Latest / Becker’s Healthcare Podcast / Building a Unified Data Foundation for AI and Healthcare Innovation with Chris Carmody

Transcript

Building a Unified Data Foundation for AI and Healthcare Innovation with Chris Carmody · Becker’s Healthcare Podcast

• 0:00
  - What we wanna do with AI is help
• 0:01
  - clinicians go from being reactive
• 0:04
  - to proactive.
• 0:05
  - At Philips, we can recognize patterns in monitoring
• 0:08
  - data across populations
• 0:10
  - and apply those insights to specific patients.
• 0:14
  - Safe and responsible AI can support sharper insights,
• 0:18
  - pave the way for faster diagnosis,
• 0:21
  - and help streamline care.
• 0:23
  - That can benefit clinicians, it can benefit health
• 0:26
  - systems. And most importantly, it can benefit patients
• 0:29
  - and their families.
• 0:33
  - Phillips.
• 0:34
  - Hi, everyone. This is Scott King with Becker's
• 0:35
  - Healthcare. Thank you so much for tuning in
• 0:37
  - to the Becker's Healthcare podcast series. Joining me
• 0:39
  - for today's discussion is Chris Carmody, chief information
• 0:42
  - officer over at UPMC. Chris, thanks so much
• 0:44
  - for being here today. How are you doing?
• 0:46
  - I'm doing great, Scott. Thanks. Thanks for having
• 0:48
  - me. Absolutely. It's great to have you here
• 0:50
  - at the conference. Before we kinda dive into
• 0:52
  - our questions, just wondering if you can share
• 0:53
  - just a little bit about your background, your
• 0:55
  - career journey with us. Sure. I have been
• 0:57
  - in health care IT for about thirty years.
• 1:00
  - The last twenty eight, I've been at UPMC
• 1:02
  - in many different roles. And and, actually, in
• 1:05
  - the last couple weeks, I was just, moved
• 1:07
  - into the CIO role, which I'm very excited
• 1:09
  - about that opportunity and follow a a great
• 1:12
  - line,
• 1:13
  - from a succession perspective.
• 1:15
  - And,
• 1:16
  - as part of UPMC,
• 1:18
  - I'm actually in a position where I can
• 1:20
  - serve the community where I grew up in
• 1:21
  - the city of Pittsburgh and the surrounding regions.
• 1:24
  - So,
• 1:25
  - I'm very humbled and and motivated by that
• 1:27
  - opportunity to continue my IT career in this
• 1:30
  - leadership position. That's awesome. Well, you're helping the
• 1:32
  - community back home, and and congrats on the
• 1:34
  - promotion. That's That's wonderful. Good for you. And
• 1:36
  - then we're gonna dive into these questions here.
• 1:37
  - I I gotta lean on you for a
• 1:38
  - lot of big topics in health care right
• 1:40
  - now. You you know, health systems are receiving
• 1:43
  - a growing stream of patient data from bedside
• 1:45
  - monitoring,
• 1:46
  - ambulatory devices, and and care delivered at home.
• 1:49
  - Where's that data creating actionable insights for clinicians,
• 1:52
  - and where does it still create more work
• 1:54
  - than insight?
• 1:55
  - That's a that's a great question. I think
• 1:57
  - it's... There's an evolving answer to it in
• 1:59
  - terms of how we're dealing with all the
• 2:01
  - data that's being generated.
• 2:03
  - We started our electronic journey
• 2:06
  - decades ago and have generated a tremendous amount
• 2:09
  - of of data that we've worked towards transforming
• 2:12
  - an information and and then actionable insights. And
• 2:15
  - I think that's the key there Yeah. Is
• 2:17
  - to reduce all the noise and identify the
• 2:20
  - signals of what an action or what next
• 2:23
  - steps a clinician needs to take from a
• 2:24
  - patient care perspective. So when you add in
• 2:27
  - all the medical devices in addition to all
• 2:29
  - the different ancillary systems that are plugging into
• 2:32
  - your your electronic health record system, there's a
• 2:34
  - lot of information for that clinical care team
• 2:37
  - to actually
• 2:38
  - consume and digest and process in a short
• 2:40
  - amount of time so they're most effective
• 2:43
  - from a patient care perspective. So
• 2:45
  - I think
• 2:46
  - the benefit and the opportunity with artificial intelligence,
• 2:49
  - whether it be generative or Genentech AI, is
• 2:51
  - to actually leverage those tools to harness
• 2:53
  - all that data and summarize that information
• 2:56
  - into something that's consumable by the clinician that's
• 2:59
  - complete and accurate and has integrity, which are
• 3:02
  - so very important when you're talking about these
• 3:04
  - types of technologies from an AI perspective. But
• 3:07
  - that... That'll help
• 3:09
  - make that burden of making sure that all
• 3:12
  - the information is in one place
• 3:14
  - and available so the the best possible outcomes
• 3:17
  - can be achieved from a care perspective. Yeah.
