1
00:00:00,180 --> 00:00:04,110
And what I've seen decade over decade,
'cause I can actually say that now
2
00:00:04,110 --> 00:00:07,290
I've been doing this for a long time,
is those who wait, I wouldn't say they
3
00:00:07,290 --> 00:00:09,510
get hurt, but they lose advantage.
4
00:00:09,570 --> 00:00:12,540
And negative advantage is the
same thing as getting hurt
5
00:00:12,540 --> 00:00:13,680
when the market moves fast.
6
00:00:13,740 --> 00:00:17,940
So I would say get connected and then
start finding tools for the specific
7
00:00:17,940 --> 00:00:19,260
things that you're looking for.
8
00:00:19,350 --> 00:00:22,440
It's not like a sea-change
moment when you into interop.
9
00:00:22,530 --> 00:00:26,040
It's an incremental movement that
leads other increments, movements
10
00:00:26,040 --> 00:00:27,390
that leads to general success.
11
00:00:35,840 --> 00:00:39,530
Welcome to Hard Problems, Smart
Solutions, the Newfire Podcast where
12
00:00:39,530 --> 00:00:43,040
we explore the toughest challenges and
the smartest solutions with leaders
13
00:00:43,040 --> 00:00:44,810
across technology and healthcare.
14
00:00:45,500 --> 00:00:48,590
I'm Brendan Iglehart, Staff Healthcare
Architects here at Newfire Global
15
00:00:48,590 --> 00:00:50,780
Partners, and your host for this episode.
16
00:00:53,476 --> 00:00:55,966
For today's episode, we have the
pleasure of speaking with Paul
17
00:00:55,966 --> 00:00:58,996
Wilder, the executive director of
the CommonWell Health Alliance.
18
00:00:59,446 --> 00:01:02,536
Paul brings more than two decades of
experience in health IT leadership,
19
00:01:02,956 --> 00:01:05,326
deeply committed to transforming
healthcare delivery through
20
00:01:05,326 --> 00:01:06,581
technology and data exchange.
21
00:01:07,550 --> 00:01:10,340
He's at the forefront of driving
nationwide interoperability,
22
00:01:10,700 --> 00:01:13,490
working to empower clinicians,
practitioners, and individuals
23
00:01:13,490 --> 00:01:15,740
with robust data exchange services.
24
00:01:15,957 --> 00:01:19,677
Paul's work at CommonWell aligns
perfectly with our podcast mission to
25
00:01:19,677 --> 00:01:22,677
discuss the toughest challenges and
the smartest solutions with leaders
26
00:01:22,677 --> 00:01:24,207
across technology and healthcare.
27
00:01:24,434 --> 00:01:27,194
He is known for his insightful
perspectives on the complex world of
28
00:01:27,194 --> 00:01:30,844
health data from regulatory shift to the
practical realities of implementation.
29
00:01:31,429 --> 00:01:32,809
Paul, thanks for joining us today.
30
00:01:33,529 --> 00:01:33,739
Great.
31
00:01:33,739 --> 00:01:34,339
Thanks for being here.
32
00:01:34,369 --> 00:01:35,299
Thanks for letting me be here.
33
00:01:35,419 --> 00:01:35,854
It should be fun.
34
00:01:36,902 --> 00:01:41,012
So to get started, Paul, I'm, I want to
begin with your journey in health IT.
35
00:01:41,462 --> 00:01:45,642
So, you've had a significant tenure
back at, before with the New York
36
00:01:45,642 --> 00:01:48,582
eHealth Collaborative before moving
into your current role as the
37
00:01:48,582 --> 00:01:50,282
executive director at CommonWell.
38
00:01:50,532 --> 00:01:54,292
So what drew you into this crazy
world of healthcare interoperability?
39
00:01:55,354 --> 00:01:55,594
Well
40
00:01:55,594 --> 00:01:56,314
First I was drawn to
41
00:01:56,314 --> 00:01:59,504
healthcare starting technology
that was quite outside.
42
00:01:59,504 --> 00:02:03,784
I was doing network support cellular
networks kind of stuff, and immediately
43
00:02:03,784 --> 00:02:07,434
got pulled over and about six months
into that first out-of-college
44
00:02:07,454 --> 00:02:10,254
job into records management.
45
00:02:10,404 --> 00:02:15,384
So, the, this is precursors to EHRs
going back in the late nineties and
46
00:02:15,384 --> 00:02:20,974
there was the document flow and the
fun thing about that one, I was young
47
00:02:20,974 --> 00:02:21,894
and got to travel on the country.
48
00:02:21,894 --> 00:02:22,454
That was fun.
49
00:02:22,454 --> 00:02:27,044
But, I really got a flavor for the
business backend side of healthcare.
50
00:02:27,104 --> 00:02:30,644
I originally thought I was going to Med
School, actually got accepted to one,
51
00:02:30,644 --> 00:02:33,824
decided not to do it, but always had
an interest in healthcare in general.
52
00:02:34,274 --> 00:02:37,064
And I realized there's multiple
areas that people can help with.
53
00:02:37,064 --> 00:02:40,604
And the really fun part that I thought
about then, but really thinking about more
54
00:02:40,604 --> 00:02:44,354
about now, is how much it affects your
everyday life and how you're influencing
55
00:02:44,354 --> 00:02:46,791
something that influences a lot of
people around that you love, admire.
56
00:02:46,791 --> 00:02:49,911
And some people that you just want to be,
you know, better humans going forward.
57
00:02:50,581 --> 00:02:53,431
For listeners who may not be familiar,
can you explain a little bit about
58
00:02:53,431 --> 00:02:57,151
what CommonWell is and the role
that you all play in this industry?
59
00:02:58,197 --> 00:02:58,437
Yeah,
60
00:02:58,437 --> 00:02:58,647
Given
61
00:02:58,647 --> 00:03:01,713
the best generic part of healthcare
is fun to play with 'cause you get to
62
00:03:01,713 --> 00:03:03,393
influence stuff and make things better.
63
00:03:03,443 --> 00:03:06,043
Interoperability i s
the focus of CommonWell.
64
00:03:06,043 --> 00:03:11,533
CommonWell was founded 2013, so a little
over a decade ago, by some preeminent
65
00:03:11,603 --> 00:03:16,523
electronic health record and other health
information technology companies with
66
00:03:16,523 --> 00:03:24,113
the belief that data availability and
exchange across providers was necessary
67
00:03:24,113 --> 00:03:26,963
for a better person-centric healthcare.
68
00:03:27,948 --> 00:03:32,688
And the original philosophy was
that it can't be a, or shouldn't be,
69
00:03:32,748 --> 00:03:34,488
shouldn't be a single vendor solution.
70
00:03:35,038 --> 00:03:38,878
'Cause patients see many different
providers, different types of specialties,
71
00:03:39,238 --> 00:03:42,928
and they're not picking by EHR,
they're picking by many other factors.
72
00:03:43,348 --> 00:03:48,318
So we were built with as a 501(c)(6),
a trade association nonprofit type.
73
00:03:49,008 --> 00:03:52,338
And our goal was to make sure all
the data was available every time
74
00:03:52,338 --> 00:03:56,623
a provider needs it across every
healthcare entity across the country.
75
00:03:57,103 --> 00:04:01,023
And we continue to strive to do that
with both within ourselves and across
76
00:04:01,023 --> 00:04:05,403
partners to move data from provider to
provider and provider to patient and
77
00:04:05,733 --> 00:04:08,253
keep adding Ps, payers, public health.
78
00:04:08,303 --> 00:04:09,233
Keep going down the list.
79
00:04:10,265 --> 00:04:13,145
One of the things that was interesting
when we were chatting last week in
80
00:04:13,145 --> 00:04:16,655
preparation for this podcast is that
you were talking about how CommonWell
81
00:04:16,985 --> 00:04:20,735
has been on the forefront of enabling
patient access to their own data for
82
00:04:20,735 --> 00:04:25,110
quite some time, which is, and currently
a big topic within interoperability.
83
00:04:25,110 --> 00:04:28,050
So can you explain a little bit about
what you've been doing there and why
84
00:04:28,050 --> 00:04:31,285
it's taken so long to get more visibility
and focus on that sort of work?
85
00:04:32,031 --> 00:04:32,241
Yeah,
86
00:04:32,241 --> 00:04:33,381
we started 2013.
87
00:04:33,381 --> 00:04:36,531
It really was a treatment centric
network provider to provider exchange.
88
00:04:36,581 --> 00:04:40,781
It was 2016 when we first added
what we called, patient access.
89
00:04:41,471 --> 00:04:44,951
And sometimes in the standards
called request, patient request.
90
00:04:45,581 --> 00:04:49,181
And I'll be honest, the first couple
years didn't go amazingly well.
91
00:04:49,211 --> 00:04:50,901
We had two general headwinds.
92
00:04:50,941 --> 00:04:54,131
One, there wasn't that much the
tracks laid for provider-provider.
93
00:04:54,151 --> 00:05:00,491
We thought we were doing well in 2016 with
a high number of thousands of connections.
94
00:05:00,921 --> 00:05:05,421
By the time I got here, and we go about
two years in, so we're around 2021,
95
00:05:05,421 --> 00:05:07,367
we were at tens of thousands, right?
96
00:05:07,367 --> 00:05:09,827
So we added in almost a factor
from that point forward.
97
00:05:10,337 --> 00:05:14,327
And the reality is most patient access
applications, or individual access as
98
00:05:14,327 --> 00:05:18,347
we like to call it now, or individual
access services, they don't work well
99
00:05:18,347 --> 00:05:19,637
if you can't get to all the data.
100
00:05:20,517 --> 00:05:23,867
Because the patient is looking
for their most recent thing.
101
00:05:23,867 --> 00:05:26,447
It's a very atomic level entity.
102
00:05:26,447 --> 00:05:31,217
It's provider Bob, provider John,
provider Lucy, that has my data or
103
00:05:31,277 --> 00:05:35,567
hospital X, Y, or Z. And when that
one's missing, they kind of disengage.
104
00:05:35,627 --> 00:05:38,237
So you needed more of a network
effect to make this work.
105
00:05:38,717 --> 00:05:44,627
The second part was the network itself
was designed for provider to provider,
106
00:05:45,017 --> 00:05:48,047
and we didn't actually write our
agreements to give this data to patients.
107
00:05:48,387 --> 00:05:51,507
And those of you who are HIPAA experts
and know things about healthcare
108
00:05:51,507 --> 00:05:53,787
data, you don't move data to things.
109
00:05:53,787 --> 00:05:57,027
You're not allowed to move them to
without a contract staying as such.
110
00:05:57,027 --> 00:05:59,487
So we had to do a fair amount of
retrofitting to get things up to
111
00:05:59,487 --> 00:06:02,067
snuff to be able to do patient access.
112
00:06:02,487 --> 00:06:07,587
Fortunately, the rest of the market
or at least the industry or us as just
113
00:06:07,587 --> 00:06:12,057
citizens and active healthcare users said,
Hey, we really want access to our data.
114
00:06:12,567 --> 00:06:15,657
And more and more people heard that
and it started to appear in other
115
00:06:15,657 --> 00:06:19,437
networking frameworks, which increased,
it increased the ability for us to
116
00:06:19,437 --> 00:06:20,997
take the time to focus and fix it.
117
00:06:21,517 --> 00:06:25,177
So we recently did that with one
of our largest members being Athena
118
00:06:25,537 --> 00:06:31,427
just moved in their entire cohort of
providers into our patient access aware
119
00:06:31,427 --> 00:06:33,497
application that is also TEFCA aware.
120
00:06:33,902 --> 00:06:37,182
And now we're off to the races with
people, I wouldn't say playing catch-up,
121
00:06:37,202 --> 00:06:39,502
I think we're all clotting to the same
point at the same time, which is great.
122
00:06:41,164 --> 00:06:43,894
So, Paul, I know a lot of us in
the industry have been tracking the
123
00:06:43,924 --> 00:06:47,524
recent administration's movements
around health IT, and specifically
124
00:06:47,524 --> 00:06:53,224
watching the recent RFI and events
that have been held by CMS and ONCASTP.
125
00:06:53,464 --> 00:06:56,104
I think that you've been involved with
some of those events in Washington.
126
00:06:56,104 --> 00:06:58,794
Do you wanna tell us a little bit
about the kind of tone and conversation
127
00:06:58,844 --> 00:07:02,349
going on in there and what you expect
to come outta those of those meetings?
128
00:07:03,114 --> 00:07:05,594
Yeah, there are really three
major events that occurred.
129
00:07:05,654 --> 00:07:12,524
One goes back to May where CMS
released an RFI, looking for commentary
130
00:07:12,524 --> 00:07:17,354
on the benefits of health IT and
what are good apps out there, as
131
00:07:17,354 --> 00:07:20,484
well as other things related to
technology and data interoperability.
132
00:07:21,004 --> 00:07:24,874
It culminated in a, well, actually the
culminated was the end of the RFI the
133
00:07:24,874 --> 00:07:27,544
day you had to submit everything, but in
the middle they did a listening session
134
00:07:28,199 --> 00:07:32,399
where they had the, the higher-ups of
HHS, so this was Secretary Kennedy,
135
00:07:32,399 --> 00:07:36,329
then Dr. Oz from CMS side, down the
line explaining what's going on.
136
00:07:36,329 --> 00:07:37,499
Amy Gleason, et cetera.
137
00:07:38,084 --> 00:07:41,984
And the morning session was presentation,
followed by a bunch of panels in the
138
00:07:41,984 --> 00:07:44,504
afternoon was more working sessions
to figure out what to do next.
139
00:07:45,104 --> 00:07:48,974
The morning panels were full of
people just banging the floor
140
00:07:48,974 --> 00:07:50,594
looking for better patient access.
141
00:07:50,644 --> 00:07:54,304
And you have to think about this for
a second, realize that the people
142
00:07:54,334 --> 00:07:58,204
putting together this listening
session had to have set a tone.
143
00:07:58,204 --> 00:07:59,584
They chose who to put up there.
144
00:07:59,644 --> 00:07:59,944
Right.
145
00:07:59,944 --> 00:08:02,464
And there were people they knew
that were gonna say these things.
146
00:08:02,834 --> 00:08:04,694
You don't have to, you don't
have to wind people up and
147
00:08:04,694 --> 00:08:05,624
tell 'em exactly what to say.
