastructure
that allows us to reach them in real time and in real place. Or for them to reach us or
resources that are on the device in real time in real place. So from a health care
perspective this is a game changing opportunity. And I can’t emphasize that
enough. There are more people on the
planet with access to a mobile phone than there are to a
working toilet or running electricity and your
smartphones. Most of you judging by the
demographics of this audience most of you have smartphones
your smartphones have greater computational capacities the
NASA had when they sent the Apollo missions to the moon and
that’s sitting in your pocket right now. So hopefully that helps
contextualize the territory that we’re in it’s a really really
exciting time. Back to serious mental illness. So when we talk about serious
mental illnesses we’re really referring to severe psychiatric
illnesses like schizophrenia schizoaffective disorder bipolar
disorder or major depressive disorder where there is also
severe functional impairment that is the core aspect of of
serious mental illnesses. And we’re roughly talking about
4 percent of the world’s population that have a
disproportionate impact on all levels in terms of days lost to
disability in terms of distress in terms of our visits. So they are highly represented
and untreated unmet need untreated serious mental illness
has really dire consequences. So if you have an SMI and you
don’t get care you are dramatically more likely to have
a shorter lifespan to be the victim of abuse to be homeless
to attempt suicide to succeed with that suicide attempt and to
be incarcerated. So it is a major major public
health concern. But while we think of serious
mental illnesses as chronic they are not necessarily static. That means that there are
periods of recovery where people even in the context of
persistent symptoms do rather well. People move between these
periods of psychiatric symptom exacerbation and perhaps
hospitalization and relapse through periods where people are
quite high functioning. And really we know now that
illness management interventions can really lengthen and increase
these recovery periods. And that’s really the backdrop
for the interventions that we tested in this study. So in this management
interventions are typically deployed in a context of a
clinic setting an academic medical center or a brick and
mortar clinic. And of course there are a host
of challenges and this type of deployment right distance to the
clinic transportation hours and schedules of the clinic. If you have a serious mental
illness and you may struggle with persecutory ideations
sitting in a room with other patients and other people might
not be on the top of your to do list that day. The stigma that is associated
both with the label of a mental illness or coming to services at
a clinic is pretty significant to this population. And of course there’s an issue
of quality of care. The majority of people with SMI
get services from individuals who are not necessarily doctoral
level trained psychiatrist and psychologists people who do
their best but may have a bachelor’s level education or
less or perhaps very little experience and background in
prompting services. And at the community mental
health center most most people get services the turnover
because these are highly under-resourced environments. The turnover is great. And so it’s very difficult to
establish a high level of expertise in delivery of illness
management strategies OK so that’s where a mobile health
from the health may come in. We mentioned that mobile phones
are ubiquitous. This is also true for people
with serious mental illness. Now meta analysis already show
men and assist with studies conducted on four continents
that showed that people with SMI can use mobile devices that the
penetration rates are not all that different from the general
population in fact that’s one of the few areas where the gaps
between people with the same mind the general population are
quite narrow because the affordability of devices and in
the US access to the federal lifeline Assistance Program
which provides mobile devices and calling text minutes for
people who have a certain level of disability. Now smartphones with those
computational capacities and multi-media resources can really
bring mobile health to life. It doesn’t have to be a staid
written in prevention. We can leverage video and audio
and a host of sensors that are embedded in the device to really
try to do something that perhaps we might not be able to do with
other forms of technology. And so there have been several
studies some from our group some certainly others from other
groups in the US and beyond showing real promise in terms of
m health for SMI. So acceptability feasibility
single arm trials m health has this problem with pilot Titus. Right. A lot of pilot studies a lot of
things have been published but relatively few are cities and
certainly very few comparative effectiveness trials and there
and health for mental health. And so we wanted to see how does
AMP health compared to clinic based care where there’s all
this promise of technology being able to transport delivery of
services outside of the brick and mortar clinic. How does this actually play out
when you put this head to head with a valid evidence based
intervention. And so I’ll describe the mobile
health intervention that was deployed in our PCORI funded
study. It’s folk. It’s called focus it’s a mobile
health it’s a multi-modal smartphone delivered
intervention that my team and I developed over the course of
multiple years the way the focus works that’s installed onto a
device and then the participant gets up to three daily prompts
that will take over the home screen and engage in an
assessment and I’ll demonstrate one of those in a moment. In addition to that all the
self-management resources are installed are downloaded onto
the phone and are available on demand 24/7 regardless of
whether the individual’s connectivity or not. So this can work even in areas
that are blackout zones in terms of access to the Internet or a
cell plan in addition to that the focused intervention package
also has a clinician dashboard so that the response is that
patients have to those self assessments are uploaded to a
secure remote dashboard and they are available to an authorized
individual in this case a person we call the health support
specialist person that would look at the data. Help with technical
troubleshooting using weekly calls to the phone and also
helped bring the intervention to life. Tailor it to the individual
specific life circumstances. And so how does this actually
look. Focus will take over the home
screen and say Can you check in with focus right now. And if the individual says yes
it will immediately launch into an assessment of a target area. That’s one of five core
treatment targets and those core treatment targets are voices or
auditory hallucinations meds or medication use. This is not necessarily
medication adherence. So part of it is remembering to