• 3:19
  - Tuning out to noise, like you said, that's
• 3:20
  - a good way to put it. Absolutely. And
• 3:22
  - you kinda walked us into our next topic
• 3:24
  - here. When clinicians are trying to to reach
• 3:26
  - a diagnosis,
• 3:27
  - they often must piece together information from radiology,
• 3:29
  - cardiology, pathology,
• 3:31
  - urology, and other systems. Where does that fragmentation
• 3:34
  - create the most friction in your organization,
• 3:36
  - and how you bring together data sources and
• 3:38
  - address this fragmentation
• 3:40
  - or close those gaps?
• 3:42
  - We're actually in the midst of solving that
• 3:44
  - problem from a fundamental perspective.
• 3:47
  - Just as as late as,
• 3:50
  - well,
• 3:51
  - a year ago, we had 10 different electronic
• 3:53
  - health records,
• 3:55
  - which are really the core platform for all
• 3:57
  - that different data to come in from the
• 3:58
  - ancillary, like radiology,
• 4:00
  - pathology, lab.
• 4:02
  - And so when you're seeing patients traverse those
• 4:04
  - different EHRs, stitching that data back together and
• 4:07
  - making it available
• 4:08
  - is a is a challenge. It's it's like
• 4:11
  - sprinting a marathon from an IT perspective. A
• 4:13
  - lot of our efforts over the last decade
• 4:15
  - have been trying to make it a more
• 4:16
  - seamless experience for clinicians to have all that
• 4:18
  - data in one place. And we've been really
• 4:21
  - successful at it, but, again, it's taken a
• 4:22
  - lot of hard work and effort
• 4:24
  - to bring all that data together and make
• 4:26
  - sense of it in one place.
• 4:28
  - What we've been doing the last few years
• 4:30
  - is on this journey to a single EHR
• 4:32
  - platform
• 4:33
  - to help create that standardization and put all
• 4:35
  - that data into one place. So we've actually
• 4:37
  - backloaded
• 4:38
  - years' worth of data from those other EHR
• 4:40
  - source
• 4:41
  - systems. And just so having that that base
• 4:43
  - foundation and having
• 4:45
  - the lab, the radiology, and all the other
• 4:46
  - data sources come into that single spot
• 4:49
  - is... It took... From our perspective is a
• 4:51
  - huge win. It's gonna create,
• 4:54
  - a better experience for the clinician, obviously, from
• 4:57
  - the patient. The patient actually accesses a lot
• 4:58
  - of that information from their... Our patient portal,
• 5:01
  - our myUPMC
• 5:02
  - system. So it's it's it's it's very exciting
• 5:05
  - for us. And, again, it positions us from
• 5:08
  - an opportunistic perspective to actually innovate off of
• 5:10
  - that and and just drive health care innovation
• 5:13
  - from there. Yeah. It sounds like a big
• 5:15
  - relief in terms of efficiency probably for everyone.
• 5:18
  - Absolutely.
• 5:20
  - You you mentioned patients. Patients increasingly move
• 5:23
  - among the hospital outpatient settings and and home.
• 5:26
  - Where does clinical context most often get lost
• 5:28
  - during those transitions,
• 5:30
  - and what would it take to give care
• 5:31
  - teams more continuous view of the patient?
• 5:34
  - You you just said it, Scott. It's it's
• 5:36
  - during that transition when they're going from inpatient
• 5:39
  - to outpatient or vice versa.
• 5:41
  - It's making sure the data
• 5:43
  - travels with the patient that's available, and that's
• 5:45
  - why we made the,
• 5:47
  - pretty big investment in time and resources
• 5:50
  - to put UPMC onto one EHR system, which
• 5:53
  - makes all that data
• 5:55
  - seamlessly available in one place for clinicians. So
• 5:58
  - as a patient traverses our system,
• 6:00
  - they're able... Their data is able to move
• 6:02
  - with them. And what our EHR does, which
• 6:05
  - it's pretty well documented,
• 6:06
  - we're implementing Epic.
• 6:08
  - But what Epic does is actually help,
• 6:11
  - that patient if they would traverse outside of
• 6:13
  - a UPMC facility for that data to be
• 6:15
  - shared back and forth, as they traverse their
• 6:18
  - health care needs.
• 6:19
  - Yeah. I get to AI now. You know,
• 6:21
  - we're gonna cover AI in some respect.
• 6:23
  - With so many AI use cases competing for
• 6:26
  - attention, how do you distinguish a compelling pilot
• 6:28
  - from something
• 6:30
  - ready to become part of everyday care, and
• 6:32
  - what must be in place to scale it
• 6:34
  - responsibly?
• 6:36
  - So I think there's great opportunity with AI,
• 6:39
  - and I think one of the challenges is
• 6:40
  - to
• 6:41
  - actually slow things down in order to accelerate.
