Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
People say this all the time we're going to reduce the cost of care
0:02
. It's like , well , we're not really going
0:04
to reduce the cost of care , we
0:06
hope to reduce
0:08
the increase in
0:11
cost growth . We
0:14
hope to be more productive
0:16
or efficient , meaning that we are getting more
0:19
health care quality per unit dollar
0:21
that we're spending . But those aren't
0:24
obvious things . Those are not obvious
0:26
things to measure , particularly when
0:28
health care quality has got
0:31
many , many , many non-monetary
0:33
aspects . What is the value of
0:36
an improvement in my quality
0:38
of life so that I have
0:41
managed diabetes , for example
0:43
, or I don't have my leg amputated or
0:45
any of those things that are about my quality
0:47
of life ? How do we actually measure that in
0:49
a way that enters that value proposition equation
0:52
in the same way that it does when I purchase
0:54
a computer or a TV , where
0:57
you can have that price quality chart and
0:59
it's pretty easy to make those estimates
1:02
of demand , supply and what additional value
1:04
you get for that next
1:06
increment in capability in my TV or
1:08
my computer ?
1:09
Welcome back to the Chilcast , a
1:11
health care podcast from Chilmark Research
1:14
, helping health care leaders make
1:16
the best decisions for the populations
1:18
they serve . Welcome
1:25
back to the Chilcast . I'm the managing
1:27
partner of Chilmark Research and your host today
1:29
, John Moore III . This is the fifth
1:32
episode of our mini series delving
1:34
into how thought leaders representing different
1:36
industry stakeholders think about assessing
1:38
the value of digital health and health IT implementations
1:41
. So we appreciate everybody that has
1:43
been tuning in and welcome to
1:45
our new listeners . This
1:47
series came about as part of a new collaborative
1:49
effort , the Health Impact Project that
1:52
Schillmark convened earlier this year with Curtis Peterson
1:54
of King Fisher Advising Marie Coppola
1:57
of Horta Health , and Pam Arlotto and Susan
1:59
Irby from MyEster Strategies . After
2:02
hearing ad nauseam over the last couple of years
2:04
how important it is for new technologies to
2:06
show efficacy in the current economic climate
2:08
, we decided that it was time to come up with
2:10
a better metric for evaluating the impact
2:12
of technology than a standard , oversimplified
2:15
financial calculation like ROI . This
2:18
series is intended to catalyze industry
2:20
conversations that we see as necessary
2:22
to reach consensus on a new model of value
2:24
that C-Suite's care organizations
2:27
, payers , technology developers , investment
2:30
firms and more can apply to their own initiatives
2:32
. So please be sure to follow
2:34
us on LinkedIn , via the Health Impact Project
2:36
page , for updates and new contact
2:38
, and be sure to subscribe to this podcast
2:41
to not miss any episodes as
2:43
they are released . So
2:45
for our fifth installment , I am pleased to introduce
2:47
none other than Micky Tripathi , PhD
2:50
, current national coordinator of health information
2:52
technology at HHS , where
2:54
he leads the formulation of
2:57
federal health IT strategy and coordinates
2:59
federal health IT policies , standards
3:01
, programs and investments . This
3:03
makes him arguably the most influential public
3:06
servant in our healthcare IT space
3:08
.
3:08
Hi , john , really glad to be here .
3:11
It's our pleasure , really looking forward to this conversation
3:13
. So , for those of you not
3:16
already familiar with Mickey , he is one of those industry
3:18
heavyweights that generally needs no introduction
3:20
, but it's definitely important to provide a
3:22
brief history of his career as valuable context
3:24
for what we'll be discussing today . Dr
3:27
Tripathi brought over 20 years of
3:29
experience across the healthcare IT landscape
3:31
to this federal appointment . He
3:33
most recently served as chief alliance
3:35
officer for Arcadiaio , a
3:38
healthcare data and software company focused
3:40
on population health management and value-based care
3:42
. He was the project
3:45
manager of the Argonaut project , an industry
3:47
collaboration to accelerate the adoption of
3:49
FIRE , and a board member of HL7
3:52
, the Sequoia project , the Commonwealth
3:54
Health Alliance and the Caron Alliance . So
3:56
all very interoperability , heavy initiatives
3:59
and standards-based bodies . Mickey
4:01
served as the president and CEO
4:04
of the Mass E Health Collaborative , a
4:06
non-profit health IT advisory and clinical
4:08
data analytics company . He was
4:10
also the founding president and CEO
4:12
of the Indiana Health Information Exchange
4:14
, a statewide HIE partnered
4:17
with the Regan Streef Institute , and
4:20
he's also an executive advisor to investment
4:22
from LRV Health , who we actually
4:24
spoke with on our last episode , and
4:27
a fellow at the Berkman Klein Center for
4:29
Internet and Society at Harvard University
4:31
. These were all roles that he had
4:34
before coming to the Fed , which is now his
4:36
only current role . He
4:38
holds a PhD in political science
4:40
from the Massachusetts Institute of
4:42
Technology , a master of public
4:44
policy from Harvard and an AB
4:47
in political science from Vassar College , which also
4:49
happens to be my alma mater . So go
4:51
Brewers , not that sports are a
4:53
really big thing there . Prior to receiving
4:56
his PhD , he was a Presidential Management
4:58
Fellow and Senior Operations Research Analyst
5:00
in the Office of the Secretary of Defense in Washington
5:02
DC , for which he received the
5:04
Secretary of Defense Meritorious Civilian
5:07
Service Medal . So
5:09
yeah , really excited to kind of go deep
5:11
into Mickey's career
5:13
and what he sees the
5:16
role of the federal , you know
5:18
, regulation and involvement with health care , it , adoption
5:20
, going forward .
5:22
Great .
5:24
So let's start with kind of
5:26
your earliest days . You got into
5:28
health care IT after you received
5:30
your PhD in political science , so
5:32
that seems like a pretty non-traditional path , but
5:34
I feel like that's how a lot of us end up
5:36
in health IT . It's not necessarily where we intended
5:38
to start or end up in , but
5:41
would you mind sharing your story how
5:43
you went for PhD in polypsi
5:45
and then , you know , ended up being as influential
5:47
and kind of the interoperability in health care IT world
5:49
.