148
00:08:05,624 --> 00:08:06,614
Just get the right people.
149
00:08:06,914 --> 00:08:09,824
So the fact that CMS and HHS put a
lot of people they knew they were
150
00:08:09,824 --> 00:08:14,834
gonna be focused on individual
access on a stage during their big
151
00:08:14,834 --> 00:08:17,504
show is interesting unto itself.
152
00:08:17,504 --> 00:08:19,399
That says the administration
is very interested.
153
00:08:19,979 --> 00:08:24,389
And then the conversation that happened
after that, which then led to this
154
00:08:24,389 --> 00:08:28,779
idea of a CMS aligned network and a
further event that happened a couple
155
00:08:28,779 --> 00:08:33,399
weeks ago at the White House where
the CEOs and other officers of many of
156
00:08:33,399 --> 00:08:36,129
the companies are at that event, and
actually a smaller group than that,
157
00:08:36,639 --> 00:08:41,664
were pledging to be aligned to CMS as
we go forward with a significant focus
158
00:08:41,664 --> 00:08:46,344
of that being, how do we help patients
make better decisions with better data?
159
00:08:46,864 --> 00:08:51,664
And I think in the first administration
round of Trump, you saw financial data.
160
00:08:51,664 --> 00:08:55,404
There was an attempt to do getting
price information, price transparency
161
00:08:55,404 --> 00:08:58,284
to patients, which I don't think went
as well as everybody wanted it to,
162
00:08:58,284 --> 00:09:01,044
but it was an interesting idea of
more information, better decisions.
163
00:09:01,404 --> 00:09:04,614
And I think this is the next evolution
to how about the clinical data, too?
164
00:09:04,614 --> 00:09:07,254
Maybe the two together
start to move the needle?
165
00:09:07,294 --> 00:09:10,669
Understanding that the first
tools probably won't do as much
166
00:09:10,669 --> 00:09:13,399
as you wanted to, but you have to
start every journey with a step.
167
00:09:13,819 --> 00:09:16,939
And a significant step would be
getting the ability for apps to
168
00:09:16,939 --> 00:09:19,519
get data in patients' hands so they
can start doing something with it.
169
00:09:20,783 --> 00:09:24,983
You mentioned Amy Gleason and just then,
um, who was the acting head of the newly
170
00:09:24,983 --> 00:09:28,563
formed department of government efficiency
or DOGE, under the federal government.
171
00:09:28,573 --> 00:09:31,678
And obviously she comes from a
background in, in healthtech.
172
00:09:31,868 --> 00:09:35,148
So I'm curious for your perspective
on someone like her, who's known
173
00:09:35,148 --> 00:09:36,648
for really getting things done,
174
00:09:36,868 --> 00:09:38,848
how do you think her tenure
and involvement in this
175
00:09:38,848 --> 00:09:40,118
will impact the outcome?
176
00:09:41,328 --> 00:09:42,038
Well, there's two things.
177
00:09:42,038 --> 00:09:45,328
She has she has a good patient story
in her daughter's story, which is
178
00:09:45,328 --> 00:09:46,838
for those who have not seen it.
179
00:09:47,488 --> 00:09:50,848
I won't give you a full summary,
but you can find it on the interweb
180
00:09:50,878 --> 00:09:53,008
looking up Amy Gleason story.
181
00:09:53,008 --> 00:09:56,278
Amy Gleason, CMS, YouTube,
whatever you wanna find it on.
182
00:09:56,548 --> 00:10:00,448
And she does a good job of delineating
a, you could think of a generic patient
183
00:10:00,448 --> 00:10:06,568
with a rare disease and without data
and AI tools and the like to be able to
184
00:10:06,568 --> 00:10:09,748
process that data, how things were missed.
185
00:10:09,878 --> 00:10:14,738
And I think there's a lesson in there
that, providers, i n the past you
186
00:10:14,738 --> 00:10:18,508
would, you have seen some resistance
to some adoption of technology.
187
00:10:18,508 --> 00:10:21,948
For example, we saw an initial bad
reaction to like direct messaging,
188
00:10:21,998 --> 00:10:25,238
provider to provider direct, but
then they tried applying patient
189
00:10:25,238 --> 00:10:29,768
to provider and we skipped that and
went instead to portal conversations.
190
00:10:29,768 --> 00:10:33,518
And when you start to learn as people
are generally pretty nice, people don't
191
00:10:33,518 --> 00:10:36,218
abuse your, you know, my, my mobile
phone number's in my signature line for
192
00:10:36,218 --> 00:10:38,498
my email, and no one calls it, right.
193
00:10:38,498 --> 00:10:39,878
, No one bothers people that way.
194
00:10:39,878 --> 00:10:44,038
They have other ways of communicating
and I think they realized after that push
195
00:10:44,038 --> 00:10:47,938
of communication direct to patient, that
they're not gonna get abused as much.
196
00:10:47,938 --> 00:10:50,998
Maybe it's time to think about using
technology in a more augmented way.
197
00:10:51,538 --> 00:10:54,418
But Amy in particular, is
like, no, go do it faster.
198
00:10:54,518 --> 00:10:59,153
You're hurting healthcare, you're
hurting yourself actually by not allowing
199
00:10:59,153 --> 00:11:01,013
patients to supplement what you're doing.
200
00:11:01,453 --> 00:11:09,583
And her brain in terms of doing things
faster but efficiently and effectively,
201
00:11:10,003 --> 00:11:13,723
on top of the passion to do it from her
own personal story, is usually the kind
202
00:11:13,723 --> 00:11:15,373
of secret sauce you need to move things.
203
00:11:15,823 --> 00:11:20,788
And you put someone in a central position
where they can convene people and blast
204
00:11:20,788 --> 00:11:24,418
through what I, I usually call the first
mover disadvantage of network effects, the
205
00:11:24,418 --> 00:11:28,188
first guy does all the work and a lot of
people follow in after, i s significant.
206
00:11:28,188 --> 00:11:33,428
Like that kind of the beat of the drum
being consistent across many parties who
207
00:11:33,428 --> 00:11:37,868
normally compete and don't wanna be synced
with each other is very powerful when
208
00:11:37,873 --> 00:11:39,278
you're trying to develop network of x.
209
00:11:40,467 --> 00:11:44,187
Speaking of building out networks, one
of the big developments that I've really
210
00:11:44,187 --> 00:11:47,127
tracked from, from my side and the
customers that we worked with in the last
211
00:11:47,127 --> 00:11:52,317
couple years is the rollout of the On
Behalf Of designation with Carequality,
212
00:11:52,337 --> 00:11:56,157
which I understand that you had a hand
in originally developing that concept.
213
00:11:56,157 --> 00:11:59,377
Can you tell us a little bit about, on
behalf of our OBO and a little bit of
214
00:11:59,377 --> 00:12:00,912
the story about how that came about?
215
00:12:02,877 --> 00:12:03,507
Yeah.
216
00:12:03,557 --> 00:12:05,447
I wanna see healthcare move faster too.
217
00:12:05,507 --> 00:12:09,537
And you can imagine running a trade
alliance, which is supposed to be
218
00:12:09,957 --> 00:12:14,727
supportive of trade and not just the
entries, entities are in it right now.
219
00:12:14,777 --> 00:12:17,597
It's supposed to be also supportive
of the next thing to enter.
220
00:12:18,137 --> 00:12:20,177
I kept thinking of what
are we missing, right?
221
00:12:20,177 --> 00:12:25,052
If I just, if all we did was work with our
primary EHRs and the like, t hat's great.
222
00:12:25,082 --> 00:12:27,902
They'll fund the enterprise and
we're a true nonprofit, we make
223
00:12:27,902 --> 00:12:28,862
too much money, we give it back.
224
00:12:28,862 --> 00:12:30,272
I was like, who else needs help?
225
00:12:30,872 --> 00:12:35,882
And I kept looking to the startups
and the neat little things being
226
00:12:35,882 --> 00:12:39,572
developed at the edge that are
the first step to their existence.
227
00:12:39,572 --> 00:12:42,712
And I had done some accelerator
type things in my past as well.
228
00:12:43,072 --> 00:12:44,962
And how did you do an
accelerator in healthcare?
229
00:12:44,962 --> 00:12:47,392
Usually it was by getting
your first customer, right?
230
00:12:47,392 --> 00:12:50,932
You would link them up with a technology
set and say, Hey, I'm bringing this
231
00:12:50,932 --> 00:12:55,822
preeminent academic in as a pilot customer
to help you develop and go forward and
232
00:12:55,822 --> 00:12:57,502
then get more contacts and go from there.
233
00:12:57,982 --> 00:13:00,592
But when you start looking
at patient stuff and littler
234
00:13:00,712 --> 00:13:03,622
tools, it doesn't fit as well.
235
00:13:03,682 --> 00:13:07,952
Like we're talking like, one
person, one diabetic, we need to
236
00:13:07,952 --> 00:13:09,342
move faster and get things going.
237
00:13:09,342 --> 00:13:13,572
So I said, why i f providers are hooking
up their EHRs at scale, and we're getting
238
00:13:13,572 --> 00:13:17,532
the data to the network, we're solving
the data availability problem, right?
239
00:13:17,532 --> 00:13:19,002
The data is going to be there.
240
00:13:19,397 --> 00:13:23,547
What I was missing was, who else needs
to use this data, the supply and demand.
241
00:13:24,147 --> 00:13:27,657
I said, can we hook up these
applications to the network?
242
00:13:28,137 --> 00:13:31,497
To get the same data to accelerate their
ability to do what they need to do.
243
00:13:31,497 --> 00:13:35,607
Iinstead of integrating to one
hospital or to one ambulatory practice,
244
00:13:35,637 --> 00:13:39,057
why don't they hook up to all of them
and do that without having to hook up
245
00:13:39,057 --> 00:13:42,177
through the high primary EHR 'cause
it's expensive and time-consuming?
246
00:13:42,687 --> 00:13:45,147
And I was more worried about the
time-consuming, to be honest.
247
00:13:45,147 --> 00:13:48,157
I mean, it's fine for, again, we're
a trade association we support trade.
248
00:13:48,517 --> 00:13:51,127
I'm not against people charging
for services that's fine.
249
00:13:51,557 --> 00:13:52,907
You just don't want it to be a barrier.
250
00:13:53,612 --> 00:13:56,042
And so I felt like we weren't getting the
right tools in front of the right people.
251
00:13:56,282 --> 00:13:57,242
So we built this OBO.
252
00:13:57,242 --> 00:14:01,487
If your EHR, if your EHR is already
connected, why can't you connect
253
00:14:01,487 --> 00:14:04,217
every other product you bought
that is for treatment purposes?
254
00:14:04,337 --> 00:14:04,847
All of them.
255
00:14:05,327 --> 00:14:08,597
In the dermatology app that takes a neat
picture from your iPhone and sends it up
256
00:14:08,597 --> 00:14:12,017
to some AI tool to do a skin lesion check.
257
00:14:12,317 --> 00:14:15,257
Why shouldn't it have access
to your history of skin cancer,
258
00:14:15,257 --> 00:14:16,427
family history, and the like?
259
00:14:16,667 --> 00:14:20,417
To better inform the answer of what it's
about, to help suggest in that clinical
260
00:14:20,417 --> 00:14:21,887
decision support thing it's doing.
261
00:14:22,277 --> 00:14:23,297
And do it cheaply.
262
00:14:23,387 --> 00:14:26,627
And by cheaply we make the
network an interface engine.
263
00:14:26,627 --> 00:14:29,957
Everybody can access the same
information at the atomic level of
264
00:14:29,957 --> 00:14:31,757
you, Brendan, Paul, whatever it is.
265
00:14:32,307 --> 00:14:37,667
There was some resistance at first
and unfortunately it got changed a
266
00:14:37,667 --> 00:14:39,857
little bit in the way it got word
that people found loopholes to
267
00:14:39,857 --> 00:14:41,927
use it for, semi-abusive things.
268
00:14:41,927 --> 00:14:44,807
And we were still addressing
that trust issue today.
269
00:14:45,767 --> 00:14:49,817
But I still stand behind the idea
that we need more access, not less.
270
00:14:50,187 --> 00:14:53,747
And we just need to do it safely and
securely so that, we maintain trust and
271
00:14:53,747 --> 00:14:55,307
provide the right solutions to patients.
272
00:14:56,699 --> 00:15:00,149
Paul, when we think about the future
of patient access to their own data,
273
00:15:00,149 --> 00:15:03,419
obviously there's a lot of different
applications for that, including
274
00:15:03,479 --> 00:15:07,019
data correction, which was one that I
have experienced as a patient myself.
275
00:15:07,019 --> 00:15:11,309
You know, seeing things in my own patient
record that is, is incorrect, and then
276
00:15:11,309 --> 00:15:13,679
trying to take steps to address that.
277
00:15:13,889 --> 00:15:18,119
Why is that specifically such an
important aspect of kind of unlocking
278
00:15:18,119 --> 00:15:22,379
data for patients that ultimately they,
they should have the right to use?
279
00:15:24,194 --> 00:15:24,404
Yeah.
280
00:15:24,404 --> 00:15:28,784
At the risk of giving the standard answer
at every technology conference, and
281
00:15:28,784 --> 00:15:32,804
especially healthcare conferences, AI, so
we got this artificial intelligence thing
282
00:15:32,804 --> 00:15:39,014
going on, and I think it's important to
remember how AI actually works, right?
283
00:15:39,044 --> 00:15:40,064
It's statistic space, right?
284
00:15:40,064 --> 00:15:43,914
It's taking... when you do an LLM
and have it write an RFP for you
285
00:15:43,994 --> 00:15:45,704
in Copilot, or whatever you use,
286
00:15:46,184 --> 00:15:50,484
it's just aware of what kind of words
go together when it's, look through
287
00:15:50,484 --> 00:15:53,514
the history of time of everybody
writing things and gives you a
288
00:15:53,514 --> 00:15:56,724
suggestion that seems to be within the
confines of what you're looking for.
289
00:15:57,264 --> 00:15:59,814
And when you get to healthcare
data, it's doing the same thing.
290
00:15:59,844 --> 00:16:04,564
Going back to the days of Watson and
other early precursors where we are today.
291
00:16:05,344 --> 00:16:08,524
The one thing you don't
wanna do is confuse it.