take your medication. Some of it is also how to strike
up a conversation with your physician and your prescriber if
you’re concerned about side effects social which really
means social skills training or coping with persecutory ideation
which is a common symptom in psychotic disorders mood rash
which really means anxiety or depression depending on what you
endorse and the assessments it’ll take you down to choose
your own intervention route and slip which is not a core symptom
of certainly no question of schizophrenia but we know time
and time again from patient reports and family members that
sleep is often an early indicator when someone is not
doing well and certainly a consequence of an exacerbation
of psychiatric symptoms when people are fatigued. Everything else becomes worse
and so we’ve integrated that into the app that was directly
informed by patients and their users in the design phase. So focus jumps into an
assessment of auditory hallucinations because this
individual the patient Rachel described that as a problem
during a baseline interview and it’ll ask killer Rachel have you
been bothered by voices today with a multiple choice response
option scream if Rachel says anything but not at all then
focus will follow up asking about common dysfunctional
beliefs. This is the type of thinking
that we use in cognitive behavioral interventions for
psychosis. The target are the beliefs and
the distress that the symptoms might have to that individual. The functional impairment that
those that the existence of the symptoms may lead to. How would you describe the
voices that can’t be controlled. They know everything. They’re powerful or unpleasant
or something else. If a common dysfunctional
beliefs and psychosis and so here Rachel endorses that the
voices know everything and focus jumps into action and again
action means ideally very streamlined uni directional
simplified screens right. Our user centered design showed
us that a lot of bells and whistles and multiple directions
to navigate a real is not really well suited for folks for users
with SMI. And so if anything we had to
whittle down and simplify over the development process. Voices may sound like they know
everything but they don’t. Can you think of any time when
the voices made a mistake. How about when the voices were
sure something was going to happen but it didn’t. If the voices got it wrong once
they probably don’t know everything right. Think about it and the idea is
to give people these quick and dirty tools to examine their
beliefs about their symptoms to engage in behavioral strategies
like relaxation or striking up conversations with others who
are socially isolated. In addition to the written
version of this though like I mentioned smartphones have these
multimedia capabilities and so there is a parallel video or
audio version of it. And this is especially relevant
for people who may have limited literacy and education and so
people can select to either read an intervention or see an
intervention if they choose to see an intervention they might
see something like this. Voices make mistakes all the
time. Try this experiment to test
through voices ask them to make a prediction about the color of
the next three cars that you see on the street. Write down the colors and the
order that the voices say. Now test it are the next three
cars exactly in the same order of colors that they mentioned if
they got it wrong. Think about it. What else could the voices
beginning wrong. Their predictions are not always
right. So the idea here is to chisel
away at these beliefs in some cases people have been coping
with this for decades right. I was with some of these ideas
by the way. Those dishes breaking were not
from the video that would be sort of an evil thing to do a
right to sort of talk about voices and at the same time have
something in the back. I like that. So just a quick clarification
about that. The culture dashboard again this
is accessible to the person who is authorized to log on to this
remotely. It summarizes the person’s
engagement over the last seven days. The treatment areas that were
assigned to that laser pointer off in the distance treatment
errors that were sent to that individual. The severity of the responses
whether they’re severe or not symptomatic and an overall log
of the responses to the symptom assessments since they were
involved in the intervention. So a motivated clinician can see
progress over time. Now to the study. So we were thankfully funded by
the court to conduct a comparative effectiveness trial
to see how mobile health for SMI compares to clinic based care. And so the study site was
thresholds in Chicago a multilateralist site large
agency that provides outstanding services to people with SMI in
the greater Chicago area. It was a randomized controlled
trial one hundred and sixty three individuals with SMI about
half of them people with psychotic disorders were
randomized into one of two arms either the health intervention
or rap wellness recovery action planning which is a clinic based
group intervention that was delivered by facilitators with
lived experience. Now the modalities of both of
these interventions are quite different but the overall
objective improving this management increased awareness
to impending symptoms to reach out to others when the crisis
that the overall objectives are quite similar in terms of the
study sample. The average age was forty nine
so not an especially young digitally native or
technologically motivated group ethnicity Sixty five percent
African-American on average a fifth grade reading level and
so. So if folks that really
represented the type of person with a smile that thresholds
provide services to and we conducted assessment at baseline
post treatment which is three months and at six month follow
up to see if the gains ideally gains are retained. And so again this is a quick
breakdown of the allocation to arms. I will cut to the chase and the
that in terms of the outcomes there were no significant
differences in the clinical outcomes or satisfaction ratings
between these conditions both produced positive gains and
satisfaction ratings were also high. But there was the magnitude of
effect in the direction of the effect was not statistically
different between these arms and again dramatically different
treatment modalities one conducted entirely remotely
through a smartphone and one in-person visit with that kind
of support specialist at the beginning of the three months
but really from that point on everything is done remotely
versus a clinic based model weekly two hour sessions with a
group usually comprised of six to 10 other patients two
facilitators but in addition to the clinical outcomes I want to
draw your attention to something that was really quite eye
opening when we when we saw the outcomes of this. We looked at engagement and
treatment commencement right. If we’re talking about access to
care will people have some sort of differential pattern when
they are allocated to a mobile health versus clinic based
models. And if you look at the left