5:50
It is definitely non-traditional , but
5:52
I guess I've always been , you know
5:55
, sort of very intrigued by and interested
5:58
in areas that kind of bring together
6:00
technology , technology
6:02
slash science , policy
6:04
and business . It
6:07
wasn't a deliberate strategy , it's just like looking
6:09
back . It's where I sort of see that I've , you
6:11
know , sort of found myself . I worked in the defense
6:13
industry , in the Pentagon , for a number of years
6:15
and
6:18
found myself sort of in the middle of that
6:20
, you know , right at that nexus as well , you know sort
6:22
of thinking about the
6:24
economics of weapons
6:26
procurement , well , real asset management
6:28
and weapons procurement and policy and
6:31
you know sort of science and technology piece of
6:33
that . So in some ways was
6:35
, you know kind of you know very similar
6:37
to where I ended up in health , it and
6:40
my in studying political science
6:42
. I was very focused on . I
6:44
mean , I started off pre-med in college and then in
6:47
studying political science I was much
6:50
more oriented toward the economics
6:52
and quantitative
6:54
methods side of the house . So
6:57
I was in the political science program at MIT
6:59
but I did the doctoral requirements for
7:01
the economics degree as well and
7:04
taught statistics . So I was always
7:06
like a weird , you know sort of more
7:09
technically , quantitatively oriented
7:11
type . And then
7:13
once I thought that I was
7:15
going to go teach , but then , as
7:17
I sort of looked at what teaching jobs might be like , realized
7:20
that that was really not going to work out . I already had
7:22
three kids at that point . I got my PhD
7:24
a little bit later in my career . I
7:26
already had three kids , was
7:28
unwilling to move from Boston and so
7:30
had to look , you know , for other opportunities
7:33
. And
7:35
right around that time this was like the late
7:37
90s , early 2000 , like late 90s the sort of the dot-com era was really
7:39
peaking and the management
7:43
consulting firms in particular were
7:47
recruiting very heavily , particularly outside of the
7:49
MBA cohort , which is where they typically went
7:52
. They were starting to look at MDs , jds and PhDs . And
7:55
I got caught up in that
7:57
, got very interested in the
7:59
Boston Consulting Group as one
8:01
firm and that's where I ended up . So
8:04
I was there for like five or six years and
8:08
that's where I got really interested in sort of getting my hands really
8:11
dirty and sort of the business elements
8:13
or the business drivers of value . But
8:16
even there was gravitating toward the areas that
8:18
had a significant public policy dimension
8:20
. So
8:22
one of the last engagements I had
8:24
was working out in Indianapolis founding the Indian
8:26
Health Information Exchange and that's
8:28
what really got me sort of involved in Health IT and
8:34
when I decided that this was
8:36
kind of the industry that I wanted to focus on .
8:40
Okay , I know that's fantastic context , especially
8:42
given that the purpose of these conversations
8:44
that we're having are to think about
8:46
kind of the economic impact , the role that
8:48
healthcare technology and data
8:50
technologies coming into healthcare can have
8:53
on truly unlocking
8:56
new value drivers in the
8:58
healthcare sector . Because , as you know
9:01
first hand , inefficiency
9:03
and administrative
9:05
ways are two of the biggest drivers of kind
9:07
of unnecessary expenses in the healthcare space
9:09
, and technology is primed to
9:12
help with those specific problems
9:14
. So , with that
9:16
context and the purpose of this endeavor
9:18
to be developing a new model for
9:20
assessing the value that technology
9:22
delivers in any healthcare
9:25
context , any use case you
9:27
know that these are very broadly
9:29
differentiated use cases and
9:31
that is no easy task . So
9:33
, given your economics
9:36
background and understanding
9:38
the public health aspect of policy
9:40
, how do you think about
9:42
defining the value
9:44
of technology ? How do you think about ensuring
9:46
that what people
9:48
are focusing on and what you can drive people
9:50
to focus on in your role , is actually going to
9:53
have a substantive value
9:56
, impact and outcome for people
9:58
that are deploying these new technologies ?
10:02
Yeah , I mean , I think you know we've
10:05
got a significant challenge in
10:07
healthcare in the US which
10:09
is just the fragmentation of
10:11
, you know , of our healthcare system , and I'll
10:13
put system in quotes . And
10:16
you know , in a country as
10:18
big as ours and as fragmented
10:21
as ours On almost every dimension
10:23
you can think of , and certainly you know , healthcare
10:26
is , you know , is
10:29
, you know , probably the biggest example
10:31
of that , if not one of the biggest examples of that . The
10:34
challenge that we have with almost anything
10:36
is you know what in political science and economics
10:38
they call a collective action problem , which
10:40
is how do you get an industry to move forward
10:43
in a more
10:45
, you know , in somewhat of a synchronized way when
10:48
it's very , very fragmented ? And
10:51
you know it's not as if you can just get , you
10:53
know , seven or eight large companies in the room
10:55
, like you can with banks or like
10:57
with airlines , and get
10:59
an industry agreement among them
11:01
. Usually they themselves , you
11:03
know , are part of driving . Because
11:06
they can , you know the industry side , they're able
11:08
to solve a collective action problem themselves
11:10
or with , you know , relatively little help from
11:12
the federal government . In healthcare we don't have
11:14
that . I mean , if you look at you know
11:16
, we have large provider organizations
11:18
that are nationwide , but
11:21
in most cases they're small in
11:24
most markets . Yes , united
11:26
Health Group is very big in Massachusetts they're
11:28
not really a driver of . You know what happens in the Massachusetts market
11:31
and
11:36
you know the same thing for Ascension Health you know
11:38
very big national integrated delivery network Don't really
11:40
drive a whole lot in , you know , in
11:42
many , many markets , indeed most of the markets in
11:44
the country . So
11:48
you've got this . You know this anomaly that we have
11:50
some very big organizations . Some of them
11:52
are nationwide organizations or maybe it's
11:54
best to think of as multi-regional organizations . And
11:58
the challenge that we have in healthcare is , you know
12:00
, how do you get coordination
12:03
to have the industry drive forward , particularly
12:06
in areas of technology that can
12:08
help to drive the kinds of efficiencies , the kind of higher value
12:10
that we've seen in other industries ? And
12:14
in the absence of market type of orchestration
12:17
you need , I would argue , you
12:19
know , in the healthcare industry you need the
12:21
federal government to be able to play
12:23
a role in kind of tying it together , providing some of that systemness that
12:25
we don't have and
12:29
I think that that's partly , you
12:31
know , sort of owing to the federal government's presence
12:33
in almost every market . So
12:37
you know , it may be that United Health Group
12:39
isn't big in every market , but Medicare and Medicaid sure are Every
12:41
single market you can think of . There are providers
12:43
who are a part of Medicaid , you know , serving
12:45
members of Medicare and Medicaid
12:47
. So it's one aspect , just the participation that only the federal
12:50
government participates in almost
12:52
every market that you can think of across the
12:54
country . The
13:00
second is just that regulatory dimension , that the
13:02
ability to judiciously
13:04
apply regulation that helps the market
13:06
itself move forward and overcome
13:08
some of these collective action problems I think is
13:10
a critical role of the federal government .
13:13
Okay . So getting back to the value piece
13:16
, how do you think about the government
13:18
actually like defining
13:21
the value that people should expect from the
13:23
various initiatives that you're trying to drive
13:25
forward ? So you know , interoperability it's
13:27
almost impossible to assign a specific ROI
13:29
to having data flow . So how do you think
13:31
about value when you
13:33
are not thinking necessarily about dollar and cent
13:36
?