292
00:16:08,864 --> 00:16:12,164
We, you'll hear people talking
about AI hallucinations and it
293
00:16:12,164 --> 00:16:13,664
starts to fill in data gaps.
294
00:16:14,174 --> 00:16:18,494
Another problem is when it has a
piece of data that doesn't belong
295
00:16:18,524 --> 00:16:21,464
and it doesn't always know what
should go around it or ignore it.
296
00:16:21,974 --> 00:16:24,884
So if you have errors,
bad things can happen.
297
00:16:24,884 --> 00:16:26,054
Lemme give you an example.
298
00:16:26,104 --> 00:16:32,284
I had a shot, a flu shot last
season and my record said I
299
00:16:32,284 --> 00:16:33,994
had gotten HPV vaccination.
300
00:16:34,714 --> 00:16:39,544
Now, it is actually more common these
days for or for especially younger males
301
00:16:39,544 --> 00:16:43,804
to get HPV vaccinations, but not very
common for a 48-year-old male to do that.
302
00:16:44,464 --> 00:16:50,554
And my doctor would probably look over
that, if he was looking at fantastic
303
00:16:50,554 --> 00:16:53,484
cancer drugs that would be the
all-end-all to cure prostate cancer
304
00:16:53,544 --> 00:16:58,014
it's available and says, but what if
that prostate cancer drug material
305
00:16:58,014 --> 00:17:01,254
said it's contraindicated to people
who have the HPV vaccination?
306
00:17:01,864 --> 00:17:05,104
My provider might skip it, but the
AI provider may not know what to do.
307
00:17:05,614 --> 00:17:07,294
And so it's statistical analysis.
308
00:17:07,294 --> 00:17:09,504
We're gonna start suggesting
different therapies.
309
00:17:10,224 --> 00:17:12,654
Could underweight the perfect drug, right?
310
00:17:12,654 --> 00:17:16,074
Everything else could fit in, but
this thing is a bang-out criteria and
311
00:17:16,074 --> 00:17:18,774
bang-out means eliminate it at all costs.
312
00:17:19,224 --> 00:17:20,064
You don't want that.
313
00:17:20,134 --> 00:17:23,314
And there are various things that,
little tiny nuances, little nudges you
314
00:17:23,314 --> 00:17:25,324
can do that make the model messed up.
315
00:17:25,864 --> 00:17:27,474
Providers can do the same thing.
316
00:17:27,684 --> 00:17:29,854
They already have clinical
decision support tools on their
317
00:17:29,854 --> 00:17:34,114
side, and contraindications X is
contraindicated to Y. You can't
318
00:17:34,114 --> 00:17:35,404
have these two things together.
319
00:17:35,809 --> 00:17:37,249
Automatically set off red flags.
320
00:17:37,999 --> 00:17:38,239
Right?
321
00:17:38,239 --> 00:17:40,279
So that would be, that'd be an
extreme example, but it could
322
00:17:40,279 --> 00:17:43,009
also say there is no data on this.
323
00:17:43,069 --> 00:17:46,639
Anything that decreases that
Superdrug now being not a Superdrug,
324
00:17:46,644 --> 00:17:49,894
because of my potential biochemistry
and the things I have in it.
325
00:17:50,284 --> 00:17:52,594
So I think it's really important
that we have the right data.
326
00:17:52,594 --> 00:17:55,714
I mean, today, most of us probably
worry about the wrong data being
327
00:17:55,714 --> 00:17:57,634
things we get billed for, right?
328
00:17:57,634 --> 00:17:58,654
I didn't have this done.
329
00:17:58,654 --> 00:18:00,034
I don't want to pay my $20 co-pay.
330
00:18:00,034 --> 00:18:02,044
This was supposed to be an annual
exam, why do I have a co-pay?
331
00:18:02,044 --> 00:18:02,794
It's supposed to be free.
332
00:18:03,244 --> 00:18:04,624
And those guy, oh, it's coded wrong.
333
00:18:04,624 --> 00:18:05,554
You like, okay, you fix it.
334
00:18:05,554 --> 00:18:08,824
Fix it to an annual wellness exam
and not just a normal office visit.
335
00:18:09,184 --> 00:18:11,554
And that's fine for the $20,
a hundred dollars, a couple
336
00:18:11,554 --> 00:18:12,994
thousand dollars even, you pay.
337
00:18:13,444 --> 00:18:17,254
But the clinical stories could get
really messed up if we confuse the AIs.
338
00:18:17,644 --> 00:18:20,359
And it is, you could argue
whether it's the future or not,
339
00:18:20,359 --> 00:18:21,439
but it's definitely a factor.
340
00:18:21,839 --> 00:18:25,799
So getting data into the hands of
patients is the most likely way
341
00:18:25,799 --> 00:18:27,119
to make sure the data is accurate.
342
00:18:27,449 --> 00:18:29,939
'cause I probably care more my
data than anybody else does.
343
00:18:32,204 --> 00:18:36,674
And on the topic of AI, Paul, as we
continue to see, meaningful advancement
344
00:18:36,734 --> 00:18:40,484
toward making data more liquid for
different use cases, involving including
345
00:18:40,484 --> 00:18:43,904
those that patients are directing,
how do we make sure that patients
346
00:18:43,964 --> 00:18:48,374
stay protected or that f olks' privacy
is respected in this whole process.
347
00:18:50,729 --> 00:18:53,859
You know, I saw, we talked about
the events that happened and then
348
00:18:53,859 --> 00:18:54,894
there was the big White House event.
349
00:18:55,674 --> 00:18:58,584
And I don't, it doesn't really
matter where you sit in the political
350
00:18:58,584 --> 00:19:02,604
spectrum, I did see something disturbing
happen that happened that, that day.
351
00:19:03,024 --> 00:19:07,439
There was an article put out by a, what I
consider a good media outlet, that talked
352
00:19:07,439 --> 00:19:11,099
about the federal government is trying to
make a system to collect all your data.
353
00:19:11,654 --> 00:19:15,854
Big tech is getting into your data,
be careful and got quotes from some
354
00:19:16,094 --> 00:19:18,804
preeminent professors about the
security and privacy concerns alike.
355
00:19:19,334 --> 00:19:24,769
And the irony there was the dialogue
was exactly opposite that, right?
356
00:19:24,769 --> 00:19:29,479
The dialogue was, we want big
tech to do the right thing, to get
357
00:19:29,479 --> 00:19:34,279
data into your hands so that you
can control where it goes, right?
358
00:19:34,279 --> 00:19:36,169
We're not trying to
collect it for ourselves.
359
00:19:36,469 --> 00:19:40,419
We're trying to collect it on your behalf
and then let go of it, right, put it
360
00:19:40,419 --> 00:19:45,949
into the storage thing of your choice,
into the application of your choice.
361
00:19:45,949 --> 00:19:47,449
And you do what you wanna do.
362
00:19:47,449 --> 00:19:49,669
If you wanna be in a research
study, you release it.
363
00:19:49,669 --> 00:19:49,849
Right?
364
00:19:49,849 --> 00:19:51,469
That's not, we're trying
to get in the middle.
365
00:19:51,919 --> 00:19:56,689
So I think we've banged around the
edges of security and privacy for a
366
00:19:56,689 --> 00:20:00,989
long time, but I think the current push
right now has the right philosophy.
367
00:20:01,679 --> 00:20:07,034
The danger of securing privacy
around healthcare data is, you
368
00:20:07,034 --> 00:20:11,924
gotta be really careful just not
disclosing data that might help you.
369
00:20:12,104 --> 00:20:14,684
You kind of have to trust your
provider to do the right thing.
370
00:20:15,194 --> 00:20:18,374
And we, if we go too far, making
it locked up, that you have to
371
00:20:18,374 --> 00:20:22,814
anatomically every piece of data one
element at a time, I think we are
372
00:20:22,814 --> 00:20:24,554
actually gonna negatively affect care.
373
00:20:25,089 --> 00:20:30,279
Because patients have not been given their
data for so long, they don't quite know
374
00:20:30,279 --> 00:20:31,869
what's important and what's not important.
375
00:20:32,529 --> 00:20:36,579
And so, if you start scaring people that
your data's flying around everywhere, and
376
00:20:36,579 --> 00:20:40,239
then say you have control, the a knee-jerk
reaction could be to not give it to
377
00:20:40,239 --> 00:20:45,009
anybody and you could be doing yourself
serious harm, which actually is a lesson.
378
00:20:45,069 --> 00:20:48,279
The lesson here is get the
data to the patients faster.
379
00:20:48,609 --> 00:20:50,589
We have to get past that episode.
380
00:20:50,589 --> 00:20:54,039
We have to get past that, that
fear bubble as quickly as possible
381
00:20:54,099 --> 00:20:55,389
to get back to the right balance.
382
00:20:55,389 --> 00:20:58,179
And we see in everything,
there's often an overshoot.
383
00:20:58,179 --> 00:21:01,899
I'm, I would not be surprised, if
we give patients all their data and
384
00:21:01,899 --> 00:21:05,469
we give them audit capabilities to
know where it's flying that many
385
00:21:05,469 --> 00:21:07,029
people will shut it down first.
386
00:21:07,029 --> 00:21:08,659
'cause they don't understand
how to control it.
387
00:21:08,779 --> 00:21:10,459
And then we had to bring it back up.
388
00:21:10,579 --> 00:21:11,659
But that's gonna happen.
389
00:21:11,659 --> 00:21:12,649
So you might as well do it now.
390
00:21:12,679 --> 00:21:13,699
'cause doing it later is worse.
391
00:21:15,374 --> 00:21:19,994
And also on the topic of data liquidity as
I think we talked about, there's a number
392
00:21:19,994 --> 00:21:23,144
of different factors in that beyond just
the technology factors that have made
393
00:21:23,144 --> 00:21:27,744
that challenging in healthcare to date,
but w ith the advancement and adoption
394
00:21:27,804 --> 00:21:31,944
and rollout of FHIR by different vendors,
we're seeing some kind of other technology
395
00:21:31,944 --> 00:21:33,964
pieces of that being addressed as well.
396
00:21:33,964 --> 00:21:36,204
So what's your kind of vantage
point on the current state of
397
00:21:36,304 --> 00:21:39,644
FHIR rollout and where that stands
to advance in the coming years?
398
00:21:40,889 --> 00:21:42,859
FHIR has finally taken some leaps forward.
399
00:21:42,909 --> 00:21:43,959
I do think it's interesting.
400
00:21:43,959 --> 00:21:46,599
You mentioned security and privacy, right,
went right to FHIR because I think they're
401
00:21:46,599 --> 00:21:51,339
actually very linked in a way that isn't
always obvious to the average consumer.
402
00:21:51,389 --> 00:21:54,539
The primary link here is there's
a fair amount of security and
403
00:21:54,539 --> 00:21:57,959
privacy control embedded in HIPAA
and other state and federal laws.
404
00:21:58,664 --> 00:21:59,984
About minimal necessary.
405
00:21:59,989 --> 00:22:03,719
Don't share more than you need to
particularly when you go outside
406
00:22:03,719 --> 00:22:06,929
our provider relationship, right,
so that be to your payer, it could
407
00:22:06,929 --> 00:22:08,279
be to public health and whatever.
408
00:22:08,849 --> 00:22:12,749
And sending a document with all
your data, it makes it difficult to
409
00:22:12,749 --> 00:22:14,429
redact what I'm not supposed to send.
410
00:22:15,099 --> 00:22:17,309
Imagine you're getting a mortgage.
411
00:22:17,714 --> 00:22:21,434
And the bank requires to know
your current salary and they wanna
412
00:22:21,434 --> 00:22:24,764
know it from your gross income
from last year's, uh, tax return.
413
00:22:25,334 --> 00:22:30,179
Sending your 110 page tax return to
get line one is a little silly, right?
414
00:22:30,179 --> 00:22:34,049
That's a way over disclosure of
a lot of things about you when
415
00:22:34,049 --> 00:22:35,309
they only needed one number.
416
00:22:35,819 --> 00:22:39,209
And today most exchange is
being done via documents.
417
00:22:39,209 --> 00:22:42,149
These large things with a
whole bunch of data in them.
418
00:22:42,719 --> 00:22:45,869
It makes it much more difficult
to assert this is the minimal
419
00:22:45,869 --> 00:22:47,159
necessary for the purpose.
420
00:22:47,579 --> 00:22:50,094
And I think that is a huge
advantage of FHIR there.
421
00:22:50,159 --> 00:22:53,009
There are public health departments
out there that are on the sidelines.
422
00:22:53,549 --> 00:22:54,209
During COVID.
423
00:22:54,989 --> 00:22:59,114
During COVID, we allowed public health
departments over Carequality and
424
00:22:59,114 --> 00:23:03,774
therefore over CommonWell as well to
access data about COVID patients that
425
00:23:03,774 --> 00:23:07,704
they were following up with through
the networks using the treatment lane.
426
00:23:07,824 --> 00:23:08,664
And I say the treatment lane.
427
00:23:08,664 --> 00:23:11,364
'cause the treatment lane, the
treatment use case is the most
428
00:23:11,364 --> 00:23:13,404
open public health, not as open.
429
00:23:13,404 --> 00:23:14,784
So they said, just go through
the treatment lane, we
430
00:23:14,784 --> 00:23:15,804
know it's you, it's cool.
431
00:23:16,404 --> 00:23:17,664
And they said, what about
the minimal necessary?
432
00:23:17,664 --> 00:23:21,864
I said, you tell us this data is minimal
necessary for this, and we're good.
433
00:23:22,734 --> 00:23:26,044
Only two states figured out how
to talk to their AGs to figure out
434
00:23:26,044 --> 00:23:27,424
they're allowed to make that request.
435
00:23:27,844 --> 00:23:28,024
Right?
436
00:23:28,024 --> 00:23:31,354
That minimal necessary for this use
case was everything and will only keep.
437
00:23:31,684 --> 00:23:32,824
They didn't know what to do.
438
00:23:33,154 --> 00:23:38,314
So we need the ability to be more
atomic or more focused on the data
439
00:23:38,314 --> 00:23:41,434
we exchange 'cause everybody's
actually trying to do a pretty good
440
00:23:41,434 --> 00:23:42,994
job on the security and privacy.