whoops if we look at the left side over here once people are
randomized into one of two arms they get an envelope that tells
them you either call the mobile health specialist to get up and
running with your smartphone intervention or you contact your
facilitator from rap and you set up a clinic appointment. Ninety percent of the people who
are assigned to the mobile health arm went on followed up
and use the smartphone intervention at least once. So they started they commenced
treatment whereas only 58 percent of people who were asked
to come to the clinic ever made it into a single session. And so we lose before people
know anything about the intervention itself. They don’t know exactly what’s
going to happen. Mobile. They just know the modality and
they don’t know much about rap. They just know that it’s a
clinic based intervention. We lose 32 percent of our target
audience right then and there. When we ask them to come to the
clinic. And that in my mind has
implications for things well beyond SMI. This speaks to mobile health the
potential for mobile health broadly in terms of engagement
throughout the study. The engagement was superior for
mobile health. We lost people in terms of fully
engaged participants throughout the study which is common with
patients with SMI. But our bar for engagement here
was actually quite high. So that. So what we thought of it as an
Gauger was either a person who used the mobile app at least
once a day for five out of seven days every week. So that is a fully engaged
patient or person who sat for 70 out of 90 minutes of that group
session. So there were certainly people
who were using the app less engage in sessions for 20
minutes and so on and so forth before they left. In terms of clinical outcomes we
saw significant but similar reductions in depression as
measured by the Beck Depression Inventory at three months and
six months. Similar reductions in Symptom
Checklist claim it’s a measure of general psychopathology
things like anxiety and obsessive beliefs similar
improvements in recovery as measured by the recovery
assessment scale. So this is an indicator of hope
of confidence and coping with symptoms. And when we look at treatment
satisfaction the outcomes were almost identical. So these are the average ratings
on five patient satisfaction ratings. I’m satisfied with the program
the treatment helped me feel better. The treatment program was not
interactive enough. That’s the third one that of
course is reverse scored. Right. So the lower the score the
better that is. I enjoyed the treatment program
and I would recommend the treatment program to a friend
again dramatically different interventions almost identical
in terms of patient satisfaction. So conclusions and have
demonstrated superior treatment commencement and engagement when
compared to click based care focus and rap produced similar
patient satisfaction and clinical outcomes. And so in terms of next steps. Well first of all to answer that
original question it seems that that mobile health can actual at
that potential. All right. And that for some people the
mobile health approach might be more relevant and more engaging
regardless of the content than other clinic based approaches. From our perspective the next
step is how do we actually disseminate and implement this
in the real world. How do we take things from
research from these sterile even though it’s funded real world
pragmatic deployments. This is still artificial
research to some extent. How do we actually integrate
this into the workflow at community health centers and
other critics that really needs to be developed. No one’s really done that yet
hasn’t really taken things the full length that full pipeline
of development all the way through then with rigorous
research and real world meaningful sustainable
implementation of health. It’s very interesting to think
about how a program like the federal lifeline assistance
program that offers devices and call plans might also leverage
things like available apps. If we know that there are
evidence based tools that are helpful couldn’t we also offer
this as part of the basic things that we think are part of your
rights as a citizen of our country to benefit from like
access to 9 1 1. And what about our building
codes and reimbursement models. How do we start incentivizing
and paying and reimbursing for these services whether it’s to
help patients navigate an app look at patient data online make
meaningful decisions in person care based on that data. There’s a lot to be done right
OK. Here you can learn more about
our study and about Macau in general. Feel free to e-mail me or to
check out our activity at MH. Oh there’s a typo there which I
should have corrected. It’s actually MH for MH and
health for mental health that OIG. But if you google me you’ll find
us at the University of Washington. And I’m happy to respond. Thanks for your attention guys
and I look forward for a lively discussion about oh I should. I should have presented the
plaque because we’re doing that we would do a bit of
improvisational discussing parts so the slides are done. But Mark maybe you can comment a
little bit on the study from your perspective. Mark again is the CEO of
thresholds. And so certainly oversaw the
overall activities but is well aware of the landscape of mental
health services in Chicago and beyond. Thank you. Thank you. I have no slides. You’re going to be pretty brief
so we can open it up to an old presentation. I have to say really quickly
that I’ve been introduced today as a doctor and I’m not smart
enough to be a doctor and I’m too young to be the founder
president. But I love both of those. It’s just a little bit about
threshold I think give it some context. We are 60 years old have 40
hundreds. Yeah 30 50000 patients and we
have a budget of around 90 100 million depending on the year. So it’s it’s a big complex
organization. But one of the reasons I’m so
excited about this kind of study at thresholds that we have eight
or nine research studies going on at any time that our center
in Chicago is that 80 percent of our work is delivered in the
community in vivo where people live. So it’s a very different model. We’re not a traditional
community mental health center where people come into services
you know people with serious mental illness don’t come in
often for 60 years. We’ve been reaching people where
they are at through aggressive assertive community treatment. So now over time it is in
Chicago. I probably have 700 outreach
workers that are out on the street at jails and emergency
rooms and nursing homes under the viaduct at a coffee shop
delivering services where they’re at. So what I’m really excited about
with something like and help and focus and mobile technology is
that we’re already out in the community and they have an
extender to our staff that people own themselves to improve
their own health is really really exciting for me. But as CEO I think about what is
what you all know is the triple A every day I’m thinking how are