13:36
Yeah , I mean , I think you know , for some of this
13:38
there was , you know there was a lot of effort over
13:40
a number of years to define
13:43
the ROI you know for
13:45
interoperability . There were a lot
13:47
of studies going back to it was an organization
13:49
called CITL , the Center
13:51
for Information Technology . I even have
13:55
their study here on my bookshelf
13:57
. That like like tried to do a very
13:59
specific value proposition
14:01
dissection of what's the value proposition for payers ? what's
14:03
the value proposition for payers large providers , small
14:06
providers what's the value proposition ? And try to sort of
14:08
craft it as almost like a business strategy
14:10
and to see if they could thread the needle
14:12
you know to have . you
14:14
know that sort of be this epiphany
14:16
for all those stakeholders of ah , there
14:18
it is . Now I will , you
14:21
know , sort of , you know , make the investments
14:23
. Because I see the ROI and
14:25
I think after a while a number
14:27
of us just threw up our hands and said
14:29
that the health care is
14:31
too fragmented and the economics
14:34
are too opaque for
14:36
us to be able to , you know , really , you
14:38
know kind of drive it in that way and
14:40
so we really have to , you know , sort of go
14:43
on the assumption which I think is a very
14:45
, very good assumption that interoperability
14:47
is good . It is good and
14:50
it will deliver an enormous number of benefits
14:53
on a lot of different dimensions , and
14:55
it is a public good in
14:57
the economic sense . So the economic definition
14:59
of public good is something that benefits society but
15:01
that no actor finds
15:04
it in their rational self-interest to make the investment right , the
15:07
tragedy of the commons , the whole thing right . I
15:09
mean , all of that is just different elements
15:11
of a public good story . Interoperability
15:13
, I think , falls into that category . Even before that
15:15
, electronic health records fell into that category , people
15:18
weren't willing to invest in electronic health records
15:20
, which is why we had industry stalemate for a number
15:22
of years , and it wasn't until
15:25
Medicare
15:27
and Medicaid essentially through high tech as
15:30
the only players across the country
15:32
who had an interest
15:35
in all markets to
15:37
invest in their supply chain that it
15:39
basically broke the prisoner's dilemma , you
15:42
know , or the public goods problem . However you want to sort of think
15:44
about that they were able to break
15:46
that stalemate to say , well , we
15:48
, the federal government , can invest in that
15:50
supply chain for our business purposes
15:53
, which is Medicare and Medicaid , but that will
15:55
also have spillover effects to the entire economy
15:57
. And so being able to sort of break that open
15:59
, first for electronic health records
16:01
and then for interoperability , is kind of I think
16:03
you know more of the approach that we're taking . And
16:06
you know , and I think you see it in the 21st Century Cores
16:08
Act , which basically said you
16:10
know , interoperability is the natural state
16:12
, which is essentially what that said . You know
16:14
, the whole idea of information blocking is that
16:16
information should just be flowing . That's
16:19
. You know , that's the way it was intended , you
16:22
know , but from by whatever higher being
16:25
you think defines these things , that
16:28
data and healthcare data should just
16:30
be natural flowing , and the idea of information
16:32
blocking is that you need to have
16:34
a very good reason to interfere
16:37
with or prevent that information
16:39
from flowing . And they weren't
16:41
. You know , they weren't thinking about oh , I need an ROI
16:43
calculation at that point . It was just
16:45
that you know , value will
16:47
come to everyone through information
16:50
flowing and for those who
16:52
feel that it
16:54
upsets their markets . Those are
16:57
organizations who are basically
16:59
making profits on the inefficient side and
17:02
obviously , as a federal government , we're , you know
17:04
, going to try to prevent that .
17:07
Yeah , inefficiencies or potentially
17:09
anti competitive behaviors that they don't want exposed
17:11
Obviously .
17:15
No different than any other industry I
17:17
think that's something we always have to remember is
17:19
it's it's a business in the US
17:21
. It's a business . It's a business . We'd
17:24
like to think that certain aspects of it are
17:26
a higher calling and you know , my parents are doctors
17:28
, my daughter's doctor . There is a higher calling
17:30
there , for sure , but the
17:33
basic drivers of you know , of
17:35
the , of you know sort of the industry
17:37
are business drivers and we just have to
17:39
remember they're gonna act like other
17:42
businesses , act in many , many ways , not
17:44
not always , but many , many ways .
17:45
So when we think about the
17:48
hard kind of value of
17:50
interoperability and that being a bit Difficult
17:53
to define very specifically the
17:55
government's kind of way around , that is , to your
17:57
point , the info blocking rules . So
17:59
I've looked at the info blocking Disincentives
18:02
that you guys published this year , so congratulations
18:04
on getting that final rule out there . But
18:06
I think for a lot of people the
18:09
way that it's worded like with a lot
18:11
of federal Legislation , a lot of
18:13
rules it's worded in ways that can be
18:15
a little bit difficult for a layperson
18:17
or for somebody that's not on the legal side to understand
18:20
. So could you give us a concrete example
18:22
of what the penalties
18:25
, the disincentives , would actually look like ? Edit , you
18:27
know an organization of some specific
18:29
side , sure ?
18:30
And so the first set
18:32
I should you know , we should be clear
18:35
and this was in , you know , secretary
18:37
, by Zara step statement with the release
18:39
of the of the property disincentives
18:41
, as it was called , draft
18:43
rule was that this
18:45
sets up a framework For
18:48
this initial set of disincentives
18:50
, with the anticipation that there will
18:52
probably be more that fit within to this
18:54
framework . So the first was establishing how are we going
18:56
to approach this from a general regulatory
18:58
policy perspective , and establishing that
19:01
first and then saying here are the first
19:03
set of you know of
19:05
, of appropriate disincentives . So the three
19:07
that are proposed right now are in
19:09
, are in the CMS program . Again
19:12
, there's nothing that limits the appropriate disincentives
19:14
Program from being limited to CMS
19:16
. That could really be any agency in the Department
19:18
of Health Services and as the biggest
19:21
grant-giving organization in the world , there
19:24
are many , many agencies that that
19:26
do have programs that
19:28
you know that that Organizations
19:30
do receive funding for . So
19:33
but you know , the first set of disincentives you know
19:35
comes through the CMS program and
19:37
it it looks at our
19:39
focuses on three . You
19:41
know , three different types of actors . So in
19:44
information blocking there's the idea of actors
19:46
who is responsible for not
19:48
information blocking , and there are three
19:50
name types providers , health
19:53
information networks and Certified
19:56
technology developers , which would be , like you
19:58
know , electronic health record vendors . So those
20:00
are the three actors who are named , and
20:03
so in identifying , electing a
20:05
set of appropriate disincentives , you first have to ask yourself
20:07
well , where are there government
20:09
programs that actually Involved
20:12
in some way those particular types
20:14
of actors ? And of course , you know , once you get to providers
20:16
You're talking about , you open that box
20:18
up and then it's like , well , there's hospitals
20:21
and there's physician offices , and there's
20:23
federally qualified health centers and there's LT
20:25
pack and there's behavioral health and right
20:27
and so , and each of those has
20:30
different programs associated with them , right
20:32
in different agencies . Samhsa works with you know , something's
20:35
used providers , and HRSA works with federal
20:37
health centers and CMS works kind
20:39
of with all of them . And yeah , so anyway
20:41
. So that's kind of the you know sort of the more detailed
20:44
work you have to do is say , well , each
20:46
actor in theory needs
20:48
a penalty associated with them , and
20:50
so how do you work your way through that ? So the first
20:53
, you know , the first set of appropriate disincentive
20:55
proposals are focused on three
20:57
general types of actors . One is
20:59
hospitals , emu
21:01
eligible providers second category , which is basically
21:04
a physician offices or physicians working in an
21:06
ambulatory setting . And then the third would
21:08
be accountable care organizations and , and
21:10
, and the first two , without
21:12
going you know deep , deep , deep into the weeds on it
21:15
, are essentially a proposed
21:17
appropriate disincentives which would
21:19
, you know , sort
21:21
of limit the , the
21:24
score that they can get for
21:26
certain CMS Incentive
21:29
programs , so promoting interoperability
21:31
, for example , or the MIPS program that you
21:33
know that maybe mumbo jumbo to lay
21:35
people , but Physician offices and
21:37
hospitals absolutely know what those programs
21:40
are . And those programs are programs
21:42
where they can get incentive