441
00:23:43,414 --> 00:23:47,164
It's just hard to do when you have a blunt
force instrument of a massive document.
442
00:23:47,464 --> 00:23:49,234
So I think FHIR goes a long way there.
443
00:23:49,234 --> 00:23:52,324
It also allows for more current
information, 'cause a lot of
444
00:23:52,324 --> 00:23:55,894
documents are created at the end
of exams, at the end of encounters.
445
00:23:56,224 --> 00:23:57,394
There's a ton of advantages.
446
00:23:57,394 --> 00:24:02,704
I will say though I am not... FHIR is the
panacea solution for interoperability.
447
00:24:03,054 --> 00:24:05,964
It solves some problems and it helps
with some things, but we still have
448
00:24:05,964 --> 00:24:10,044
a lot of our things to work on,
including just laying the tracks and the
449
00:24:10,044 --> 00:24:14,724
philosophical change from holding data to
exposing data as the right thing to do.
450
00:24:17,334 --> 00:24:20,904
We've talked a lot about the kind of
current advancements in interoperability
451
00:24:20,904 --> 00:24:24,174
and the progress that's being made that
advances a number of different use cases.
452
00:24:24,939 --> 00:24:28,239
Here at Newfire, we obviously serve a
lot of different innovators, including
453
00:24:28,239 --> 00:24:30,909
those in the digital health space
vendors that just sell to health
454
00:24:30,909 --> 00:24:32,439
systems and payers and things like that.
455
00:24:32,439 --> 00:24:34,925
So, what's in this broad world of interop?
456
00:24:34,925 --> 00:24:39,635
Like one mistake that you see
organizations make most commonly when
457
00:24:39,635 --> 00:24:42,665
they approach data exchange that really
ends up hurting them in the long run?
458
00:24:44,295 --> 00:24:44,955
Waiting.
459
00:24:45,405 --> 00:24:47,265
I think waiting is the
worst thing you can do.
460
00:24:47,725 --> 00:24:52,695
There's this desire to say I'll jump in
when X. And for a while it was jumping
461
00:24:52,695 --> 00:24:55,185
in when there was enough data, right?
462
00:24:55,185 --> 00:24:58,850
If you went back when I first started
doing this, going back to 2009, 2010,
463
00:25:00,045 --> 00:25:03,405
the adoption of interoperability and
connecting provider-provider was pretty
464
00:25:03,405 --> 00:25:08,775
low to the point that if a provider
pressed the button, the likelihood they
465
00:25:08,775 --> 00:25:11,475
found data for the patient in front
of 'em was actually relatively low.
466
00:25:12,155 --> 00:25:13,415
Now the problem is flipped.
467
00:25:13,415 --> 00:25:14,735
There's too much data, right?
468
00:25:14,735 --> 00:25:17,555
So all these documents are flying
around and they do have a fair amount
469
00:25:17,555 --> 00:25:18,330
of duplication right within them.
470
00:25:18,335 --> 00:25:21,215
So it is sometime difficult to work with.
471
00:25:21,845 --> 00:25:25,715
But waiting for technology to make
that better, for example, waiting
472
00:25:25,715 --> 00:25:28,775
for FHIR, like I got news for you,
the documents we sent around today
473
00:25:29,225 --> 00:25:31,325
are pretty well coded, right?
474
00:25:31,325 --> 00:25:32,945
They're just, we were lazy.
475
00:25:33,005 --> 00:25:34,325
There's a lot of industries that do this.
476
00:25:34,325 --> 00:25:37,805
They bandaided knowing that the other
side may not have a good parser.
477
00:25:38,165 --> 00:25:41,495
We made a human readable format version
of it, and that's with all display on the
478
00:25:41,495 --> 00:25:44,375
screen you get a discharge summary, which
has a human readable displayed version.
479
00:25:44,975 --> 00:25:48,635
Underneath that human readable thing
is a machine readable version that
480
00:25:48,635 --> 00:25:53,255
has lab results and times, and all
the stuff you would need to do like a
481
00:25:53,255 --> 00:25:57,620
lab chart showing how things changed
over time, both inpatient and then
482
00:25:57,620 --> 00:26:00,530
before inpatient and after, you can
show the differences in between.
483
00:26:00,910 --> 00:26:03,820
But we've seen people now look
and go, there's too much data.
484
00:26:03,820 --> 00:26:05,780
Everybody's gotta organize it
better or I'm not coming in.
485
00:26:06,250 --> 00:26:10,300
And what I've seen decade over decade,
'cause I actually say that now I've
486
00:26:10,300 --> 00:26:13,810
been doing this for a long time, is
those who wait I wouldn't say they
487
00:26:13,810 --> 00:26:16,120
get hurt, but they lose advantage.
488
00:26:16,600 --> 00:26:19,590
And negative advantage is the
same thing as getting hurt
489
00:26:19,590 --> 00:26:20,760
when the market moves fast.
490
00:26:21,175 --> 00:26:25,315
So I would say get connected and then
start finding tools for the specific
491
00:26:25,315 --> 00:26:26,635
things that you're looking for.
492
00:26:27,005 --> 00:26:30,125
It's not like a sea-change
moment when you into interop.
493
00:26:30,485 --> 00:26:33,605
It's a incremental movement that
leads to other incremental movement
494
00:26:33,605 --> 00:26:35,385
movements that leads to general success.
495
00:26:37,365 --> 00:26:41,615
And on that topic uh, what are,
what's a capability or tool in
496
00:26:41,615 --> 00:26:45,455
the interop space that you feel is
most underutilized or overlooked by
497
00:26:45,455 --> 00:26:46,925
folks looking to do data exchange?
498
00:26:48,760 --> 00:26:52,830
There are two things that that are, that
apply there and I'm a little surprised
499
00:26:52,830 --> 00:26:54,390
'cause they seem almost obvious.
500
00:26:54,450 --> 00:26:57,640
And I'll be honest a fair amount
of entities have recognized
501
00:26:57,640 --> 00:26:58,660
this and are starting to fix it.
502
00:26:59,230 --> 00:27:01,800
But the data is large.
503
00:27:01,800 --> 00:27:03,480
The data sets are not tiny.
504
00:27:03,550 --> 00:27:05,020
They do take time to move.
505
00:27:05,560 --> 00:27:09,040
The movement of my data from one
provider or a collection provider
506
00:27:09,040 --> 00:27:12,415
to another is generally not a
good thing to do interactively.
507
00:27:12,505 --> 00:27:15,565
Interactively, meaning I click
the button and I see it come up.
508
00:27:15,625 --> 00:27:17,125
It takes a little bit of time.
509
00:27:17,635 --> 00:27:22,225
I'm amazed at how many EHRs that I
consider relatively mature, don't have
510
00:27:22,225 --> 00:27:27,235
a pre-access workflow, you know, like
a, a prefetch that is based on triggers.
511
00:27:27,685 --> 00:27:30,715
Most healthcare and a lot of people
think about healthcare as having,
512
00:27:30,765 --> 00:27:32,775
different types of encounters,
513
00:27:32,775 --> 00:27:35,505
you have a thing, your ambulatory,
your wellness exam every year, you have
514
00:27:35,505 --> 00:27:39,735
specialty stuff, you have inpatient that's
scheduled, then you have things that go
515
00:27:39,735 --> 00:27:41,055
through the emergency department alike.
516
00:27:41,595 --> 00:27:44,985
We often focus in interop, or at least in
the beginning, we focused on the ED stuff.
517
00:27:44,985 --> 00:27:46,635
I don't know anything about this patient
518
00:27:46,995 --> 00:27:49,215
so I'll hit this button to
go find something out about
519
00:27:49,215 --> 00:27:50,355
them before I start a workup.
520
00:27:50,715 --> 00:27:51,645
And that makes perfect sense.
521
00:27:51,645 --> 00:27:53,985
If the person's coming through the
ED, absolutely positively get their
522
00:27:53,985 --> 00:27:56,775
allergies before you start injecting
penicillin or whatever you're gonna do.
523
00:27:56,775 --> 00:27:57,585
That makes perfect sense.
524
00:27:58,155 --> 00:28:02,485
But there's so much information
between episodes and most of the
525
00:28:02,545 --> 00:28:07,845
tolerance for delay amongst most
clinicians is incredibly short.
526
00:28:08,665 --> 00:28:10,105
Three seconds and you're gone.
527
00:28:10,105 --> 00:28:10,945
I'd say a second.
528
00:28:10,945 --> 00:28:12,605
Most people don't like the application.
529
00:28:12,875 --> 00:28:14,465
You have to prefetch stuff.
530
00:28:14,855 --> 00:28:18,395
And then if you prefetch it, based
upon the scheduled exam for tomorrow
531
00:28:18,395 --> 00:28:20,855
or the scheduled exam for next
week, 'cause again, most of it's not
532
00:28:20,855 --> 00:28:24,815
emergent, most is planned a little
bit or at least sometimes a lot.
533
00:28:25,325 --> 00:28:28,385
And from there, pre-chart
some of it do something what,
534
00:28:28,835 --> 00:28:30,545
what do computers do for us?
535
00:28:30,545 --> 00:28:33,455
Computers are programmed that
do things that are monotonous,
536
00:28:33,455 --> 00:28:35,375
that advance our cause.
537
00:28:35,375 --> 00:28:37,145
Make the computer do some work.
538
00:28:37,145 --> 00:28:42,305
Have it pre-chart out medication history
so you can do reconciliation faster
539
00:28:42,305 --> 00:28:44,585
and say These are all duplicated out,
540
00:28:44,635 --> 00:28:46,615
I see four things that
look relatively current.
541
00:28:46,855 --> 00:28:48,565
Hello, Mr. Patient or Miss Patient.
542
00:28:49,375 --> 00:28:51,025
Which one of these are
really currently active?
543
00:28:51,025 --> 00:28:51,955
What'd you take today?
544
00:28:51,955 --> 00:28:53,725
And I see four things that are candidates.
545
00:28:53,995 --> 00:28:57,205
As opposed to, tell me everything
that you wanna tell me about yourself.
546
00:28:57,485 --> 00:29:01,475
Use the data to pre-stage that,
that first interaction so you
547
00:29:01,475 --> 00:29:02,495
can get to the meaty stuff.
548
00:29:02,495 --> 00:29:06,435
I think the great power of interop
and data availability in healthcare is
549
00:29:06,435 --> 00:29:10,935
gonna be allowing for providers to work
to the level of their capabilities.
550
00:29:11,745 --> 00:29:13,845
By having more time to
interact with the patient.
551
00:29:13,935 --> 00:29:16,875
The data and the computer should
help get out of the way, not
552
00:29:16,875 --> 00:29:18,315
in the way if done correctly.
553
00:29:19,584 --> 00:29:22,644
Paul, thanks so much for a thoughtful
and insightful conversation.
554
00:29:22,674 --> 00:29:26,184
Before we close it out today,
what's one last piece of advice that
555
00:29:26,184 --> 00:29:30,114
you'd offer healthtech leaders and
innovators to navigate the sometimes
556
00:29:30,114 --> 00:29:33,439
confusing world of interop and to
drive great outcomes for patients?
557
00:29:34,730 --> 00:29:37,540
Well, first of all the main thing
I have to say, especially if you're
558
00:29:37,540 --> 00:29:41,800
on the startup side, is be prepared
for a longer haul than you'd like.
559
00:29:42,150 --> 00:29:43,465
We think of healthcare going fast.
560
00:29:44,345 --> 00:29:48,845
Healthcare technology does not
change as fast as you think it does.
561
00:29:48,845 --> 00:29:53,285
There are some brilliant engineers I've
met who worked at amazing institutions
562
00:29:53,285 --> 00:29:56,855
that paid them a boatload of money
because they were PhD, computer science,
563
00:29:56,905 --> 00:30:00,445
super geeks that saw healthcare as a
place that needed help and said, I'm
564
00:30:00,445 --> 00:30:01,915
gonna go over there and help this thing.
565
00:30:02,455 --> 00:30:05,965
And what you find out when you
get here is there is bureaucracy
566
00:30:05,965 --> 00:30:07,075
partially for a reason.
567
00:30:08,125 --> 00:30:11,485
It's a science, not just a
technology and science, it's
568
00:30:11,485 --> 00:30:13,105
actually an artful science, right?
569
00:30:13,105 --> 00:30:16,195
It's not, you're not gonna get the same
result from two different providers.
570
00:30:16,195 --> 00:30:20,215
So there's some artistic license involved
here and there's a lot of regulation.
571
00:30:20,585 --> 00:30:24,910
Be prepared with your VC and whatever
money you've got, it's gonna take
572
00:30:24,910 --> 00:30:26,395
a couple years to get traction.
573
00:30:26,445 --> 00:30:28,465
And I'm not saying that doesn't
mean you shouldn't start.
574
00:30:28,495 --> 00:30:29,695
It just means be ready.
575
00:30:30,125 --> 00:30:36,955
I often talk about leadership and how you
should act with your staff and employees
576
00:30:36,955 --> 00:30:38,305
and how you interact with others.
577
00:30:38,615 --> 00:30:41,565
Be fair, but also
understand what's going on.
578
00:30:42,105 --> 00:30:46,755
And healthcare is a it's a
hard industry to crack into.
579
00:30:46,785 --> 00:30:49,785
You need to get that those
first customers, but be patient.
580
00:30:49,785 --> 00:30:53,145
You have to have a good sense for how
long to stay the course before you pivot.
581
00:30:53,565 --> 00:30:56,235
Over pivoting in healthcare
sometimes makes you irrelevant
582
00:30:56,235 --> 00:30:57,195
'cause no one knows what you do.
583
00:30:58,618 --> 00:31:01,798
Paul, thanks again for joining us
and for sharing your perspective.
584
00:31:02,068 --> 00:31:05,818
And to our listeners, I hope you've
gained valuable insights into the ongoing
585
00:31:05,818 --> 00:31:09,718
challenges and promising solutions in
healthcare interoperability, especially
586
00:31:09,958 --> 00:31:13,688
unique opportunities around patient
access and the future of data exchange.
587
00:31:17,860 --> 00:31:20,200
Thanks again for joining
us on Hard Problems, Smart
588
00:31:20,200 --> 00:31:22,150
Solutions, the Newfire Podcast.
589
00:31:22,390 --> 00:31:23,200
Until next time.