we meeting the triple aim at Thresholds. Are we improving the patient
experience improving population health and client outcomes and
saving money or at least putting our limited resources in the
right place. So what excites me about this is
that we found that in terms of client satisfaction quality
might not on Oh I’m so loud I just figured I was on WoW that
focused and wrap study that compared the mobile technology
to equivalent high satisfaction which is really exciting
improvements in recovery and health health outcomes and
quality of life indicators high which is really fantastic. And the big question that I as
the CEO think about every day are the costs and what are the
costs of doing this now. Corey we know does not focus on
cost our cost effectiveness looks at cost but that’s what I
have to figure out how do we take this great idea that’s
proving in the field to work and how do we how do we pay for it. So very quickly you know I think
of translating this wonderful research into practice is going
to be really really hard but we have to figure out ways that we
can make it happen and so I use the acronym or the monotony of
when we have to focus on the workforce investment especially
investment infrastructure and what are the natural supports
really the unnatural supports of the systems in the sectors that
are going to help us win in this field. We have 15000 patients this
wonderful study had 80 in 96 160 total 18 in the ritual in each
arm. Right. So you know how do I figure out
a way to if we’re going to implement this would reach 15000
people and more and what are the workforce challenges we provide. Most see evidence based
practices or evidence and practices and we know how
important the mental health interventionist was to the
success or the the how the results of this work and how
would we incorporate that kind of position into a system now
that doesn’t have a code for mobile health interventionists. The investments are huge and
daunting. And we have to figure out ways
to purchase smart phones. Flip phones are available. The technology that we need
smart phones are not at least in Illinois. The data plans are expensive. There’s no current reimbursement
for that. How we figure out how to invest
in an H.R. system which we’re always investing in and make
sure that the information from from the mobile apps is
appropriately integrated into the H.R. Not to mention into the clinical
workflows of my staff and just how we get public sector and the
CEOs and the managed care companies too and the state of
Illinois to move beyond a fee for service system to a value
based or capitated payment that allows us to just do what we
need to do to achieve the outcomes that we need to
achieve. And finally there’s just the
natural supports of integrating this into the technology and
really coming up with a less complex funding system to
support our good work. So I believe we can win this and
I believe we can achieve the Triple Aim and very quickly I do
have a disclosure this fabulous guy was my research director
until he moved on to bigger and better stuff so I’m different
not different. So I’m just very very grateful
that we got to be the place not where our clients were the
objects of a study but very much the subjects under his great
under his great vision. So thank you thank you very
much. So we’re going to have our first
reaction from Silvio Trujillo and if you she’s gonna tell you
a little bit about what she does in the AMA regarding telehealth
and then give a little bit of reaction to the panel. Well I think both presentations
could not be more timely I have a big announcement to me but
I’ll first tell you and it is historic and I think the people
on the panel will be super delighted. So I work with a very large team
within the AMA as well as with external advisors to identify
who actually can get the clicker if possible to identify pathways
to clinical integration of digital medicine interventions
that advance the quadrupling that they improve value they
improve outcomes in patient experience they improve the
overall population health disposition of the patient
panels and more globally within the health system and they also
help the healthcare team not just physicians but the whole
healthcare team have those less administrative burdens and more
delight and joy in the practice of medicine. And Michael really is to work
with this very sprawling big external and internal team to
help focus on some key areas that are depicted up here. And the reason why it’s
important for me to emphasize it’s a very large team is that
if you can imagine those sort of represent the big pieces of
medicine and they represent what clinicians ask when they’re
going to integrate a disruptive potentially disruptive type of
technology or pathway or intervention. They want to know whether or not
it’s clinically efficacious and the presentations today
underscore I think very compelling examples of both the
need and the merit of doing these types of comparative
studies so that we are able to assess the sufficiency of the
clinical evidence to support these types of interventions. We then have to once there’s
demonstration of clinical efficacy figure out a payment
pathway both under a fee for service and alternative payment
models which may include Accountable Care Organizations
or capitated payments we have to think about any liability that
clinicians may face or the health care team and provider
organizations. We think a great deal it turns
out about interoperability integration with each cars and
infrastructure capacity. And then of course we also focus
on the training and workforce needs. Once you have such technologies
in place so both of the presentations talking both about
mental health and dermatological conditions could not be more
appropriate for what was just announced maybe about 30 minutes
ago and I was very anxious because I am the Army’s
Washington counsel and regulatory counsel on the
Medicare physician fee schedule with regard to any of the
digital medicine modalities and see us just after four o’clock
issue. The final rule for Medicare for
cost your 20 19 and announced some pretty historic decisions
around digital medicine. And so some of the examples both
examples today actually could very much benefit in terms of
the payment pathway under fee for service for these types of
modalities. And so CNS announced that they
will be covering an additional set of chronic care management
codes that include a component for patient education and setup
which certainly I think is something that you will look at
in terms of clinical integration. It also includes the technical
component around the actual transmission of data and
creating a meaningful set of transmissions to the treating
clinicians. And the last one relates to the
actual service involved. They also for the first time
ever have agreed to cover a virtual check ins which are very
brief check ins which can be used in substance abuse
treatment programs or in mental health programs not needing an
immediate appointment or a full appointment. And then I think for the