21:44
dollars for using , you
21:47
know , electronic health records in certain ways
21:49
that will benefit interoperability . And
21:51
so there's a you know sort of a straightforward
21:54
scoring mechanism they
21:56
get for doing different things and For
21:58
the physician offices in hospitals . That basically would
22:00
say , if you were found on
22:03
investigation by the office of inspector
22:05
general , to be an
22:07
information blocker , as it were . That
22:10
CMS would then have the
22:12
the ability to essentially for
22:15
one part of your score , which is about interoperability
22:17
, I think are using health information exchange
22:19
, or I think it's that categories Would
22:22
essentially say that you get a zero
22:24
score in that category , and
22:26
so at the end of the year , once you're going to rank all that
22:28
up , you know your total score , or then
22:30
calibrates to how much incentive payment do
22:32
you get . So essentially , we'll put a cap on
22:34
you know on how much incentive
22:37
payment you would get , so you know
22:39
the estimates . When we provide this in a rule might be
22:41
, you know , and it depends on you know the
22:43
size of the organization . So it could be like
22:46
for an individual provider , it could be something
22:48
like if you're just a solo practitioner on
22:50
the very low end , that might amount
22:52
to less than a thousand dollar penalty , which
22:55
is really not a penalty per se , it's
22:57
money that you could have
23:00
gotten that as incentive , but you're not getting right
23:02
. But it could amount to
23:04
like for a hospital or a large physician practice
23:06
it could amount to something that's in a hundreds of thousands
23:08
of dollars . So you
23:11
know , kind of read it varies . I mean
23:13
that way for the ACO's the
23:15
proposed penalties a little bit different . It
23:17
would actually preclude an ACO
23:20
organization from participating in
23:22
the program for up to a year . Is
23:24
the proposal , which again is , you know , pretty
23:26
significant Penalty . I mean , if you're an ACO , you
23:29
know , in in an MSSP ACO
23:31
, for example at CMS , you know that's that's a big
23:34
deal for sure . But but
23:36
I think you're supposed to have bite right , I mean there's . So
23:40
anyway , you know there are many , many more
23:42
details , but hopefully that gives a sense just broadly
23:44
of you know how those opposed incentives
23:47
work . And again , this is a draft rule . So you
23:49
know we welcome comments or you know we're
23:51
it's a department rule , it's not no one see rule . We
23:54
were asked by the secretary's office to just
23:56
basically have the pen because it involves
23:58
multiple different agencies . But
24:00
you know we look forward to , you know , to people's
24:03
feedback and comments .
24:05
So do you think that those Disincentives
24:07
will end up getting stronger over time ? Because right now
24:09
they feel fairly soft and I get the
24:11
Sense that part of that's because it's
24:13
still taking people a little while to get to the point
24:16
where they have the tools , they have
24:18
the infrastructure to not be blocking
24:20
, not be providing info blocking . So
24:23
do you think that over time
24:25
you will start adding to that disincentive
24:27
? I mean , a thousand dollars , that's just that's
24:29
a write-off right for some certain smaller
24:31
providers , and even a couple hundred thousand
24:33
dollars isn't a huge impact to the
24:35
revenue stream for a lot of these organizations . They're
24:37
bringing in , you know , hundreds of millions of dollars
24:39
in that patient revenue . So
24:42
Do you think you'll have to
24:44
implement something with a little bit more bite going
24:46
forward , you know , after this initial rule
24:48
gets pushed through ?
24:50
Yeah , I mean I think it's , you know it's hard to know . I
24:52
mean we , you know it's hard
24:54
to know what's . You know what's soft and what's hard . Because
24:58
you know there is one other aspect of it that , from a
25:00
transparency perspective , onc
25:03
, on our website , will post
25:05
or publish the names of all
25:07
the organizations who were found to
25:10
Be
25:12
. You know , I'm not in compliance with the information blocking
25:14
regulations , so there's be a public
25:16
aspect to it as well which could affect
25:18
our hand . And you know , and all that , it's about
25:20
transparency . I mean it's not , it's not disincentive
25:23
per se , it's about just creating transparency
25:25
for the market . But there's that dimension of
25:27
it as well . I think you know we will always
25:29
, will be closely monitoring it for sure
25:31
. And if the eye , if you know , if it looks
25:33
like there's more needed , then
25:36
I think we would always have the opportunity , you
25:39
know , to say well , we're gonna up the you know we're
25:41
gonna up the penalty or find different ways of
25:43
having penalties that have that
25:45
. Have you know more bite ? If you know If
25:47
that's , if that's , you know how it
25:49
looks like it's playing out .
25:51
Okay , make sense . So , obviously
25:54
, in your role , you're talking to a lot of different lobby groups
25:56
. You're talking to different individuals , you're
25:58
doing the speaking circuit , you're hearing
26:01
, you know you got to keep tabs on what's going on with the market
26:03
and where people are thinking on . You
26:05
know the patient side , the provider side
26:07
, the payor side , everybody . So , taking
26:11
into account all those conversations that you're having , what
26:13
are some common themes that you are seeing
26:16
emerge about how different stakeholders assess
26:18
the value of health data technology ?
26:22
No , I think in general and
26:25
I may be just hearing things that
26:27
I want to hear , so we're always
26:29
to be conscious of that but
26:31
in general I feel like
26:33
there is a lot of excitement to
26:36
finally start to get some stuff
26:38
done . But I think we
26:41
feel like it feels like across the board
26:43
, there is a sense that we've spent a
26:45
lot of time converting
26:47
from paper to an electronic
26:50
foundation , to a digital foundation , and that took
26:52
a lot of time and , yes , it was messy
26:55
all of that . But
26:58
in something like 10 , 12 years we've
27:00
completely converted the
27:03
most complex sector of the biggest and most
27:05
complex economy of the world from a
27:07
paper-based system to one that
27:09
is in many , many ways
27:11
, starting to become digitally native , and
27:13
that's a tremendous , tremendous accomplishment
27:16
, and I think
27:18
people don't get enough credit for that , both on
27:20
the public and private side for that
27:22
. But
27:25
I think , that being said , now
27:27
we want to be able to say well , yes
27:29
, you had to get billing done right . Billing
27:32
has to get done . People have to get paid If
27:34
you're not , if those systems don't do billing . My experience helping
27:36
to implement thousands and thousands and thousands
27:39
of electronic medical records was in those
27:41
implementation implementations
27:43
. The minute that revenue cycle had
27:46
a hiccup was the minute that you lost
27:48
everyone . They were like time
27:50
out . We are not doing this . I don't give a shit
27:52
about your EHR , I don't give a shit about
27:54
your vision , but I'm not getting paid anymore
27:56
. So , and that means my nurse isn't getting
27:59
paid , my front staff isn't getting paid and my janitor is
28:01
not getting paid . No one is getting paid . So
28:03
you need to make sure that , first
28:05
and foremost , we're getting paid . I know people criticize
28:07
oh , these are billing systems . It's like well , all
28:10
right , let's ask every one of you do you like getting
28:12
paid or not Not getting paid ? If you don't like
28:14
getting paid , great , fantastic
28:16
, come and volunteer , implement this technology
28:18
on behalf of society and
28:20
don't get paid in the process . But
28:23
so people need to get paid . Yes , we
28:26
need to convert from digital , from paper , to digital . Yes
28:28
, but I think all of us have higher ambitions
28:31
for this . How is this actually going
28:33
to help with curing cancer ? Cancer
28:36
moonshot goals , 50% reduction in
28:38
the next 10 years , or I forget the specific goal in
28:41
cancer , those kinds of things
28:43
. Improvements , significant , substantial improvements
28:45
in public health , maternal health
28:47
all of the things that we're kind of focusing
28:50
on , I think , is where there's
28:52
genuine excitement Just say let's get to that next
28:54
level , let's get to the level of being
28:56
able to say we're now past
28:58
the trough of disillusionment in the Gartner
29:00
hype cycle and we're on the path of the
29:02
slow of enlightenment and let's sort
29:05
of grab that opportunity . So that's kind of
29:07
where I feel like there's a tremendous
29:09
amount of excitement and
29:11
it sort of comes through and
29:13
enthusiasm we've gotten for the US CDI
29:15
, that minimum data set , us Core Data for Interoperability
29:18
both in the market as well as across
29:20
our federal agency partners . The
29:22
enthusiasm we're getting for Tefcom that
29:25
just went live and happy
29:27
to talk more about that , and the enthusiasm that
29:29
we're getting for all the work we're doing to
29:32
make sure that FHIR API
29:34
adoption is moving forward
29:36
aggressively to finally
29:39
have healthcare firmly
29:41
discover the internet in the ways
29:43
that almost every other sector .