00:00:00,180 --> 00:00:04,110
And what I've seen decade over decade,
'cause I can actually say that now
2
00:00:04,110 --> 00:00:07,290
I've been doing this for a long time,
is those who wait, I wouldn't say they
3
00:00:07,290 --> 00:00:09,510
get hurt, but they lose advantage.
4
00:00:09,570 --> 00:00:12,540
And negative advantage is the
same thing as getting hurt
5
00:00:12,540 --> 00:00:13,680
when the market moves fast.
6
00:00:13,740 --> 00:00:17,940
So I would say get connected and then
start finding tools for the specific
7
00:00:17,940 --> 00:00:19,260
things that you're looking for.
8
00:00:19,350 --> 00:00:22,440
It's not like a sea-change
moment when you into interop.
9
00:00:22,530 --> 00:00:26,040
It's an incremental movement that
leads other increments, movements
10
00:00:26,040 --> 00:00:27,390
that leads to general success.
11
00:00:35,840 --> 00:00:39,530
Welcome to Hard Problems, Smart
Solutions, the Newfire Podcast where
12
00:00:39,530 --> 00:00:43,040
we explore the toughest challenges and
the smartest solutions with leaders
13
00:00:43,040 --> 00:00:44,810
across technology and healthcare.
14
00:00:45,500 --> 00:00:48,590
I'm Brendan Iglehart, Staff Healthcare
Architects here at Newfire Global
15
00:00:48,590 --> 00:00:50,780
Partners, and your host for this episode.
16
00:00:53,476 --> 00:00:55,966
For today's episode, we have the
pleasure of speaking with Paul
17
00:00:55,966 --> 00:00:58,996
Wilder, the executive director of
the CommonWell Health Alliance.
18
00:00:59,446 --> 00:01:02,536
Paul brings more than two decades of
experience in health IT leadership,
19
00:01:02,956 --> 00:01:05,326
deeply committed to transforming
healthcare delivery through
20
00:01:05,326 --> 00:01:06,581
technology and data exchange.
21
00:01:07,550 --> 00:01:10,340
He's at the forefront of driving
nationwide interoperability,
22
00:01:10,700 --> 00:01:13,490
working to empower clinicians,
practitioners, and individuals
23
00:01:13,490 --> 00:01:15,740
with robust data exchange services.
24
00:01:15,957 --> 00:01:19,677
Paul's work at CommonWell aligns
perfectly with our podcast mission to
25
00:01:19,677 --> 00:01:22,677
discuss the toughest challenges and
the smartest solutions with leaders
26
00:01:22,677 --> 00:01:24,207
across technology and healthcare.
27
00:01:24,434 --> 00:01:27,194
He is known for his insightful
perspectives on the complex world of
28
00:01:27,194 --> 00:01:30,844
health data from regulatory shift to the
practical realities of implementation.
29
00:01:31,429 --> 00:01:32,809
Paul, thanks for joining us today.
30
00:01:33,529 --> 00:01:33,739
Great.
31
00:01:33,739 --> 00:01:34,339
Thanks for being here.
32
00:01:34,369 --> 00:01:35,299
Thanks for letting me be here.
33
00:01:35,419 --> 00:01:35,854
It should be fun.
34
00:01:36,902 --> 00:01:41,012
So to get started, Paul, I'm, I want to
begin with your journey in health IT.
35
00:01:41,462 --> 00:01:45,642
So, you've had a significant tenure
back at, before with the New York
36
00:01:45,642 --> 00:01:48,582
eHealth Collaborative before moving
into your current role as the
37
00:01:48,582 --> 00:01:50,282
executive director at CommonWell.
38
00:01:50,532 --> 00:01:54,292
So what drew you into this crazy
world of healthcare interoperability?
39
00:01:55,354 --> 00:01:55,594
Well
40
00:01:55,594 --> 00:01:56,314
First I was drawn to
41
00:01:56,314 --> 00:01:59,504
healthcare starting technology
that was quite outside.
42
00:01:59,504 --> 00:02:03,784
I was doing network support cellular
networks kind of stuff, and immediately
43
00:02:03,784 --> 00:02:07,434
got pulled over and about six months
into that first out-of-college
44
00:02:07,454 --> 00:02:10,254
job into records management.
45
00:02:10,404 --> 00:02:15,384
So, the, this is precursors to EHRs
going back in the late nineties and
46
00:02:15,384 --> 00:02:20,974
there was the document flow and the
fun thing about that one, I was young
47
00:02:20,974 --> 00:02:21,894
and got to travel on the country.
48
00:02:21,894 --> 00:02:22,454
That was fun.
49
00:02:22,454 --> 00:02:27,044
But, I really got a flavor for the
business backend side of healthcare.
50
00:02:27,104 --> 00:02:30,644
I originally thought I was going to Med
School, actually got accepted to one,
51
00:02:30,644 --> 00:02:33,824
decided not to do it, but always had
an interest in healthcare in general.
52
00:02:34,274 --> 00:02:37,064
And I realized there's multiple
areas that people can help with.
53
00:02:37,064 --> 00:02:40,604
And the really fun part that I thought
about then, but really thinking about more
54
00:02:40,604 --> 00:02:44,354
about now, is how much it affects your
everyday life and how you're influencing
55
00:02:44,354 --> 00:02:46,791
something that influences a lot of
people around that you love, admire.
56
00:02:46,791 --> 00:02:49,911
And some people that you just want to be,
you know, better humans going forward.
57
00:02:50,581 --> 00:02:53,431
For listeners who may not be familiar,
can you explain a little bit about
58
00:02:53,431 --> 00:02:57,151
what CommonWell is and the role
that you all play in this industry?
59
00:02:58,197 --> 00:02:58,437
Yeah,
60
00:02:58,437 --> 00:02:58,647
Given
61
00:02:58,647 --> 00:03:01,713
the best generic part of healthcare
is fun to play with 'cause you get to
62
00:03:01,713 --> 00:03:03,393
influence stuff and make things better.
63
00:03:03,443 --> 00:03:06,043
Interoperability i s
the focus of CommonWell.
64
00:03:06,043 --> 00:03:11,533
CommonWell was founded 2013, so a little
over a decade ago, by some preeminent
65
00:03:11,603 --> 00:03:16,523
electronic health record and other health
information technology companies with
66
00:03:16,523 --> 00:03:24,113
the belief that data availability and
exchange across providers was necessary
67
00:03:24,113 --> 00:03:26,963
for a better person-centric healthcare.
68
00:03:27,948 --> 00:03:32,688
And the original philosophy was
that it can't be a, or shouldn't be,
69
00:03:32,748 --> 00:03:34,488
shouldn't be a single vendor solution.
70
00:03:35,038 --> 00:03:38,878
'Cause patients see many different
providers, different types of specialties,
71
00:03:39,238 --> 00:03:42,928
and they're not picking by EHR,
they're picking by many other factors.
72
00:03:43,348 --> 00:03:48,318
So we were built with as a 501(c)(6),
a trade association nonprofit type.
73
00:03:49,008 --> 00:03:52,338
And our goal was to make sure all
the data was available every time
74
00:03:52,338 --> 00:03:56,623
a provider needs it across every
healthcare entity across the country.
75
00:03:57,103 --> 00:04:01,023
And we continue to strive to do that
with both within ourselves and across
76
00:04:01,023 --> 00:04:05,403
partners to move data from provider to
provider and provider to patient and
77
00:04:05,733 --> 00:04:08,253
keep adding Ps, payers, public health.
78
00:04:08,303 --> 00:04:09,233
Keep going down the list.
79
00:04:10,265 --> 00:04:13,145
One of the things that was interesting
when we were chatting last week in
80
00:04:13,145 --> 00:04:16,655
preparation for this podcast is that
you were talking about how CommonWell
81
00:04:16,985 --> 00:04:20,735
has been on the forefront of enabling
patient access to their own data for
82
00:04:20,735 --> 00:04:25,110
quite some time, which is, and currently
a big topic within interoperability.
83
00:04:25,110 --> 00:04:28,050
So can you explain a little bit about
what you've been doing there and why
84
00:04:28,050 --> 00:04:31,285
it's taken so long to get more visibility
and focus on that sort of work?
85
00:04:32,031 --> 00:04:32,241
Yeah,
86
00:04:32,241 --> 00:04:33,381
we started 2013.
87
00:04:33,381 --> 00:04:36,531
It really was a treatment centric
network provider to provider exchange.
88
00:04:36,581 --> 00:04:40,781
It was 2016 when we first added
what we called, patient access.
89
00:04:41,471 --> 00:04:44,951
And sometimes in the standards
called request, patient request.
90
00:04:45,581 --> 00:04:49,181
And I'll be honest, the first couple
years didn't go amazingly well.
91
00:04:49,211 --> 00:04:50,901
We had two general headwinds.
92
00:04:50,941 --> 00:04:54,131
One, there wasn't that much the
tracks laid for provider-provider.
93
00:04:54,151 --> 00:05:00,491
We thought we were doing well in 2016 with
a high number of thousands of connections.
94
00:05:00,921 --> 00:05:05,421
By the time I got here, and we go about
two years in, so we're around 2021,
95
00:05:05,421 --> 00:05:07,367
we were at tens of thousands, right?
96
00:05:07,367 --> 00:05:09,827
So we added in almost a factor
from that point forward.
97
00:05:10,337 --> 00:05:14,327
And the reality is most patient access
applications, or individual access as
98
00:05:14,327 --> 00:05:18,347
we like to call it now, or individual
access services, they don't work well
99
00:05:18,347 --> 00:05:19,637
if you can't get to all the data.
100
00:05:20,517 --> 00:05:23,867
Because the patient is looking
for their most recent thing.
101
00:05:23,867 --> 00:05:26,447
It's a very atomic level entity.
102
00:05:26,447 --> 00:05:31,217
It's provider Bob, provider John,
provider Lucy, that has my data or
103
00:05:31,277 --> 00:05:35,567
hospital X, Y, or Z. And when that
one's missing, they kind of disengage.
104
00:05:35,627 --> 00:05:38,237
So you needed more of a network
effect to make this work.
105
00:05:38,717 --> 00:05:44,627
The second part was the network itself
was designed for provider to provider,
106
00:05:45,017 --> 00:05:48,047
and we didn't actually write our
agreements to give this data to patients.
107
00:05:48,387 --> 00:05:51,507
And those of you who are HIPAA experts
and know things about healthcare
108
00:05:51,507 --> 00:05:53,787
data, you don't move data to things.
109
00:05:53,787 --> 00:05:57,027
You're not allowed to move them to
without a contract staying as such.
110
00:05:57,027 --> 00:05:59,487
So we had to do a fair amount of
retrofitting to get things up to
111
00:05:59,487 --> 00:06:02,067
snuff to be able to do patient access.
112
00:06:02,487 --> 00:06:07,587
Fortunately, the rest of the market
or at least the industry or us as just
113
00:06:07,587 --> 00:06:12,057
citizens and active healthcare users said,
Hey, we really want access to our data.
114
00:06:12,567 --> 00:06:15,657
And more and more people heard that
and it started to appear in other
115
00:06:15,657 --> 00:06:19,437
networking frameworks, which increased,
it increased the ability for us to
116
00:06:19,437 --> 00:06:20,997
take the time to focus and fix it.
117
00:06:21,517 --> 00:06:25,177
So we recently did that with one
of our largest members being Athena
118
00:06:25,537 --> 00:06:31,427
just moved in their entire cohort of
providers into our patient access aware
119
00:06:31,427 --> 00:06:33,497
application that is also TEFCA aware.
120
00:06:33,902 --> 00:06:37,182
And now we're off to the races with
people, I wouldn't say playing catch-up,
121
00:06:37,202 --> 00:06:39,502
I think we're all clotting to the same
point at the same time, which is great.
122
00:06:41,164 --> 00:06:43,894
So, Paul, I know a lot of us in
the industry have been tracking the
123
00:06:43,924 --> 00:06:47,524
recent administration's movements
around health IT, and specifically
124
00:06:47,524 --> 00:06:53,224
watching the recent RFI and events
that have been held by CMS and ONCASTP.
125
00:06:53,464 --> 00:06:56,104
I think that you've been involved with
some of those events in Washington.
126
00:06:56,104 --> 00:06:58,794
Do you wanna tell us a little bit
about the kind of tone and conversation
127
00:06:58,844 --> 00:07:02,349
going on in there and what you expect
to come outta those of those meetings?
128
00:07:03,114 --> 00:07:05,594
Yeah, there are really three
major events that occurred.
129
00:07:05,654 --> 00:07:12,524
One goes back to May where CMS
released an RFI, looking for commentary
130
00:07:12,524 --> 00:07:17,354
on the benefits of health IT and
what are good apps out there, as
131
00:07:17,354 --> 00:07:20,484
well as other things related to
technology and data interoperability.
132
00:07:21,004 --> 00:07:24,874
It culminated in a, well, actually the
culminated was the end of the RFI the
133
00:07:24,874 --> 00:07:27,544
day you had to submit everything, but in
the middle they did a listening session
134
00:07:28,199 --> 00:07:32,399
where they had the, the higher-ups of
HHS, so this was Secretary Kennedy,
135
00:07:32,399 --> 00:07:36,329
then Dr. Oz from CMS side, down the
line explaining what's going on.
136
00:07:36,329 --> 00:07:37,499
Amy Gleason, et cetera.
137
00:07:38,084 --> 00:07:41,984
And the morning session was presentation,
followed by a bunch of panels in the
138
00:07:41,984 --> 00:07:44,504
afternoon was more working sessions
to figure out what to do next.
139
00:07:45,104 --> 00:07:48,974
The morning panels were full of
people just banging the floor
140
00:07:48,974 --> 00:07:50,594
looking for better patient access.
141
00:07:50,644 --> 00:07:54,304
And you have to think about this for
a second, realize that the people
142
00:07:54,334 --> 00:07:58,204
putting together this listening
session had to have set a tone.
143
00:07:58,204 --> 00:07:59,584
They chose who to put up there.
144
00:07:59,644 --> 00:07:59,944
Right.
145
00:07:59,944 --> 00:08:02,464
And there were people they knew
that were gonna say these things.
146
00:08:02,834 --> 00:08:04,694
You don't have to, you don't
have to wind people up and
147
00:08:04,694 --> 00:08:05,624
tell 'em exactly what to say.