• 6:44
  - Sure. That we have to have the right
• 6:45
  - processes in place, the right governance to identify,
• 6:47
  - hey. These are the these are the top
• 6:50
  - three,
• 6:51
  - five, 10 problems that we're trying to solve
• 6:53
  - as an organization
• 6:54
  - and focus and channel our efforts on making
• 6:57
  - those successful
• 6:58
  - so we can build off of that and
• 7:00
  - and and move things forward and and take
• 7:02
  - advantage of the opportunity with AI. But it
• 7:04
  - really does start with that governance piece Right.
• 7:06
  - And the processes around to support it. And
• 7:09
  - then the last piece, which is just as
• 7:11
  - important, is is making sure that the the
• 7:13
  - users that are having their workflows changed to
• 7:17
  - make sure that they're trained to understand the
• 7:18
  - why behind
• 7:20
  - utilizing that technology and what the benefit is.
• 7:23
  - Right. Right. The reason you're using it and
• 7:25
  - and the result you're trying to achieve.
• 7:27
  - As technology relationships move beyond
• 7:29
  - one time purchases, what should a partner remain
• 7:32
  - accountable for
• 7:33
  - after
• 7:34
  - go live to make the relationship
• 7:36
  - genuinely valuable over time?
• 7:38
  - Accountability is a great term to use with
• 7:40
  - our technology partners.
• 7:43
  - What is unfortunately the case
• 7:45
  - far too often for my liking
• 7:48
  - is our technology partners come in and
• 7:50
  - they make the sale, we do the transaction,
• 7:53
  - and then they disappear.
• 7:55
  - And what we look for, and I encourage
• 7:57
  - all health care IT to look for,
• 7:59
  - are partners that stay engaged, that stay accountable,
• 8:03
  - and work with you to actually solve the
• 8:05
  - problems. Because their technology is not the end
• 8:07
  - all be all, especially when you're acquiring something
• 8:09
  - out of the gate. You have to integrate
• 8:11
  - it and make sure it works with the
• 8:13
  - rest of your technology platforms, let alone integrate
• 8:16
  - into the workflows for the users of that
• 8:18
  - technology. But
• 8:20
  - I I personally think it's a a great
• 8:23
  - place to be in to... That your technology
• 8:25
  - partners are always a little uncomfortable.
• 8:27
  - Anytime they get comfortable, that's a bad sign
• 8:30
  - that they're being complacent, and we need to
• 8:32
  - work together and partner and innovate to drive
• 8:34
  - health care forward. Yeah. Yeah. Absolutely. For our
• 8:37
  - last question, what's the future of integrated diagnostics
• 8:41
  - or just diagnostic services, and where do you
• 8:43
  - see yourself in the next five years?
• 8:46
  - That's... Those are some big questions.
• 8:48
  - The five the five year outlook,
• 8:51
  - I'm I'm an opportunistic
• 8:53
  - person, and I... I'm... I believe that health
• 8:56
  - care is only gonna get... Be better,
• 8:58
  - be more accessible,
• 9:00
  - have higher quality and better outcomes for our
• 9:02
  - patients that we serve across
• 9:05
  - the entire US and beyond because of technology.
• 9:08
  - And I I hope for... And to me,
• 9:11
  - a successful technology implementation, whatever that might be,
• 9:13
  - however we use technology,
• 9:15
  - really is...
• 9:17
  - It's at its best when the technology fades
• 9:19
  - to the background Yeah. When it's not a
• 9:22
  - distraction for our clinical users and end users
• 9:25
  - trying to use something, do something when they
• 9:27
  - should be focusing on the patient. So to
• 9:30
  - me, that... That's where I hope to be
• 9:31
  - in five years and how technology will help
• 9:33
  - us get there. And that, to me, that's
• 9:35
  - the the the measuring stick on whether we're
• 9:38
  - successful or not. On the other, like, diagnostic
• 9:41
  - side, I think there's great opportunity there as
• 9:43
  - well, but I think we have to
• 9:45
  - be measured with our approach and adoption
• 9:48
  - of leveraging things like AI to help perform
• 9:51
  - some of that
• 9:52
  - diagnostic work. There's obviously been a lot of
• 9:55
  - success cases in the imaging space,
• 9:57
  - and that's a great place to build off
• 9:59
  - of. But there's also a lot of data,
• 10:00
  - and, again, we just have to be very
• 10:02
  - cautious about how we approach that and and
• 10:04
  - and leverage the technology. We don't wanna just
• 10:06
  - adopt technology for technology's sake. Right. Again, it
• 10:09
  - gets back to... Which I believe is all
• 10:11
  - of our mission, is to provide the best
• 10:13
  - patient care, to improve patient safety, and ultimately
• 10:15
  - have those great outcomes that we're that we're
• 10:18
  - searching for. A 100%. Chris, thanks so much
• 10:20
  - for joining us on the podcast. It was
• 10:22
  - a great conversation. I look forward to work
• 10:23
  - with you again soon. Thanks, Scott. Appreciate being
• 10:25
  - here. And, again, appreciate everything Becker's HealthCare does
• 10:28
  - for the industry. Such such a great event,
• 10:30
  - and I encourage all to join us here
• 10:32
  - next time. Oh, thanks so much. We appreciate
• 10:33
  - the support.