29:44
I mean it's still kind of crazy that not
29:47
everybody has online scheduling tools for their
29:49
organization yet . Right , getting back to that
29:51
kind of payment piece , so
29:54
real quick , kind of sidebar question that isn't super
29:56
relevant to the rest of this discussion
29:58
, but do you think that RCEM
30:01
solutions are going to be necessary post
30:03
fee for service ? So if we're actually able to make a full
30:06
transition to value-based care , do you think that those RCEM
30:08
technologies are still going to be relevant ?
30:12
Yeah , I mean , I'm not an expert in
30:14
that , but
30:17
in theory you can imagine that
30:19
they would take a different flavor . But
30:22
we're pretty far away from our goal of having
30:25
everything in value-based care . So I think
30:27
there's going to be a new
30:30
RCEM solutions for a long time
30:32
.
30:32
No , absolutely , but thinking about that being
30:34
one of the first big needs for digital and
30:36
why a lot of the EHRs and
30:39
core systems are so flat and not necessarily
30:41
systems of engagement or , more just , systems of
30:43
record . But as we think
30:45
about that shift away from fee for service , hopefully
30:49
I see some of that investment going more towards
30:51
tracking , value measures and the other kind
30:53
of performance metrics that are going to be tied to
30:55
value-based care , and I think
30:57
that's a really important capitation .
30:59
But to your point , that's just a different flavor of RCEM . My
31:03
experience in implementing was
31:06
pretty consistently been that when
31:10
organizations move from fee for service
31:12
to some kind of value-based
31:14
purchasing arrangement , whatever it is , is
31:16
the minute that they really start focusing
31:18
on interoperability , because
31:21
it's moved from overhead to
31:23
my cost
31:26
of good soul essentially it's
31:31
a part of doing business at that point
31:33
and to delivering
31:35
value , and
31:37
so I generally agree with you that that's
31:40
sort of one of the biggest drivers
31:42
of being able to think more creatively and innovatively
31:44
about how technology can really
31:47
deliver value versus just getting the
31:49
bills paid .
31:51
Yeah , enable new models of care , enable just
31:53
performance improvement , ongoing
31:55
consistent performance improvement , specifically
31:57
, et cetera . So
32:00
, getting back to your training as
32:02
an economist , why
32:05
do you think the value assessment has been
32:07
so hard to achieve in a consistent
32:10
manner in the healthcare tech
32:12
space for so many years ? Obviously
32:15
, it's a relatively immature space , so that's part
32:17
of it , the immaturity . We
32:19
didn't have the data to effectively assess the
32:21
impact . But what else is going on here
32:23
that's made it so hard to actually assess
32:25
the value of IT and healthcare ?
32:29
Well , I think it's a little bit of what we were talking about . That
32:34
, first off , it's in a fee
32:36
for service . World interoperability
32:39
and indeed in many parts of health
32:41
, information technology are just overhead
32:44
. I mean , you're basically
32:46
in a cost plus contract
32:48
kind of arrangement and there
32:50
aren't huge incentives
32:52
to get more efficient
32:55
and or deliver higher value , and
32:59
so I think that's just one of the challenges that we
33:01
have . And again , I don't think
33:03
it's different than any other business . It's just that
33:06
that's kind of the way that a lot
33:08
of the economics flow in
33:10
healthcare . So I think that's been one challenge
33:12
for sure . I think that
33:14
the opaque nature of
33:16
healthcare
33:19
economics is also right
33:21
, and this is like in every healthcare
33:23
economics 101 or in every economics
33:25
101 course you'll get is
33:27
well , here's how markets work , and then let's
33:29
talk about markets that don't have
33:31
some of these core sort
33:34
of principles . So meet these core principles
33:36
of plausibly efficient markets
33:38
, and healthcare and defense are the two right
33:40
that every single economics processor across
33:42
the country teaches . And
33:45
the part of the healthcare equation is that
33:47
consumers are shielded
33:49
essentially from the
33:52
purchase decisions that they make in
33:55
certain significant ways , that
33:57
there are information and symmetries . It's really
33:59
hard for them to assess value and
34:01
to assess incremental value for the incremental
34:03
value of the next unit of payment . And
34:09
also there is a very asymmetric view of
34:11
risk , which is actually very similar to defense
34:13
, where I did a lot of work with the defense industry I was talking about
34:16
before . It's very similar in defense right
34:18
, where you don't have sort of a symmetric
34:20
idea of risk . No-transcript
34:22
risk , wait a minute , I increase my risk of
34:25
dying five percentage points . That's
34:27
not good , that's
34:29
really not good . In the same way , with defense
34:32
, we're always like wait a minute . Every
34:34
increment in risk you're adding is a
34:36
risk to our entire country and our entire way
34:38
of life . That's not good , right ? So
34:41
we do everything that we can to move
34:44
away from that risk and to be very risk
34:46
averse . On that dimension , healthcare
34:48
is very similar , I think , and individuals behave
34:50
that way . So I think that's part of the challenge is that
34:53
it doesn't have those
34:56
simple or relatively simple kinds
34:59
of value metrics that we can apply in other
35:01
parts of the economy
35:04
. I think the other part of this
35:06
which makes it especially challenging is
35:09
that it's a very
35:13
fast moving technology space and
35:16
so for anyone who is
35:19
trying to say well , I'm going to cut
35:21
, people
35:23
say this all the time we're going to reduce the cost of care
35:25
. It's like , well , we're not really going
35:27
to reduce the cost of care , we
35:29
hope to reduce
35:31
the increase in
35:34
cost
35:36
growth . We hope to
35:38
be more productive or efficient
35:41
, meaning that we are getting more healthcare
35:43
quality per unit dollar
35:45
that we're spending . But
35:48
those aren't obvious things , right ? Those are
35:50
not obvious things to measure , particularly
35:52
when healthcare quality has
35:54
got many
35:57
, many , many non-monetary aspects . What
36:00
is the value of an
36:02
improvement in my quality of life so
36:05
that I have
36:07
managed diabetes , for example
36:09
, or I don't have my leg amputated , or
36:13
any of those things that are about my quality
36:15
of life ? How do we actually measure that in
36:17
a way that enters that
36:19
value proposition equation
36:22
in the same way that it does when I purchase
36:24
a computer or a TV
36:26
, where you can have that price
36:29
quality chart and it's pretty easy
36:31
to make those estimates of demand supply
36:33
and what additional value
36:36
you get for that next
36:38
increment in capability
36:40
in my TV or my computer
36:42
. I think all of those things make it really
36:45
hard to have that kind of calculation .