dermatological presentation that we just heard CNS is also for
the first time ever in history. I’m going to cover evaluation of
remote patient recorded images or video for clinician
evaluation and to make determinations of how that that
image or video should be staged. CNS is also going to cover for
the first time ever internet into professional consultations. So it is a big day. But in saying that the road to
this day is also complemented by historic federal legislation
which will expand mental health and substance abuse telehealth
two way audio visual real time coverage. Congress has mandated coverage
in the Medicare program without restrictions on the geographic
location or originating site meaning it can happen anywhere
at any time. And it is to be effective by
July 2019. There are some additional major
historic changes that happen this year legislatively. But I want to emphasize that the
type of research that we’ve just heard was really the linchpin of
the advocacy that has been undertaken to secure this type
of coverage that the head to head comparison with in-person
versus virtual evaluating the staffing needs the technology
needs are very important to demonstrate to commercial payers
and frankly to clinicians that these services are not only
clinically valid but they can be integrated into their practice. I think the dermatological
example is I think a compelling example where we have a
disconnect between the specialties that are needed and
where the patient is located and ease of access. So it’s not simply geographic
proximity but it may be mobility impairments and resources in
order to see their clinician and these challenges are only going
to grow. And so I think both of the
presentations you have seen are just the tip of the iceberg of
the type of research we will continue to need for other areas
of clinical diagnostic interventions in order to expand
coverage beyond what we’re looking at right now is a
historic expansion for mental health services and substance
abuse in the Medicare program. On tap for Bill July 1 and
augmented by these new services that CNS is says they will cover
on January 1 and on the dermatological example I think
we’re seeing that this is an area where imaging and images
increasingly are going to be very important outside of
radiology and even dermatology. We’re going to start to see that
digital and virtual services will play a big role. But the type of research you’ve
heard today which was rigorous that was focused not only on the
clinical outcomes but key issues like how did this how does this
integrate into workflow. How do you scale this up. Also for our purposes in making
the case to both clinicians and to payers we need to demonstrate
value and we’re not talking about cost effectiveness it
often means. Can you address a condition when
it’s lower acuity out or can you address a condition in a lower
cost site of care that is more accessible and affordable to the
patient and I think both of these examples demonstrated that
and I think I’m at time so I’m going to go ahead and hand this
over to us really quickly before you said Medicare three times
you meant to say and Medicaid. Right. So the Medicare physician fee
schedule is Medicare specific. And so the rule that came out
today would impact dual eligibles as well. So to the extent that Medicare
beneficiaries who are also eligible under Medicaid have
been limited in their access to telehealth and other services
this would lift those restrictions that they would
face in those categories that I just identified. No it’s wonderful. I’m just being facetious. 5 percent of our clients are
Medicaid. No no I understand off. But it will positively impact
the dual eligibles. Right. So before we handed over to
Jennifer I just wanted to say one that was just fabulous. And you heard it first here. Right. Yeah. This is really fantastic great
news. Okay. Go ahead Jenna beat that. For this. Oh here we go. I’ve get the clicker. I’m really happy to be here
today as part of this panel to represent the perspective of
state health policymakers. I’m a project director at Nash
B. Our work is guided by an academy
of leaders in state health policy working across agencies
and branches of state government. We have had the pleasure to work
with Pecora on a number of engagement awards. For purposes of time today I’ll
just mention the work that we’re currently doing. We convene a work group of state
policy leaders to explore the policy implications of emerging
research coming out of precautionary. That’s our peace corps work
group and the the work that we do without work group is what is
informing the comments and then perspective that I’m sharing
today. So first off why are state
policy makers interested in telehealth. Before answering that I just
want to note that there is an extremely high level of interest
in telehealth. It is a very pervasive topic I
feel like a very high percentage of the things that pop up in my
inbox are telehealth. It is definitely the topic of
the day. And I think part of the reason
is because it’s seen as a strategy with a lot of potential
to improve access and to address workforce workforce shortages
that states struggle with in terms of serving underserved
populations especially rural areas. It can provide greater access to
specialty providers as we’ve seen today and it also comes up
a lot as a potential response to the opiate epidemic. And in that context really
quickly just as an indicator of the level of interest and
activity at the state level around telehealth. I just want to give you this
quick snapshot of telehealth legislation at the state level
in the last session we saw thirty eight state laws enacted
on a range of topics here from broadband access to cross state
licensing and reimbursement I’m going to talk briefly about some
barriers to the adoption of telehealth interventions related
to payment. As one of our Peace Corps work
group members very aptly put it it’s a little risky to think of
telemedicine as a solution to rural health access. If you don’t think about payment
for it and I also want to say that these brief remarks
obviously are not a complete list of the barriers to the
adoption of telehealth but where the things that rose to the top
and were the points of discussions shared with our work
group. So the first payment related
barrier has to do with lack of alignment across payers in terms
of telehealth reimbursement. In her publication Dr. Armstrong
noted that tele dermatology has not been adopted as widely as
might be expected on the state health policy makers that we
work with. Note the same thing. There’s Medicaid coverage for
telehealth but they just don’t see it used as pervasively or
widely as they might imagine or hope it might be. And when we ask them why this
might be happening they pointed to the patchwork of varying
coverage policy across payers as a factor that could be
inhibiting the use of telehealth the policy response to that
would be parody laws requiring coverage of telehealth across
payers. State lawmakers in 35 states in