148
00:08:05,624 --> 00:08:06,614
Just get the right people.
149
00:08:06,914 --> 00:08:09,824
So the fact that CMS and HHS put a
lot of people they knew they were
150
00:08:09,824 --> 00:08:14,834
gonna be focused on individual
access on a stage during their big
151
00:08:14,834 --> 00:08:17,504
show is interesting unto itself.
152
00:08:17,504 --> 00:08:19,399
That says the administration
is very interested.
153
00:08:19,979 --> 00:08:24,389
And then the conversation that happened
after that, which then led to this
154
00:08:24,389 --> 00:08:28,779
idea of a CMS aligned network and a
further event that happened a couple
155
00:08:28,779 --> 00:08:33,399
weeks ago at the White House where
the CEOs and other officers of many of
156
00:08:33,399 --> 00:08:36,129
the companies are at that event, and
actually a smaller group than that,
157
00:08:36,639 --> 00:08:41,664
were pledging to be aligned to CMS as
we go forward with a significant focus
158
00:08:41,664 --> 00:08:46,344
of that being, how do we help patients
make better decisions with better data?
159
00:08:46,864 --> 00:08:51,664
And I think in the first administration
round of Trump, you saw financial data.
160
00:08:51,664 --> 00:08:55,404
There was an attempt to do getting
price information, price transparency
161
00:08:55,404 --> 00:08:58,284
to patients, which I don't think went
as well as everybody wanted it to,
162
00:08:58,284 --> 00:09:01,044
but it was an interesting idea of
more information, better decisions.
163
00:09:01,404 --> 00:09:04,614
And I think this is the next evolution
to how about the clinical data, too?
164
00:09:04,614 --> 00:09:07,254
Maybe the two together
start to move the needle?
165
00:09:07,294 --> 00:09:10,669
Understanding that the first
tools probably won't do as much
166
00:09:10,669 --> 00:09:13,399
as you wanted to, but you have to
start every journey with a step.
167
00:09:13,819 --> 00:09:16,939
And a significant step would be
getting the ability for apps to
168
00:09:16,939 --> 00:09:19,519
get data in patients' hands so they
can start doing something with it.
169
00:09:20,783 --> 00:09:24,983
You mentioned Amy Gleason and just then,
um, who was the acting head of the newly
170
00:09:24,983 --> 00:09:28,563
formed department of government efficiency
or DOGE, under the federal government.
171
00:09:28,573 --> 00:09:31,678
And obviously she comes from a
background in, in healthtech.
172
00:09:31,868 --> 00:09:35,148
So I'm curious for your perspective
on someone like her, who's known
173
00:09:35,148 --> 00:09:36,648
for really getting things done,
174
00:09:36,868 --> 00:09:38,848
how do you think her tenure
and involvement in this
175
00:09:38,848 --> 00:09:40,118
will impact the outcome?
176
00:09:41,328 --> 00:09:42,038
Well, there's two things.
177
00:09:42,038 --> 00:09:45,328
She has she has a good patient story
in her daughter's story, which is
178
00:09:45,328 --> 00:09:46,838
for those who have not seen it.
179
00:09:47,488 --> 00:09:50,848
I won't give you a full summary,
but you can find it on the interweb
180
00:09:50,878 --> 00:09:53,008
looking up Amy Gleason story.
181
00:09:53,008 --> 00:09:56,278
Amy Gleason, CMS, YouTube,
whatever you wanna find it on.
182
00:09:56,548 --> 00:10:00,448
And she does a good job of delineating
a, you could think of a generic patient
183
00:10:00,448 --> 00:10:06,568
with a rare disease and without data
and AI tools and the like to be able to
184
00:10:06,568 --> 00:10:09,748
process that data, how things were missed.
185
00:10:09,878 --> 00:10:14,738
And I think there's a lesson in there
that, providers, i n the past you
186
00:10:14,738 --> 00:10:18,508
would, you have seen some resistance
to some adoption of technology.
187
00:10:18,508 --> 00:10:21,948
For example, we saw an initial bad
reaction to like direct messaging,
188
00:10:21,998 --> 00:10:25,238
provider to provider direct, but
then they tried applying patient
189
00:10:25,238 --> 00:10:29,768
to provider and we skipped that and
went instead to portal conversations.
190
00:10:29,768 --> 00:10:33,518
And when you start to learn as people
are generally pretty nice, people don't
191
00:10:33,518 --> 00:10:36,218
abuse your, you know, my, my mobile
phone number's in my signature line for
192
00:10:36,218 --> 00:10:38,498
my email, and no one calls it, right.
193
00:10:38,498 --> 00:10:39,878
, No one bothers people that way.
194
00:10:39,878 --> 00:10:44,038
They have other ways of communicating
and I think they realized after that push
195
00:10:44,038 --> 00:10:47,938
of communication direct to patient, that
they're not gonna get abused as much.
196
00:10:47,938 --> 00:10:50,998
Maybe it's time to think about using
technology in a more augmented way.
197
00:10:51,538 --> 00:10:54,418
But Amy in particular, is
like, no, go do it faster.
198
00:10:54,518 --> 00:10:59,153
You're hurting healthcare, you're
hurting yourself actually by not allowing
199
00:10:59,153 --> 00:11:01,013
patients to supplement what you're doing.
200
00:11:01,453 --> 00:11:09,583
And her brain in terms of doing things
faster but efficiently and effectively,
201
00:11:10,003 --> 00:11:13,723
on top of the passion to do it from her
own personal story, is usually the kind
202
00:11:13,723 --> 00:11:15,373
of secret sauce you need to move things.
203
00:11:15,823 --> 00:11:20,788
And you put someone in a central position
where they can convene people and blast
204
00:11:20,788 --> 00:11:24,418
through what I, I usually call the first
mover disadvantage of network effects, the
205
00:11:24,418 --> 00:11:28,188
first guy does all the work and a lot of
people follow in after, i s significant.
206
00:11:28,188 --> 00:11:33,428
Like that kind of the beat of the drum
being consistent across many parties who
207
00:11:33,428 --> 00:11:37,868
normally compete and don't wanna be synced
with each other is very powerful when
208
00:11:37,873 --> 00:11:39,278
you're trying to develop network of x.
209
00:11:40,467 --> 00:11:44,187
Speaking of building out networks, one
of the big developments that I've really
210
00:11:44,187 --> 00:11:47,127
tracked from, from my side and the
customers that we worked with in the last
211
00:11:47,127 --> 00:11:52,317
couple years is the rollout of the On
Behalf Of designation with Carequality,
212
00:11:52,337 --> 00:11:56,157
which I understand that you had a hand
in originally developing that concept.
213
00:11:56,157 --> 00:11:59,377
Can you tell us a little bit about, on
behalf of our OBO and a little bit of
214
00:11:59,377 --> 00:12:00,912
the story about how that came about?
215
00:12:02,877 --> 00:12:03,507
Yeah.
216
00:12:03,557 --> 00:12:05,447
I wanna see healthcare move faster too.
217
00:12:05,507 --> 00:12:09,537
And you can imagine running a trade
alliance, which is supposed to be
218
00:12:09,957 --> 00:12:14,727
supportive of trade and not just the
entries, entities are in it right now.
219
00:12:14,777 --> 00:12:17,597
It's supposed to be also supportive
of the next thing to enter.
220
00:12:18,137 --> 00:12:20,177
I kept thinking of what
are we missing, right?
221
00:12:20,177 --> 00:12:25,052
If I just, if all we did was work with our
primary EHRs and the like, t hat's great.
222
00:12:25,082 --> 00:12:27,902
They'll fund the enterprise and
we're a true nonprofit, we make
223
00:12:27,902 --> 00:12:28,862
too much money, we give it back.
224
00:12:28,862 --> 00:12:30,272
I was like, who else needs help?
225
00:12:30,872 --> 00:12:35,882
And I kept looking to the startups
and the neat little things being
226
00:12:35,882 --> 00:12:39,572
developed at the edge that are
the first step to their existence.
227
00:12:39,572 --> 00:12:42,712
And I had done some accelerator
type things in my past as well.
228
00:12:43,072 --> 00:12:44,962
And how did you do an
accelerator in healthcare?
229
00:12:44,962 --> 00:12:47,392
Usually it was by getting
your first customer, right?
230
00:12:47,392 --> 00:12:50,932
You would link them up with a technology
set and say, Hey, I'm bringing this
231
00:12:50,932 --> 00:12:55,822
preeminent academic in as a pilot customer
to help you develop and go forward and
232
00:12:55,822 --> 00:12:57,502
then get more contacts and go from there.
233
00:12:57,982 --> 00:13:00,592
But when you start looking
at patient stuff and littler
234
00:13:00,712 --> 00:13:03,622
tools, it doesn't fit as well.
235
00:13:03,682 --> 00:13:07,952
Like we're talking like, one
person, one diabetic, we need to
236
00:13:07,952 --> 00:13:09,342
move faster and get things going.
237
00:13:09,342 --> 00:13:13,572
So I said, why i f providers are hooking
up their EHRs at scale, and we're getting
238
00:13:13,572 --> 00:13:17,532
the data to the network, we're solving
the data availability problem, right?
239
00:13:17,532 --> 00:13:19,002
The data is going to be there.
240
00:13:19,397 --> 00:13:23,547
What I was missing was, who else needs
to use this data, the supply and demand.
241
00:13:24,147 --> 00:13:27,657
I said, can we hook up these
applications to the network?
242
00:13:28,137 --> 00:13:31,497
To get the same data to accelerate their
ability to do what they need to do.
243
00:13:31,497 --> 00:13:35,607
Iinstead of integrating to one
hospital or to one ambulatory practice,
244
00:13:35,637 --> 00:13:39,057
why don't they hook up to all of them
and do that without having to hook up
245
00:13:39,057 --> 00:13:42,177
through the high primary EHR 'cause
it's expensive and time-consuming?
246
00:13:42,687 --> 00:13:45,147
And I was more worried about the
time-consuming, to be honest.
247
00:13:45,147 --> 00:13:48,157
I mean, it's fine for, again, we're
a trade association we support trade.
248
00:13:48,517 --> 00:13:51,127
I'm not against people charging
for services that's fine.
249
00:13:51,557 --> 00:13:52,907
You just don't want it to be a barrier.
250
00:13:53,612 --> 00:13:56,042
And so I felt like we weren't getting the
right tools in front of the right people.
251
00:13:56,282 --> 00:13:57,242
So we built this OBO.
252
00:13:57,242 --> 00:14:01,487
If your EHR, if your EHR is already
connected, why can't you connect
253
00:14:01,487 --> 00:14:04,217
every other product you bought
that is for treatment purposes?
254
00:14:04,337 --> 00:14:04,847
All of them.
255
00:14:05,327 --> 00:14:08,597
In the dermatology app that takes a neat
picture from your iPhone and sends it up
256
00:14:08,597 --> 00:14:12,017
to some AI tool to do a skin lesion check.
257
00:14:12,317 --> 00:14:15,257
Why shouldn't it have access
to your history of skin cancer,
258
00:14:15,257 --> 00:14:16,427
family history, and the like?
259
00:14:16,667 --> 00:14:20,417
To better inform the answer of what it's
about, to help suggest in that clinical
260
00:14:20,417 --> 00:14:21,887
decision support thing it's doing.
261
00:14:22,277 --> 00:14:23,297
And do it cheaply.
262
00:14:23,387 --> 00:14:26,627
And by cheaply we make the
network an interface engine.
263
00:14:26,627 --> 00:14:29,957
Everybody can access the same
information at the atomic level of
264
00:14:29,957 --> 00:14:31,757
you, Brendan, Paul, whatever it is.
265
00:14:32,307 --> 00:14:37,667
There was some resistance at first
and unfortunately it got changed a
266
00:14:37,667 --> 00:14:39,857
little bit in the way it got word
that people found loopholes to
267
00:14:39,857 --> 00:14:41,927
use it for, semi-abusive things.
268
00:14:41,927 --> 00:14:44,807
And we were still addressing
that trust issue today.
269
00:14:45,767 --> 00:14:49,817
But I still stand behind the idea
that we need more access, not less.
270
00:14:50,187 --> 00:14:53,747
And we just need to do it safely and
securely so that, we maintain trust and
271
00:14:53,747 --> 00:14:55,307
provide the right solutions to patients.
272
00:14:56,699 --> 00:15:00,149
Paul, when we think about the future
of patient access to their own data,
273
00:15:00,149 --> 00:15:03,419
obviously there's a lot of different
applications for that, including
274
00:15:03,479 --> 00:15:07,019
data correction, which was one that I
have experienced as a patient myself.
275
00:15:07,019 --> 00:15:11,309
You know, seeing things in my own patient
record that is, is incorrect, and then
276
00:15:11,309 --> 00:15:13,679
trying to take steps to address that.
277
00:15:13,889 --> 00:15:18,119
Why is that specifically such an
important aspect of kind of unlocking
278
00:15:18,119 --> 00:15:22,379
data for patients that ultimately they,
they should have the right to use?
279
00:15:24,194 --> 00:15:24,404
Yeah.
280
00:15:24,404 --> 00:15:28,784
At the risk of giving the standard answer
at every technology conference, and
281
00:15:28,784 --> 00:15:32,804
especially healthcare conferences, AI, so
we got this artificial intelligence thing
282
00:15:32,804 --> 00:15:39,014
going on, and I think it's important to
remember how AI actually works, right?
283
00:15:39,044 --> 00:15:40,064
It's statistic space, right?
284
00:15:40,064 --> 00:15:43,914
It's taking... when you do an LLM
and have it write an RFP for you
285
00:15:43,994 --> 00:15:45,704
in Copilot, or whatever you use,
286
00:15:46,184 --> 00:15:50,484
it's just aware of what kind of words
go together when it's, look through
287
00:15:50,484 --> 00:15:53,514
the history of time of everybody
writing things and gives you a
288
00:15:53,514 --> 00:15:56,724
suggestion that seems to be within the
confines of what you're looking for.
289
00:15:57,264 --> 00:15:59,814
And when you get to healthcare
data, it's doing the same thing.
290
00:15:59,844 --> 00:16:04,564
Going back to the days of Watson and
other early precursors where we are today.
291
00:16:05,344 --> 00:16:08,524
The one thing you don't
wanna do is confuse it.