36:48
Yeah , I completely agree . That's something that
36:50
at Chillmark we've been watching play out since we first
36:52
entered the space . It's a
36:55
conversation that hasn't been as
36:57
tantamount to the
36:59
activity in the industry , because for a while there
37:02
we founded the company shortly before Hightep
37:04
got passed For
37:06
a good chunk of the last decade people were
37:09
focused on that area , on meaningful use , on
37:11
just core infrastructure getting implemented , getting
37:13
the government incentive payments
37:15
for implementing this core infrastructure
37:17
. I think that now that
37:20
some of that money is dried up and there
37:22
isn't as much new money coming in
37:24
for core infrastructure investments
37:27
, you are seeing the tight
37:29
margins and the financial constraints of
37:32
healthcare organizations really come into play , with more
37:34
of these decision-making and having
37:36
to much more align any new investments with
37:39
their own organizational strategy versus some
37:41
outside strategy that's being pushed onto them
37:43
.
37:44
Right , I think that's the nature of the public
37:46
goods problem I think you just
37:49
said it which is that no organization for
37:51
all of these organizations they feel like
37:53
, well , it's not , how much
37:55
additional value do I get by
37:57
that next incremental investment
38:00
in interoperability to
38:02
get that next increment of information ? And
38:04
for most organizations they'll be like
38:07
pretty low . I'd rather just ask the patient
38:10
Right
38:12
which is not the best way of doing this
38:14
, and I would argue that that's
38:16
one of the aspects of public goods that
38:19
the federal government needs to sort of help
38:21
to solve that inefficiency
38:23
, that market failure or
38:25
dimension of a market failure . That is a public
38:28
good . What is the value to our country
38:30
of having an effective public health system
38:32
? What's the value to our country
38:34
of having patients have
38:36
as much information available to their
38:38
providers to provide the best
38:41
possible care ? Those are all
38:43
public goods that the market itself
38:45
is not going to value in the ways that we
38:47
, as a country , I think should want the
38:49
market value .
38:52
Yeah , I think that's a really good way of putting it . And
38:54
so that kind of gets into the
38:57
next section of this discussion , which we're
38:59
recording this a few days before
39:01
Christmas 2023 . And last
39:03
week you guys had a very busy week
39:06
with your annual ONC
39:08
conference , the first one since 2019
39:10
. And you had a whole bunch of news coming
39:12
out over the last week and a half . So QHINZ
39:15
went live . Hti one final rule
39:17
came out . You got your AI
39:19
transparency element got added into HTI
39:22
one from the initial proposed rule . So
39:24
can we start at the top and just you
39:26
know what's the big deal about QHINZ ? Just a quick
39:29
summary of why this is so exciting , why it's been
39:31
such a big focus of the government for the last few years
39:33
making this happen and getting us to the point that we are today
39:35
.
39:38
Sure , really , since the founding of ONC
39:40
in 2004 , there's
39:42
been the vision of essentially a nationwide
39:45
network for
39:47
medical record interoperability and
39:51
you know there's a need for some type of governance
39:53
for that . If you look at any other network
39:55
, any other industry , there are networks
39:57
that provide that backbone
39:59
of assurance for
40:02
reliability , for trust , for security
40:04
, for , you
40:07
know , collaborative problem solving
40:10
that are important to
40:12
you know , make sure that you've got that
40:14
you know sort of the backbone
40:17
to have information sharing , to
40:20
make sure that information gets where
40:22
it needs to get when it's needed
40:24
. And if you look at , you know , banking
40:26
system , for example , ach payments , right
40:28
, there's a huge , you know
40:30
sort of infrastructure behind the scenes that makes
40:32
sure that when I do my Benmo
40:35
payment or I write a check or I do a credit card
40:37
transaction , that all that gets reconciled
40:39
on the back end and I don't have to worry
40:41
about it as an individual . We don't , you know we've
40:43
been building toward that . In healthcare we
40:45
had some HIEs
40:47
early on , indian Health Information Exchange , which I
40:49
you know , which I've found , you know was
40:52
one of them . And then we had
40:54
the ONC program which looked from
40:56
a geographic perspective and sort
40:58
of made an assumption that , well , each state
41:00
will have an HIE and then
41:02
those HIEs will get , you know , connected up
41:05
as it turned out . You know that isn't the way it worked . Some
41:08
states , you know geography
41:10
was a good basis for health information
41:12
exchange network in
41:14
some places , but in most places
41:17
it actually wasn't a great you know sort
41:19
of dimension of affinity or you know sort of
41:21
a vector of trust , however you want to think
41:23
about that . And so then we moved to
41:25
the next level , which is you've got some
41:27
state you know on regional HIEs
41:29
but more nationwide type of
41:32
networks forming the health exchange which spun
41:34
out from the federal government . Care
41:36
of quality , the Commonwealth of Health Alliance , epics
41:39
, care Everywhere , for example , obviously
41:42
was , you know , was designed for , you know with an EPIC , but all of
41:44
those you know kind
41:46
of really being more organic efforts
41:48
to say , where is the business connection that we have here ? And
41:52
then how do we build on that to create a network to support
41:54
you know , to serve our participants ? So we're at the point now that we've got you
41:56
know sort of a number of networks and
42:00
the direction from the 21st Century
42:02
Cures Act was to , you know , try to
42:04
once and for all say all
42:06
right , at this point we need to connect
42:08
them up so that we have the user experience , the
42:11
network of networks , user experience , so that whichever
42:14
network you're on , you will be connected to
42:16
everyone else for a basic set
42:19
of , you know , of services , that basic interoperability . And that's
42:22
what we've been working toward and , again , that's you know . If you look at the phone
42:24
industry , it went through
42:26
a very similar type of process . There
42:29
was , you know , for there was a long time where you
42:32
know , the old Ma Bell AT&T , you
42:34
know , covered , you know , I actually
42:36
know this information because I did some research on it
42:39
and I like to talk about it . Like , in the early 1900s
42:41
there were something like 2 million phone
42:43
users across the country , which itself
42:45
was , I think , just sort of shocking that there were
42:47
that many phone users , but something like 50%
42:50
of them were connected in the
42:52
AT&T network , the old Ma Bell network
42:54
, right ? So you got about a million or
42:56
maybe it was a little bit more . Maybe it was like two thirds were
42:58
connected on the Ma Bell network , one
43:01
third remaining , one third
43:03
were covered by
43:06
over 2,000 networks . So
43:09
you think about that sort of feels
43:12
a little bit like we are today , right , you've got a couple of big
43:14
networks and then you have a whole bunch of small networks
43:16
. Then the problem for us as a country
43:18
was to say all right , how do we connect up all those networks
43:21
? That took like 50 years . In
43:23
telephone history it took like 50 years for
43:25
that to all be consolidated in
43:28
a network of networks type of structure where
43:30
, whether I use AT&T or
43:32
Verizon or T-Mobile or Sprint , I
43:34
don't for a second worry about whether I
43:37
will be able to talk to someone on another phone on
43:39
another proprietary system . That's
43:42
what we wanted to be able to do with Tefka and that was the direction
43:44
we got from the Congress to build . So
43:47
that's the importance of these networks . You can think
43:49
of them almost as like the AT&T , verizon
43:52
, t-mobile and Sprint of medical
43:54
record interoperability . You've got to have
43:56
some organization that's already got a network
43:59
and you want to connect them up to
44:01
solve that problem network to network
44:03
, when they're already providing services within their
44:05
network . So we had five organizations
44:08
who stepped forward to do this voluntarily . There
44:11
was nothing in the 21st Century Cures Act that gave us
44:13
either funding or authority , meaning
44:15
the ability to tell everyone
44:18
in the country you are required to
44:20
join this network . We didn't get any of
44:22
that direction from the Congress . So we
44:24
really needed to put it together as a model
44:26
that the industry saw
44:28
as being valuable from an industry perspective
44:31
, and I think what you see is these
44:33
five organizations , as well as two who
44:35
are in the last stages of their
44:37
implementation . Plus we've got a couple who
44:39
are in the application stage , two more
44:41
who have publicly announced very large
44:43
, influential organizations who have announced
44:45
that they have submitted their applications and intend
44:48
to become qualified networks as
44:50
well . That starts to sort of fill
44:52
out that idea of we've got
44:54
some networks that connect with
44:56
each other . Regardless , if you're a provider
44:58
, if you're a hospital , regardless of which network
45:00
you join , you will have the confidence that
45:02
you can exchange information with anyone else
45:05
, regardless of which network they're on .