Washington D.C. have passed parody laws to require private
insurers to cover telehealth so we’re more than halfway there. But but not all the way there
yet. The next payment related burial
or barrier I’ll discuss is facility fees and facility fees
are fees that hospitals or hospital owned clinics often
charge to cover their overhead costs. In addition to charges for the
actual services that they’re provided that they provide. Currently only 32 state Medicaid
programs allow a payment of a facility fee or as it might be
called a transmission fee when telehealth is used so we can
imagine how that could be a barrier for the adoption of
telehealth if it leads to a decline in revenues for four
clinics. What additional information may
be needed to change practice. State health policymakers are
definitely interested in clinical outcomes related to
tell in health but because they are policymakers they also have
to ask questions like our telehealth interventions cost
effective and they might look at a study from the angle of
wondering if the intervention replaces existing in-person care
or creates a new type of care and therefore a new cost and
also need to understand how telehealth fits into existing
payment models. So you know the economic impact
of a telehealth and intervention might be beyond the scope of a
given study. It’s basically beyond the scope
of what Pecora is authorized to do. But it’s very vital and
important information for policymakers. So I would just say that even if
it’s not something that a study could speak to directly even if
there is indirect information it could still be of use to
policymakers. So it doesn’t apply to the
studies we’ve heard about today but for example if a study shows
an increase in patient advocate activation as an outcome if
there’s any related research that might show a connection
between increased patient advocate activation and better
clinical outcomes and or decreased cost that type of
information would be really important and meaningful context
for policymakers. And I also just want to close by
saying that I I have heard and I and I’m sympathetic to the
perspective of some researchers feeling frustrated that all
policymakers care about our cost. And I would just like to say on
behalf of the folks that I work with in our work group that
that’s fortunately really not the case but unfortunately it is
part of their reality and what we do. States are responsible in most
cases for having balanced budgets and have to make
difficult decisions. And that’s the part of the
reality of what they do. So I’ll stop now and leave time
for a discussion. So I actually have a lot of
questions that I wanted to make sure that the audience has some
questions too so I just wanted to raise a couple of them and I
think one of the things that came up this morning I don’t
know how many of you saw Dr. Ray Dorsey talk about the tell a
medicine intervention that he had with for Parkinson’s disease
but he talked about the digital divide and I know that Dr.
Armstrong your population you had a large Hispanic population. And I think when you were
implementing it it was quite a struggle to actually get
acceptance to the technology. Can you talk a little bit more
about that and what that means in order to for implementation
into practice. Sure. So I work at University of
Southern California which is in East L.A. different from UCLA
which is in West L.A. And and then the predominant
patient population that we serve in our radius are Hispanic. And in fact when I got there my
research team we didn’t have any. We had bilingual research
coordinators but not only did I have bi control research
coordinators essentially made some staff changes because I
realized that to be responsive to the needs of the community
that we’re serving it’s really important to have not only
bilingual but also by cultural staff members will really
understand their needs. A few things is that some of
them while they had access to internet it was not at their
homes. So some of them had to go to
their relatives homes to to access. They wanted to be in the study. And I said you know you have to
have access to internet and imaging device. And so they actually got their
family involved and had to access the Internet that way. And we just we realized that
it’s very important when we do research to ensure that our
staff really understands the needs of the population. So so everything that we do for
this study are translated The platform is itself is entirely
translated also into Spanish so so that we could we could do
this and we one thing is that our providers may not speak
Spanish. And so for part of the study
then we had coordinators essentially making sure all
those were translated. So I think it is to say that
it’s very important that we we really take a patient centric
approach. And even if that involves
difficult choices like staff selections I had to make staff
changes that that need to really be responsive to our patients. Dr. Benjamin do you have some
comments about this digital divide and I know it’s a very
different and difficult population that you’re working
with and you did talk about access to cell phones was pretty
common. Are there any sort of cultural
issues that came up with the different populations you work
with. So again if anything I’m a
little bit more optimistic. I’m used to having such
disparities between the population that I’m interested
in on every other level that in this area of the digital divide
I’m really quite optimistic and seeing the patterns that I’m
seeing. I think that the technologies
broadly are being accessed. And of course there are I think
that there are certain regions that we found in the context of
the Chicago study and my program of research involves mobile
health interventions deployed I think we’re now up to eleven
states where we’ve done this. So certainly we see some pockets
where access to federal lifeline Assistance Program is poor where
we see where that that in Washington state for example
it’s much easier to get access to a smartphone as part as part
of these programs rather than other areas which of course
opens up opportunities. But if anything my message is an
optimistic I think I think that’s the overall trend that
I’m seeing OK. I have to add that there’s the
it’s the economic debate. I mean it gets to my question. Sylvia you know 90 percent of
our clients have Social Security Disability Income. I mean maybe have seven or eight
hundred dollars a month if they’re really lucky and if they
can find a four hundred dollar apartment and have four hundred
dollars for food and transportation there is no money
for the monthly data plan and there is no money to replace the
mobile phone when it is lost is lost. So so we have to figure out not
just the cultural divide but the class divide that is limiting
access to many of the future modalities that we know work and
will change the course of people’s lives. I have one more question before
I hand it off to the audience and it’s a long one. I don’t know if you guys can
make your answer short for this but I know that both Dr.