292
00:16:08,864 --> 00:16:12,164
We, you'll hear people talking
about AI hallucinations and it
293
00:16:12,164 --> 00:16:13,664
starts to fill in data gaps.
294
00:16:14,174 --> 00:16:18,494
Another problem is when it has a
piece of data that doesn't belong
295
00:16:18,524 --> 00:16:21,464
and it doesn't always know what
should go around it or ignore it.
296
00:16:21,974 --> 00:16:24,884
So if you have errors,
bad things can happen.
297
00:16:24,884 --> 00:16:26,054
Lemme give you an example.
298
00:16:26,104 --> 00:16:32,284
I had a shot, a flu shot last
season and my record said I
299
00:16:32,284 --> 00:16:33,994
had gotten HPV vaccination.
300
00:16:34,714 --> 00:16:39,544
Now, it is actually more common these
days for or for especially younger males
301
00:16:39,544 --> 00:16:43,804
to get HPV vaccinations, but not very
common for a 48-year-old male to do that.
302
00:16:44,464 --> 00:16:50,554
And my doctor would probably look over
that, if he was looking at fantastic
303
00:16:50,554 --> 00:16:53,484
cancer drugs that would be the
all-end-all to cure prostate cancer
304
00:16:53,544 --> 00:16:58,014
it's available and says, but what if
that prostate cancer drug material
305
00:16:58,014 --> 00:17:01,254
said it's contraindicated to people
who have the HPV vaccination?
306
00:17:01,864 --> 00:17:05,104
My provider might skip it, but the
AI provider may not know what to do.
307
00:17:05,614 --> 00:17:07,294
And so it's statistical analysis.
308
00:17:07,294 --> 00:17:09,504
We're gonna start suggesting
different therapies.
309
00:17:10,224 --> 00:17:12,654
Could underweight the perfect drug, right?
310
00:17:12,654 --> 00:17:16,074
Everything else could fit in, but
this thing is a bang-out criteria and
311
00:17:16,074 --> 00:17:18,774
bang-out means eliminate it at all costs.
312
00:17:19,224 --> 00:17:20,064
You don't want that.
313
00:17:20,134 --> 00:17:23,314
And there are various things that,
little tiny nuances, little nudges you
314
00:17:23,314 --> 00:17:25,324
can do that make the model messed up.
315
00:17:25,864 --> 00:17:27,474
Providers can do the same thing.
316
00:17:27,684 --> 00:17:29,854
They already have clinical
decision support tools on their
317
00:17:29,854 --> 00:17:34,114
side, and contraindications X is
contraindicated to Y. You can't
318
00:17:34,114 --> 00:17:35,404
have these two things together.
319
00:17:35,809 --> 00:17:37,249
Automatically set off red flags.
320
00:17:37,999 --> 00:17:38,239
Right?
321
00:17:38,239 --> 00:17:40,279
So that would be, that'd be an
extreme example, but it could
322
00:17:40,279 --> 00:17:43,009
also say there is no data on this.
323
00:17:43,069 --> 00:17:46,639
Anything that decreases that
Superdrug now being not a Superdrug,
324
00:17:46,644 --> 00:17:49,894
because of my potential biochemistry
and the things I have in it.
325
00:17:50,284 --> 00:17:52,594
So I think it's really important
that we have the right data.
326
00:17:52,594 --> 00:17:55,714
I mean, today, most of us probably
worry about the wrong data being
327
00:17:55,714 --> 00:17:57,634
things we get billed for, right?
328
00:17:57,634 --> 00:17:58,654
I didn't have this done.
329
00:17:58,654 --> 00:18:00,034
I don't want to pay my $20 co-pay.
330
00:18:00,034 --> 00:18:02,044
This was supposed to be an annual
exam, why do I have a co-pay?
331
00:18:02,044 --> 00:18:02,794
It's supposed to be free.
332
00:18:03,244 --> 00:18:04,624
And those guy, oh, it's coded wrong.
333
00:18:04,624 --> 00:18:05,554
You like, okay, you fix it.
334
00:18:05,554 --> 00:18:08,824
Fix it to an annual wellness exam
and not just a normal office visit.
335
00:18:09,184 --> 00:18:11,554
And that's fine for the $20,
a hundred dollars, a couple
336
00:18:11,554 --> 00:18:12,994
thousand dollars even, you pay.
337
00:18:13,444 --> 00:18:17,254
But the clinical stories could get
really messed up if we confuse the AIs.
338
00:18:17,644 --> 00:18:20,359
And it is, you could argue
whether it's the future or not,
339
00:18:20,359 --> 00:18:21,439
but it's definitely a factor.
340
00:18:21,839 --> 00:18:25,799
So getting data into the hands of
patients is the most likely way
341
00:18:25,799 --> 00:18:27,119
to make sure the data is accurate.
342
00:18:27,449 --> 00:18:29,939
'cause I probably care more my
data than anybody else does.
343
00:18:32,204 --> 00:18:36,674
And on the topic of AI, Paul, as we
continue to see, meaningful advancement
344
00:18:36,734 --> 00:18:40,484
toward making data more liquid for
different use cases, involving including
345
00:18:40,484 --> 00:18:43,904
those that patients are directing,
how do we make sure that patients
346
00:18:43,964 --> 00:18:48,374
stay protected or that f olks' privacy
is respected in this whole process.
347
00:18:50,729 --> 00:18:53,859
You know, I saw, we talked about
the events that happened and then
348
00:18:53,859 --> 00:18:54,894
there was the big White House event.
349
00:18:55,674 --> 00:18:58,584
And I don't, it doesn't really
matter where you sit in the political
350
00:18:58,584 --> 00:19:02,604
spectrum, I did see something disturbing
happen that happened that, that day.
351
00:19:03,024 --> 00:19:07,439
There was an article put out by a, what I
consider a good media outlet, that talked
352
00:19:07,439 --> 00:19:11,099
about the federal government is trying to
make a system to collect all your data.
353
00:19:11,654 --> 00:19:15,854
Big tech is getting into your data,
be careful and got quotes from some
354
00:19:16,094 --> 00:19:18,804
preeminent professors about the
security and privacy concerns alike.
355
00:19:19,334 --> 00:19:24,769
And the irony there was the dialogue
was exactly opposite that, right?
356
00:19:24,769 --> 00:19:29,479
The dialogue was, we want big
tech to do the right thing, to get
357
00:19:29,479 --> 00:19:34,279
data into your hands so that you
can control where it goes, right?
358
00:19:34,279 --> 00:19:36,169
We're not trying to
collect it for ourselves.
359
00:19:36,469 --> 00:19:40,419
We're trying to collect it on your behalf
and then let go of it, right, put it
360
00:19:40,419 --> 00:19:45,949
into the storage thing of your choice,
into the application of your choice.
361
00:19:45,949 --> 00:19:47,449
And you do what you wanna do.
362
00:19:47,449 --> 00:19:49,669
If you wanna be in a research
study, you release it.
363
00:19:49,669 --> 00:19:49,849
Right?
364
00:19:49,849 --> 00:19:51,469
That's not, we're trying
to get in the middle.
365
00:19:51,919 --> 00:19:56,689
So I think we've banged around the
edges of security and privacy for a
366
00:19:56,689 --> 00:20:00,989
long time, but I think the current push
right now has the right philosophy.
367
00:20:01,679 --> 00:20:07,034
The danger of securing privacy
around healthcare data is, you
368
00:20:07,034 --> 00:20:11,924
gotta be really careful just not
disclosing data that might help you.
369
00:20:12,104 --> 00:20:14,684
You kind of have to trust your
provider to do the right thing.
370
00:20:15,194 --> 00:20:18,374
And we, if we go too far, making
it locked up, that you have to
371
00:20:18,374 --> 00:20:22,814
anatomically every piece of data one
element at a time, I think we are
372
00:20:22,814 --> 00:20:24,554
actually gonna negatively affect care.
373
00:20:25,089 --> 00:20:30,279
Because patients have not been given their
data for so long, they don't quite know
374
00:20:30,279 --> 00:20:31,869
what's important and what's not important.
375
00:20:32,529 --> 00:20:36,579
And so, if you start scaring people that
your data's flying around everywhere, and
376
00:20:36,579 --> 00:20:40,239
then say you have control, the a knee-jerk
reaction could be to not give it to
377
00:20:40,239 --> 00:20:45,009
anybody and you could be doing yourself
serious harm, which actually is a lesson.
378
00:20:45,069 --> 00:20:48,279
The lesson here is get the
data to the patients faster.
379
00:20:48,609 --> 00:20:50,589
We have to get past that episode.
380
00:20:50,589 --> 00:20:54,039
We have to get past that, that
fear bubble as quickly as possible
381
00:20:54,099 --> 00:20:55,389
to get back to the right balance.
382
00:20:55,389 --> 00:20:58,179
And we see in everything,
there's often an overshoot.
383
00:20:58,179 --> 00:21:01,899
I'm, I would not be surprised, if
we give patients all their data and
384
00:21:01,899 --> 00:21:05,469
we give them audit capabilities to
know where it's flying that many
385
00:21:05,469 --> 00:21:07,029
people will shut it down first.
386
00:21:07,029 --> 00:21:08,659
'cause they don't understand
how to control it.
387
00:21:08,779 --> 00:21:10,459
And then we had to bring it back up.
388
00:21:10,579 --> 00:21:11,659
But that's gonna happen.
389
00:21:11,659 --> 00:21:12,649
So you might as well do it now.
390
00:21:12,679 --> 00:21:13,699
'cause doing it later is worse.
391
00:21:15,374 --> 00:21:19,994
And also on the topic of data liquidity as
I think we talked about, there's a number
392
00:21:19,994 --> 00:21:23,144
of different factors in that beyond just
the technology factors that have made
393
00:21:23,144 --> 00:21:27,744
that challenging in healthcare to date,
but w ith the advancement and adoption
394
00:21:27,804 --> 00:21:31,944
and rollout of FHIR by different vendors,
we're seeing some kind of other technology
395
00:21:31,944 --> 00:21:33,964
pieces of that being addressed as well.
396
00:21:33,964 --> 00:21:36,204
So what's your kind of vantage
point on the current state of
397
00:21:36,304 --> 00:21:39,644
FHIR rollout and where that stands
to advance in the coming years?
398
00:21:40,889 --> 00:21:42,859
FHIR has finally taken some leaps forward.
399
00:21:42,909 --> 00:21:43,959
I do think it's interesting.
400
00:21:43,959 --> 00:21:46,599
You mentioned security and privacy, right,
went right to FHIR because I think they're
401
00:21:46,599 --> 00:21:51,339
actually very linked in a way that isn't
always obvious to the average consumer.
402
00:21:51,389 --> 00:21:54,539
The primary link here is there's
a fair amount of security and
403
00:21:54,539 --> 00:21:57,959
privacy control embedded in HIPAA
and other state and federal laws.
404
00:21:58,664 --> 00:21:59,984
About minimal necessary.
405
00:21:59,989 --> 00:22:03,719
Don't share more than you need to
particularly when you go outside
406
00:22:03,719 --> 00:22:06,929
our provider relationship, right,
so that be to your payer, it could
407
00:22:06,929 --> 00:22:08,279
be to public health and whatever.
408
00:22:08,849 --> 00:22:12,749
And sending a document with all
your data, it makes it difficult to
409
00:22:12,749 --> 00:22:14,429
redact what I'm not supposed to send.
410
00:22:15,099 --> 00:22:17,309
Imagine you're getting a mortgage.
411
00:22:17,714 --> 00:22:21,434
And the bank requires to know
your current salary and they wanna
412
00:22:21,434 --> 00:22:24,764
know it from your gross income
from last year's, uh, tax return.
413
00:22:25,334 --> 00:22:30,179
Sending your 110 page tax return to
get line one is a little silly, right?
414
00:22:30,179 --> 00:22:34,049
That's a way over disclosure of
a lot of things about you when
415
00:22:34,049 --> 00:22:35,309
they only needed one number.
416
00:22:35,819 --> 00:22:39,209
And today most exchange is
being done via documents.
417
00:22:39,209 --> 00:22:42,149
These large things with a
whole bunch of data in them.
418
00:22:42,719 --> 00:22:45,869
It makes it much more difficult
to assert this is the minimal
419
00:22:45,869 --> 00:22:47,159
necessary for the purpose.
420
00:22:47,579 --> 00:22:50,094
And I think that is a huge
advantage of FHIR there.
421
00:22:50,159 --> 00:22:53,009
There are public health departments
out there that are on the sidelines.
422
00:22:53,549 --> 00:22:54,209
During COVID.
423
00:22:54,989 --> 00:22:59,114
During COVID, we allowed public health
departments over Carequality and
424
00:22:59,114 --> 00:23:03,774
therefore over CommonWell as well to
access data about COVID patients that
425
00:23:03,774 --> 00:23:07,704
they were following up with through
the networks using the treatment lane.
426
00:23:07,824 --> 00:23:08,664
And I say the treatment lane.
427
00:23:08,664 --> 00:23:11,364
'cause the treatment lane, the
treatment use case is the most
428
00:23:11,364 --> 00:23:13,404
open public health, not as open.
429
00:23:13,404 --> 00:23:14,784
So they said, just go through
the treatment lane, we
430
00:23:14,784 --> 00:23:15,804
know it's you, it's cool.
431
00:23:16,404 --> 00:23:17,664
And they said, what about
the minimal necessary?
432
00:23:17,664 --> 00:23:21,864
I said, you tell us this data is minimal
necessary for this, and we're good.
433
00:23:22,734 --> 00:23:26,044
Only two states figured out how
to talk to their AGs to figure out
434
00:23:26,044 --> 00:23:27,424
they're allowed to make that request.
435
00:23:27,844 --> 00:23:28,024
Right?
436
00:23:28,024 --> 00:23:31,354
That minimal necessary for this use
case was everything and will only keep.
437
00:23:31,684 --> 00:23:32,824
They didn't know what to do.
438
00:23:33,154 --> 00:23:38,314
So we need the ability to be more
atomic or more focused on the data
439
00:23:38,314 --> 00:23:41,434
we exchange 'cause everybody's
actually trying to do a pretty good
440
00:23:41,434 --> 00:23:42,994
job on the security and privacy.