45:08
So can you just give me a quick overview
45:11
, very brief , of what exactly
45:13
the data is that is going to be included
45:16
in the QHINs , because it's not
45:18
going to be all the data . There's a lot of unstructured data
45:20
. There's still pending
45:22
standards coming out for USCDI
45:24
. So at this
45:26
point in the development
45:28
of the QHIN concept , what data
45:30
elements and what exchange is actually
45:33
happening on this national network ?
45:36
Sure . So the requirement
45:38
in theory is
45:41
EHI Electronic Health Information
45:43
which is basically
45:45
the same definition that's used for
45:47
the information walking rules
45:49
, which is basically all electronic
45:51
health information . But we recognize
45:54
that that's really hard to do when you
45:56
don't have standards for all of that , and
45:59
so the minimum expectation
46:01
, the base expectation , is that
46:03
what's exchanged is the US core data
46:05
for interoperability , the USCDI , and
46:08
that's required by EHR vendors
46:11
. It's required in various federal programs
46:13
to be supported for quality measures or other
46:16
things , so it's kind of the
46:18
minimum data set of the healthcare delivery system . Already
46:20
I will note that the USCDI does
46:22
have eight note types in it as well
46:24
. So it's not as if the USCDI is only
46:26
structured data . It's not . It
46:28
includes notes , eight
46:31
note types that are required to be supported as part
46:33
of the USCDI . So that's the
46:35
minimum expectation , that that's what will be exchanged
46:37
. Again
46:40
, the requirement and it's a loose requirement
46:42
is that you send whatever you have electronically
46:44
. We just think
46:46
that we can have a good expectation that everyone
46:48
can send the USCDI , because that's
46:50
what most of those networks do today . In some
46:52
way , shape or form , they send the USCDI and
46:55
we expect that that would grow over time and certainly the information
46:57
blocking rules will provide
46:59
added impetus for people to share
47:01
more information than the USCDI , because the
47:03
information blocking rule is not limited to
47:05
the USCDI . It's limited to everything electronic
47:08
, everything in quotes , because that's
47:10
a pretty hard term to get your head around
47:12
. But
47:15
as the information blocking rule sort
47:17
of comes into place and people start
47:19
to , with enforcement and all of that coming
47:21
into place , we think
47:23
that the TEFCA would replace that . Organizations will feel like , okay
47:27
, I can share more and more
47:29
of that electronic information in the TEFCA network
47:31
and then I
47:33
will be complying with the information blocking or substantial
47:35
portions of the information blocking regulation because
47:38
I'm just making the information available and
47:41
it'd be pretty hard at least for the information
47:43
that someone's asking for , could be
47:45
pretty hard for someone to make a case that I'm information blocking
47:47
If they're on TEFCA
47:49
and I'm on TEFCA and I'm making this information available . That helps
47:51
from a compliance perspective as well . So
47:55
we think that that adds added impetus
47:57
for people to want to do it .
48:01
Okay , so we're getting a little bit late in the recording session so
48:03
I've got to jump through some of these a little bit
48:05
fast . I
48:08
think in general , most people can look
48:10
up HTI1 and see the summary that other folks have put out . I'll
48:12
share a couple of those in the show notes , but
48:15
can you just give a really quick overview
48:17
of what exactly
48:19
is contained in HTI1 ? And
48:23
how you perceive the evolution
48:25
of this ? Because the organization
48:28
has telegraphed out that
48:30
HTI1 is the first of these new certification
48:32
requirements and there's got to most
48:35
likely be an HTI2 in 2024 . I
48:37
mean that's going to be dropped in 2024
48:39
for commentary to be implemented 2025 . So
48:42
, as we think about the evolution
48:44
of these new rules for certification
48:46
, how do you think that will evolve and what exactly
48:49
are we starting with and what do you
48:51
see as being the next steps ?
48:53
Sure , yeah , so HTI1
48:55
isn't just about certification but it's
48:58
got . I mean , certainly certification is a big part of it , because
49:00
there are some things in there on information blocking
49:02
which technically isn't about certification . That's a whole
49:04
different thing which is information blocking , but
49:08
the things I mean there are many things . It's a 958
49:10
page rule , I think . So there's lots of things in it and
49:14
but you know , things that I would point to
49:16
as being really significant , I think , are
49:20
three things . One is the
49:23
establishment of the USCDI version
49:25
3 as the baseline
49:27
standard for the minimum data set , essentially
49:30
for the industry , because that includes pretty significant
49:32
health equity related data elements
49:34
that weren't a part of the USCDI
49:36
. That you know , that we inherited coming
49:38
in and so you know . So that's
49:41
a pretty big deal and you know the industry has been asking
49:43
for that , federal agency programs
49:46
are asking for that , so that's a really important part of
49:48
it . The second is it's the
49:50
first , you know , set
49:52
of regulations that this administration
49:55
has put into place related to AI
49:57
and healthcare . There's obviously
49:59
already different regulatory mechanisms you
50:02
know that are in place . You
50:04
know OCR I shouldn't say it's first , I mean OCR
50:06
has draft rule related to non-discrimination
50:09
in the use of AI-enabled tools . Fda
50:12
regulation has been placed for many years . That does
50:14
, you know , have authority over medical
50:17
devices that have AI-enabled
50:19
features in them . But this is a , you
50:21
know , pretty substantial regulatory
50:24
step forward , specifically with regard
50:26
to AI and healthcare and the requirement
50:28
there is for transparency . It's
50:30
really to empower clinicians , whether
50:34
it's physicians in physician offices or in
50:36
hospitals . It's to empower them
50:38
so that they have better
50:40
awareness of the AI-enabled technologies
50:43
that are embedded in their electronic
50:45
health record system and
50:47
to give them some information
50:50
so that they can make better judgments about whether
50:52
that particular AI-enabled tool is
50:55
appropriate for in their particular
50:57
setting . So it's really not about regulating
50:59
the AI-enabled tools themselves . We're not assessing those . We're
51:03
not monitoring those . What we are doing is saying that if
51:05
you have an AI-enabled tool in your
51:08
electronic health record system , you , as the EHR
51:11
, have a responsibility to
51:13
make available information
51:18
that will allow the user of your system
51:20
to judge whether that particular AI-enabled tool makes
51:23
sense in their environment . And then the third thing that's
51:25
part of the rule was that the
51:27
statutory requirement that came from
51:29
the 21st Century Cures Act , which is
51:32
a new kind of
51:34
program called the Insights Program
51:36
, which requires that electronic health
51:38
record vendors produce
51:40
information that
51:42
will be aggregated and made available
51:44
publicly . That gives us a sense
51:46
of where are we with the EHR
51:49
implementations and interoperability . So
51:51
how do we think at an industry level , like an industry
51:53
level dashboard of health information
51:56
technology adoption and use
51:58
and again , that was a statutory
52:00
requirement , so it was required that
52:02
we put that into place and
52:04
the first set of measures that
52:06
we are proposing to be part
52:09
of that dashboard are part of that HTI one
52:11
.