Armstrong and Dr. Ben and your systems existed outside of the
electronic health record and so that potentially could pose a
barrier for adoption of both of your systems. Can you talk a little bit about
kind of any changes that you see movement in this area or
potentially what needs to be done to address that. I have a thought rather than a
solution. So I think that the question of
interoperability and the ability to connect with electronic
records is a crucial one. And I think that many
stakeholders have incentives to keep things the way that they
are and have electronic health records that are siloed by
nature and that require a fair amount of resources to make
changes even in the context of the same health care system. And so along with the theme of
Patient Centered Outcomes and patient advocacy I think it’s I
think the pressure needs to come from from outside to create
conditions where where it’s mandated to have the flow of
certain forms of data that are likely going to be collected
whether through one particular mobile platform or another of
things like self reported patient data things like
accelerometer data things like geo location devices have a host
of sensors that in one form or another cutting across health
health conditions will need to flow into electronic health
records. And so that’s I find that to
probably be the number one barrier to meaningful
integration even within our health care system where we
might have pockets where it’s not directly fee for service. There might be contractual
agreements and other ways of reimbursement that we don’t
necessarily need to build for the particular deployment of a
device or a mobile home specialist. But still when you need to
provide an external laptop that that works outside of an
electronic health record to a system in order for a clinician
to access a dashboard that creates an enormous barrier to
meaningful use. I think that tech I think the
technology is there to integrate. I think the bigger issue is
getting stakeholders to talk together. There’s a lot of issues around
proprietary ness of each technology so. So I think it’s actually not a
technological issue even that seems like a technological
issue. It’s it’s really about playing
together in the sandbox. And I agree that perhaps outside
pressure is needed another thing. Another way to think about is
that for the larger nature companies that are out there
within that system how do we buff up the telehealth arm of
that. Thanks. And I’m gonna open it up for
questions now and I’m hoping since this is the end of the day
people can stand a little bit later we can stand a little bit
later too. So now I think of I say hi
Alistair Glidden University of Rochester. Quick question for the panel
about machine learning. So obviously when clinicians are
limited when budgets are limited it can be hard to reach all of
the patients in the time that you need to reach them for
something asynchronous like dermatology photos or
interactions with conversation based app. Can we use machine learning to
help screen you know the least complicated cases let’s say. And what are the clinical and
legal considerations that we should take into account if we
want to do something like that. So the aim a just released
policy this June from our house delegates you’re talking about
one of the systems is a subsystem within artificial
intelligence. So I think there are tremendous
promise promises related to not just machine learning but other
A.I. systems. You mentioned actually not just
machine learning but computer vision which are two systems
Mary together we’re delving into what are quite potentially very
significant opportunities to scale using these types of
systems. I’m also an FDA lawyer and I am
up to my neck in the FDA work around regulation oversight of
software as a medical device that is challenging even with
out machine learning. And just to be clear for
everyone to hear what people are really talking about when they
talk about machine learning is that I think you’re talking
about continuously learning systems potentially that are
allowed to change themselves. They’re called learner
algorithms as they obtain more information. They take that information in
and then they modify themselves and then they have a different
output. The short answer to that is it
sounds great. The only machine learning that’s
ready for primetime and I am not unique in this regard and every
responsible researcher from Harvard partners to Duke to
Stanford and from major manufacturers and technology
companies will tell you that unlocked continuous learning
systems that use and machine learning. That’s not responsible. You cannot turn those loose on a
clinical patient population not know what it’s doing and
basically have some pretty pernicious outcomes. Now having said that you can use
machine learning systems some tremendously powerful algorithms
that are locked and then tested validated and then used for
scale for things like interactions through a very
predictable set of when I say very predictable they’re not
that predictable because the machine learning figured it out
and we didn’t. But through a machine learning
validated set of interactions on genomic sequencing they find
variants for cancer that then have diagnostics that can find
things that we couldn’t find. So there are some interesting
questions. But if you can’t explain why the
algorithm is doing what it’s doing you have to convince a
physician if there has been little to no validation or
research to demonstrate that in a head to head way it’s as
effective or appropriate as the average clinician then you’re
probably going to have both a liability problem a coverage
problem and or an adoption problem. There’s a question right now or
if either side one. One comment. So maybe we can think about
machine learning as part of augmented intelligence. That is the clinician is still
at the helm and using tools of artificial intelligence and
machine learning to have augmented intelligence of the
clinician. I do think there like Sylvia
said you know when left it by itself unfettered there could be
problems. I think there are a few
technology already that’s really taking off in dermatology image
capturing using same concept. How many of you use online
banking that they’d take the check. So I use it Bank of America. Right. You put the check there. Now I know you might have to. It says move closer move and
boom and then it snaps the image right. So it’s getting better and
better. So a lot of these are being
adopted into telehealth top DERM where we are optimizing the
image capturing from the patients and right when patients
take a picture. What’s hard is how do you take a
good quality picture. So a lot of that is incorporated
with regards to just that the program is self try to figure
out the optimal optimal way of capturing the image to its
diagnosis. There are a number of apps not
apps. There are programs that are out
there that have shown that for example detection of melanoma. Some of them are scary is the
same between some of the melanoma detection programs