441
00:23:43,414 --> 00:23:47,164
It's just hard to do when you have a blunt
force instrument of a massive document.
442
00:23:47,464 --> 00:23:49,234
So I think FHIR goes a long way there.
443
00:23:49,234 --> 00:23:52,324
It also allows for more current
information, 'cause a lot of
444
00:23:52,324 --> 00:23:55,894
documents are created at the end
of exams, at the end of encounters.
445
00:23:56,224 --> 00:23:57,394
There's a ton of advantages.
446
00:23:57,394 --> 00:24:02,704
I will say though I am not... FHIR is the
panacea solution for interoperability.
447
00:24:03,054 --> 00:24:05,964
It solves some problems and it helps
with some things, but we still have
448
00:24:05,964 --> 00:24:10,044
a lot of our things to work on,
including just laying the tracks and the
449
00:24:10,044 --> 00:24:14,724
philosophical change from holding data to
exposing data as the right thing to do.
450
00:24:17,334 --> 00:24:20,904
We've talked a lot about the kind of
current advancements in interoperability
451
00:24:20,904 --> 00:24:24,174
and the progress that's being made that
advances a number of different use cases.
452
00:24:24,939 --> 00:24:28,239
Here at Newfire, we obviously serve a
lot of different innovators, including
453
00:24:28,239 --> 00:24:30,909
those in the digital health space
vendors that just sell to health
454
00:24:30,909 --> 00:24:32,439
systems and payers and things like that.
455
00:24:32,439 --> 00:24:34,925
So, what's in this broad world of interop?
456
00:24:34,925 --> 00:24:39,635
Like one mistake that you see
organizations make most commonly when
457
00:24:39,635 --> 00:24:42,665
they approach data exchange that really
ends up hurting them in the long run?
458
00:24:44,295 --> 00:24:44,955
Waiting.
459
00:24:45,405 --> 00:24:47,265
I think waiting is the
worst thing you can do.
460
00:24:47,725 --> 00:24:52,695
There's this desire to say I'll jump in
when X. And for a while it was jumping
461
00:24:52,695 --> 00:24:55,185
in when there was enough data, right?
462
00:24:55,185 --> 00:24:58,850
If you went back when I first started
doing this, going back to 2009, 2010,
463
00:25:00,045 --> 00:25:03,405
the adoption of interoperability and
connecting provider-provider was pretty
464
00:25:03,405 --> 00:25:08,775
low to the point that if a provider
pressed the button, the likelihood they
465
00:25:08,775 --> 00:25:11,475
found data for the patient in front
of 'em was actually relatively low.
466
00:25:12,155 --> 00:25:13,415
Now the problem is flipped.
467
00:25:13,415 --> 00:25:14,735
There's too much data, right?
468
00:25:14,735 --> 00:25:17,555
So all these documents are flying
around and they do have a fair amount
469
00:25:17,555 --> 00:25:18,330
of duplication right within them.
470
00:25:18,335 --> 00:25:21,215
So it is sometime difficult to work with.
471
00:25:21,845 --> 00:25:25,715
But waiting for technology to make
that better, for example, waiting
472
00:25:25,715 --> 00:25:28,775
for FHIR, like I got news for you,
the documents we sent around today
473
00:25:29,225 --> 00:25:31,325
are pretty well coded, right?
474
00:25:31,325 --> 00:25:32,945
They're just, we were lazy.
475
00:25:33,005 --> 00:25:34,325
There's a lot of industries that do this.
476
00:25:34,325 --> 00:25:37,805
They bandaided knowing that the other
side may not have a good parser.
477
00:25:38,165 --> 00:25:41,495
We made a human readable format version
of it, and that's with all display on the
478
00:25:41,495 --> 00:25:44,375
screen you get a discharge summary, which
has a human readable displayed version.
479
00:25:44,975 --> 00:25:48,635
Underneath that human readable thing
is a machine readable version that
480
00:25:48,635 --> 00:25:53,255
has lab results and times, and all
the stuff you would need to do like a
481
00:25:53,255 --> 00:25:57,620
lab chart showing how things changed
over time, both inpatient and then
482
00:25:57,620 --> 00:26:00,530
before inpatient and after, you can
show the differences in between.
483
00:26:00,910 --> 00:26:03,820
But we've seen people now look
and go, there's too much data.
484
00:26:03,820 --> 00:26:05,780
Everybody's gotta organize it
better or I'm not coming in.
485
00:26:06,250 --> 00:26:10,300
And what I've seen decade over decade,
'cause I actually say that now I've
486
00:26:10,300 --> 00:26:13,810
been doing this for a long time, is
those who wait I wouldn't say they
487
00:26:13,810 --> 00:26:16,120
get hurt, but they lose advantage.
488
00:26:16,600 --> 00:26:19,590
And negative advantage is the
same thing as getting hurt
489
00:26:19,590 --> 00:26:20,760
when the market moves fast.
490
00:26:21,175 --> 00:26:25,315
So I would say get connected and then
start finding tools for the specific
491
00:26:25,315 --> 00:26:26,635
things that you're looking for.
492
00:26:27,005 --> 00:26:30,125
It's not like a sea-change
moment when you into interop.
493
00:26:30,485 --> 00:26:33,605
It's a incremental movement that
leads to other incremental movement
494
00:26:33,605 --> 00:26:35,385
movements that leads to general success.
495
00:26:37,365 --> 00:26:41,615
And on that topic uh, what are,
what's a capability or tool in
496
00:26:41,615 --> 00:26:45,455
the interop space that you feel is
most underutilized or overlooked by
497
00:26:45,455 --> 00:26:46,925
folks looking to do data exchange?
498
00:26:48,760 --> 00:26:52,830
There are two things that that are, that
apply there and I'm a little surprised
499
00:26:52,830 --> 00:26:54,390
'cause they seem almost obvious.
500
00:26:54,450 --> 00:26:57,640
And I'll be honest a fair amount
of entities have recognized
501
00:26:57,640 --> 00:26:58,660
this and are starting to fix it.
502
00:26:59,230 --> 00:27:01,800
But the data is large.
503
00:27:01,800 --> 00:27:03,480
The data sets are not tiny.
504
00:27:03,550 --> 00:27:05,020
They do take time to move.
505
00:27:05,560 --> 00:27:09,040
The movement of my data from one
provider or a collection provider
506
00:27:09,040 --> 00:27:12,415
to another is generally not a
good thing to do interactively.
507
00:27:12,505 --> 00:27:15,565
Interactively, meaning I click
the button and I see it come up.
508
00:27:15,625 --> 00:27:17,125
It takes a little bit of time.
509
00:27:17,635 --> 00:27:22,225
I'm amazed at how many EHRs that I
consider relatively mature, don't have
510
00:27:22,225 --> 00:27:27,235
a pre-access workflow, you know, like
a, a prefetch that is based on triggers.
511
00:27:27,685 --> 00:27:30,715
Most healthcare and a lot of people
think about healthcare as having,
512
00:27:30,765 --> 00:27:32,775
different types of encounters,
513
00:27:32,775 --> 00:27:35,505
you have a thing, your ambulatory,
your wellness exam every year, you have
514
00:27:35,505 --> 00:27:39,735
specialty stuff, you have inpatient that's
scheduled, then you have things that go
515
00:27:39,735 --> 00:27:41,055
through the emergency department alike.
516
00:27:41,595 --> 00:27:44,985
We often focus in interop, or at least in
the beginning, we focused on the ED stuff.
517
00:27:44,985 --> 00:27:46,635
I don't know anything about this patient
518
00:27:46,995 --> 00:27:49,215
so I'll hit this button to
go find something out about
519
00:27:49,215 --> 00:27:50,355
them before I start a workup.
520
00:27:50,715 --> 00:27:51,645
And that makes perfect sense.
521
00:27:51,645 --> 00:27:53,985
If the person's coming through the
ED, absolutely positively get their
522
00:27:53,985 --> 00:27:56,775
allergies before you start injecting
penicillin or whatever you're gonna do.
523
00:27:56,775 --> 00:27:57,585
That makes perfect sense.
524
00:27:58,155 --> 00:28:02,485
But there's so much information
between episodes and most of the
525
00:28:02,545 --> 00:28:07,845
tolerance for delay amongst most
clinicians is incredibly short.
526
00:28:08,665 --> 00:28:10,105
Three seconds and you're gone.
527
00:28:10,105 --> 00:28:10,945
I'd say a second.
528
00:28:10,945 --> 00:28:12,605
Most people don't like the application.
529
00:28:12,875 --> 00:28:14,465
You have to prefetch stuff.
530
00:28:14,855 --> 00:28:18,395
And then if you prefetch it, based
upon the scheduled exam for tomorrow
531
00:28:18,395 --> 00:28:20,855
or the scheduled exam for next
week, 'cause again, most of it's not
532
00:28:20,855 --> 00:28:24,815
emergent, most is planned a little
bit or at least sometimes a lot.
533
00:28:25,325 --> 00:28:28,385
And from there, pre-chart
some of it do something what,
534
00:28:28,835 --> 00:28:30,545
what do computers do for us?
535
00:28:30,545 --> 00:28:33,455
Computers are programmed that
do things that are monotonous,
536
00:28:33,455 --> 00:28:35,375
that advance our cause.
537
00:28:35,375 --> 00:28:37,145
Make the computer do some work.
538
00:28:37,145 --> 00:28:42,305
Have it pre-chart out medication history
so you can do reconciliation faster
539
00:28:42,305 --> 00:28:44,585
and say These are all duplicated out,
540
00:28:44,635 --> 00:28:46,615
I see four things that
look relatively current.
541
00:28:46,855 --> 00:28:48,565
Hello, Mr. Patient or Miss Patient.
542
00:28:49,375 --> 00:28:51,025
Which one of these are
really currently active?
543
00:28:51,025 --> 00:28:51,955
What'd you take today?
544
00:28:51,955 --> 00:28:53,725
And I see four things that are candidates.
545
00:28:53,995 --> 00:28:57,205
As opposed to, tell me everything
that you wanna tell me about yourself.
546
00:28:57,485 --> 00:29:01,475
Use the data to pre-stage that,
that first interaction so you
547
00:29:01,475 --> 00:29:02,495
can get to the meaty stuff.
548
00:29:02,495 --> 00:29:06,435
I think the great power of interop
and data availability in healthcare is
549
00:29:06,435 --> 00:29:10,935
gonna be allowing for providers to work
to the level of their capabilities.
550
00:29:11,745 --> 00:29:13,845
By having more time to
interact with the patient.
551
00:29:13,935 --> 00:29:16,875
The data and the computer should
help get out of the way, not
552
00:29:16,875 --> 00:29:18,315
in the way if done correctly.
553
00:29:19,584 --> 00:29:22,644
Paul, thanks so much for a thoughtful
and insightful conversation.
554
00:29:22,674 --> 00:29:26,184
Before we close it out today,
what's one last piece of advice that
555
00:29:26,184 --> 00:29:30,114
you'd offer healthtech leaders and
innovators to navigate the sometimes
556
00:29:30,114 --> 00:29:33,439
confusing world of interop and to
drive great outcomes for patients?
557
00:29:34,730 --> 00:29:37,540
Well, first of all the main thing
I have to say, especially if you're
558
00:29:37,540 --> 00:29:41,800
on the startup side, is be prepared
for a longer haul than you'd like.
559
00:29:42,150 --> 00:29:43,465
We think of healthcare going fast.
560
00:29:44,345 --> 00:29:48,845
Healthcare technology does not
change as fast as you think it does.
561
00:29:48,845 --> 00:29:53,285
There are some brilliant engineers I've
met who worked at amazing institutions
562
00:29:53,285 --> 00:29:56,855
that paid them a boatload of money
because they were PhD, computer science,
563
00:29:56,905 --> 00:30:00,445
super geeks that saw healthcare as a
place that needed help and said, I'm
564
00:30:00,445 --> 00:30:01,915
gonna go over there and help this thing.
565
00:30:02,455 --> 00:30:05,965
And what you find out when you
get here is there is bureaucracy
566
00:30:05,965 --> 00:30:07,075
partially for a reason.
567
00:30:08,125 --> 00:30:11,485
It's a science, not just a
technology and science, it's
568
00:30:11,485 --> 00:30:13,105
actually an artful science, right?
569
00:30:13,105 --> 00:30:16,195
It's not, you're not gonna get the same
result from two different providers.
570
00:30:16,195 --> 00:30:20,215
So there's some artistic license involved
here and there's a lot of regulation.
571
00:30:20,585 --> 00:30:24,910
Be prepared with your VC and whatever
money you've got, it's gonna take
572
00:30:24,910 --> 00:30:26,395
a couple years to get traction.
573
00:30:26,445 --> 00:30:28,465
And I'm not saying that doesn't
mean you shouldn't start.
574
00:30:28,495 --> 00:30:29,695
It just means be ready.
575
00:30:30,125 --> 00:30:36,955
I often talk about leadership and how you
should act with your staff and employees
576
00:30:36,955 --> 00:30:38,305
and how you interact with others.
577
00:30:38,615 --> 00:30:41,565
Be fair, but also
understand what's going on.
578
00:30:42,105 --> 00:30:46,755
And healthcare is a it's a
hard industry to crack into.
579
00:30:46,785 --> 00:30:49,785
You need to get that those
first customers, but be patient.
580
00:30:49,785 --> 00:30:53,145
You have to have a good sense for how
long to stay the course before you pivot.
581
00:30:53,565 --> 00:30:56,235
Over pivoting in healthcare
sometimes makes you irrelevant
582
00:30:56,235 --> 00:30:57,195
'cause no one knows what you do.
583
00:30:58,618 --> 00:31:01,798
Paul, thanks again for joining us
and for sharing your perspective.
584
00:31:02,068 --> 00:31:05,818
And to our listeners, I hope you've
gained valuable insights into the ongoing
585
00:31:05,818 --> 00:31:09,718
challenges and promising solutions in
healthcare interoperability, especially
586
00:31:09,958 --> 00:31:13,688
unique opportunities around patient
access and the future of data exchange.
587
00:31:17,860 --> 00:31:20,200
Thanks again for joining
us on Hard Problems, Smart
588
00:31:20,200 --> 00:31:22,150
Solutions, the Newfire Podcast.
589
00:31:22,390 --> 00:31:23,200
Until next time.