52:14
Okay , so I have one
52:16
final question that that all raised
52:18
. I had a whole bunch more questions
52:20
in our discussion guide but due to limited time
52:22
there's only so much we can cover . But
52:25
a few weeks back the
52:28
White House , the Biden-Harris administration
52:31
, released their SDOH playbook and
52:33
earlier this year at a congressional briefing , you stated
52:36
that quote . We are
52:38
really thinking about health equity as a core
52:40
design principle , starting
52:42
with the data itself . You've got to have that data
52:44
available in order to be able to identify where
52:46
there might be communities that are getting different types of care . So
52:50
how much does that tie into
52:52
this new SDOH playbook initiative and
52:54
how much is the
52:57
ONC involved with the development
52:59
of this playbook ? That is supposed to be if they're
53:01
able to pursue their dream
53:03
. As it's laid out , it's supposed to be an all-of-government
53:06
initiative , pulling in data from
53:08
various disparate sources to better identify
53:10
what are the social ails
53:12
that can be traced back to social determinants of health
53:15
and health resource needs . So
53:18
can you talk a little bit about ONC's
53:20
role in moving that forward and
53:22
scoping that out ?
53:24
Sure , so , yeah , so we were very
53:26
involved from minute
53:28
one of the SDOH . It
53:30
was what's called an IPC , an Inter-HC
53:32
Policy Committee , that created the
53:34
SDOH playbook . So
53:37
it's , you know , ipcs
53:39
are kind of , you know , sort
53:41
of started or initiated
53:43
by the White House , someone in the White
53:45
House or a group in the White House and , as the name
53:47
implies , it brings together federal
53:50
agencies together to
53:52
work on a particular area . And in this case , sdoh
53:54
was , you know , was there , there was one for AI and
53:56
there's , you know , others on the many of them , but
53:59
this one we were involved in , you know , literally
54:01
from the very beginning , as they were
54:03
just forming the IPC . We
54:07
were directly involved with them and , you know , trying
54:09
to sort of think about how we think about this from
54:11
a health and information technology perspective , because
54:13
, obviously , electronic health records and interoperability , but
54:16
you , I mean you , can't have a healthcare
54:18
discussion meaningfully , I would
54:20
argue , at this point , without talking
54:23
about the health information technology aspect
54:25
of it and how that will either enhance
54:28
or , you know , be a detriment
54:30
to whatever it is you're trying to accomplish . And then how
54:32
we can , you know , sort of figure out how to , you know , make up
54:34
something that's helping to advance . You
54:36
know , important policy objectives . So when
54:38
you look at the SDOH playbook , I think you see significant
54:41
portions that talk about interoperability . It talks about
54:43
, you know , social and mental health . It talks about
54:45
the importance of having the USCDI , you
54:47
know , be a core construct of that . As
54:50
I mentioned before , you know , having the USCDI version
54:52
3 is a really important step forward
54:54
because that has those health information , that
54:56
health equity data elements that
54:59
all of us are looking for to be able to build
55:01
programs around . First we're going to identify
55:03
where there are issues and then , second , to be able to have
55:05
programs , you know , built , leveraging those data elements
55:07
. And you know , uscdi
55:09
version 3 , assuming that it becomes
55:12
well , it is actually we've
55:14
got the final rule out it will now
55:16
, you know , become the basis for
55:18
Tefka exchange , for example . We start with USCDI
55:20
1 because that's what we have , but that
55:23
will up to USCDI version 3 , which starts
55:25
to have those data elements that allow us to
55:27
be able to have that health equity by design kind
55:30
of construct . That it's starting with the data and then you build up from
55:32
there .
55:35
Okay , yeah , I mean I would
55:37
love to talk to you more about the SDOH playbook
55:39
and how much staying power
55:41
that has post this administration
55:43
, because I think that's one of the big unknowns right now . No-transcript
55:49
. I think that there's a huge need
55:51
for that across healthcare . But figuring out the kind of economic
55:53
model for that , While
55:58
we're still operating under largely fee for service
56:00
kind of infrastructure , I think that's a huge need
56:02
for that across healthcare . But figuring out the
56:04
kind of economic model for that While
56:15
we're still operating under largely fee for
56:17
service kind of infrastructure , is something that you and I could probably talk about for
56:19
days . It's such a big need and such an interesting area . So maybe
56:21
that's maybe we can have a follow up conversation
56:23
on that sometime in the future , but for now , thank you so much for joining us today
56:26
, Mickey , Particularly as we
56:28
explore the many ways that industry stays in the future
56:30
, and I think that's a really important part of the process . So
56:35
if any of our listeners want to be more involved with the process
56:37
of developing new policy , how would you advise they go about making their
56:39
voices heard ? How would they get involved with the government
56:41
? Yeah
56:45
, I mean there's a variety of ways , you know . Certainly . We
56:47
meet with stakeholders all the time , so we're always happy to do that . Our
56:50
high tech , which is our federal advisers , we're always happy to do
56:52
that .
56:54
Our high tech , which is our federal advisers committee
56:57
, does everything in public , both the full committees as well
56:59
as the subcommittees , and always at every meeting has
57:01
an opportunity for public comment . So
57:03
I think , getting on our website looking up the high tech and
57:05
there's a number of different work groups there and people
57:08
are interested in that that's
57:18
a great opportunity to sort of , you know , just dive
57:20
in deep on the policy issues that we're contending
57:23
with and be able to offer the opportunity to provide
57:25
public comment , all of which is recorded
57:27
, and you know that we always take into account .
57:30
All right . Well , there you have it , folks . A big
57:32
thank you for Mickey Tripathi for taking this time
57:34
. I know that December is always busy
57:37
for everyone , so I really appreciate you meeting
57:39
with us to record the session and sharing your expertise with us . Happy
57:44
holidays and have a great new year .
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More