versus a dermatologist. There were no statistical
differences. Right. So that scares most
dermatologists. But but I think it’s actually
it’s good it’s augmented intelligence. So I think if if a tool that’s
going to perform better than me I would want to go to that tool
you know so and there’s obviously a lot of variability. So I think there’s promise and
needs to be augmented intelligence to complement what
we do. And I’m glad to see there are a
lot of evolving research in that area. If I may. I really have to respond. I really do. I spend a tremendous amount of
time on my issues. So they may just call it
augmented intelligence just for the record as do a lot of I
think very responsible developers and manufacturers and
technology companies. Having said that to be very
clear the FDA has already identified 10 cage authorization
of a program called I.D. XDR which provides for screening of
diabetic retinopathy tremendous right. This is great. And the reason why is because we
don’t have enough referrals and so I think the number is
ridiculous like 60 thousand people go blind because they
didn’t have access to screening. This software does not augment
position decision making. It is in lieu of physician
decision making. And what’s fascinating about it
is that when the FDA actually approved it it’s for deployment
in a primary care physicians office and it is it has been
evaluated to make sure not only that it’s as good as or better
than an optimal allergist. The FDA actually if you look at
the memorandum which I urge you all to do with the 5 10k K de
novo authorization evaluating the conditions under which it
should be deployed. So the skill set of this staff
that would be setting up excuse me setting up that machine and
evaluating the likely unexpected adverse consequences that could
flow if it was not properly deployed. I bring that up not to disparage
our health care A.I. but to underscore that that particular
system which is called autonomous A.I. is a locked
system. So I want to emphasize that it’s
very important when people talk about machine learning and deep
learning. They make clear that they are
talking about a locked system being deployed into health care. This this is not banking. If my check doesn’t show up in
my deposit doesn’t show up it’s a major inconvenience. If I have more than a nominal
amount a diabetic retinopathy and the machine says that I have
less than a normal amount of diabetic retinopathy that can
have very serious consequences for my care. This is a really important area
in fact it’s a great area for PCORI and is a whole session
should be dedicated. Yeah I apologize. We are at time but I will take
one more question. But before I take that one more
question I just wanted to encourage people this morning. We had an absolutely stupendous
Facebook Live event that draw Ben Zev and Mark. We’re at and it was really
fantastic it’s up on our website. So if you’re interested in that
very dynamic and really a great way to see what some of the work
that they’re doing. Yes 1 1 1 0. I guess I’ll take two more
questions because I’ve been looking here. Oh OK. Go ahead. My name is Katrina Lane Lindsey
and I am a social worker and I’m a professor at Alabama State
University and one of the things that we’re doing is a colleague
and I are working in regarding to telehealth in a mental rural
area called the Bible Belt the black belt I’m sorry of Alabama
where there are socio economic disparities in that particular
area but my question is. My question is regarding the
gaps in mental health services and the consideration or if you
have had asked clinical social workers who also can also help
and regarding psychosocial issues as it relates to the
population that you’re serving with mental health we are
beginning to do the beginning of the work as it relates to
bridging some of the gap. But like I said we’re at the
beginning of that. And one of the things I wanted
to know is whether or not you’ve done that. Great question. Yes. So while we have psychiatry and
AP and and now integrated co-located primary health care
because people with mental illnesses die earlier than
everybody else in part because they don’t have access to basic
primary care I would say that 90 percent of our work to help me
here you know is really meeting the social determinants of
people’s mental health and behavioral health care. So our staff were focused almost
exclusively on ensuring that people have basic income that
ensuring people have access to food ensuring people have access
to shelter. So you know 80 percent of health
outcomes are related to social determinants and that’s what we
that’s what we focus on. So you I can’t. I do not believe you can do
Community Behavioral Health with poor people anywhere unless your
focus is principally on meeting the social determinants of their
health and behavioral health. So everyone if you feel free to
leave if you want we are at the end of the session. But I do know can the panel stay
for a little bit longer. Are you Are you OK need to go. Yeah I just saw one more
question because I know that you’ve been waiting patiently. Thanks. I understand my name is Julianna
Perry beginning of I am at Johns Hopkins School of Education. I have a question about the
platforms for telehealth or especially health and app
production and wonder if your experience in creating the apps
or your experience at the policy level leads you to be hopeful or
not about possibly creating a uniform initial launch space. When I open up my phone for apps
related to mental health or other health areas for app so
that I can open up my phone and whether you’re made aware I’m
going to get to work and there is a basic platform opening that
then drives me to my jewel my jaw or my jaw or LA raised my
Gellar eyes app that’s specific to me. So from a design perspective
I’ve got some health related apps that I would like to create
but I would like to not reinvent the wheel and create a design
interface that’s different from the one that you’ve created from
the one that you’ve created etc. But I have a easy entree point
that is universal. From a software design
perspective so I can I can I think the vision of having a
platform that that hosts multiple apps whether it’s
complete standalone apps or module elements like you’re
describing is really the focus of industry right now. So there are multiple companies
that for various reasons I won’t name out loud but you can find
them by googling and most mental health platform readily and they
are creating many of these and apps that focus that focus on
things like relaxation strategies all the way through
management of complex disorders comorbid with other comorbid
with diabetes and other things are being developed to sit on
these platforms and I think that the platform rather than the
specific apps that are being hosted is what they are
intending to market and are already trying to do that some
of them start with a V. Some of them start with a P. Thanks very much to our
fantastic panel and thank you for coming to see.

Leave a Reply

Your email address will not be published. Required fields are marked *