Latest / Becker’s Healthcare Podcast / Introducing AI While Maintaining Clinician Trust

Transcript

Introducing AI While Maintaining Clinician Trust · Becker’s Healthcare Podcast

• 0:00
  - Hello, everyone. This is Erica Spicer Mason with
• 0:02
  - Becker's Healthcare. Thank you so much for tuning
• 0:05
  - in to the Becker's Healthcare podcast series.
• 0:08
  - Today, we're going to talk about how health
• 0:10
  - system leaders are thoughtfully evaluating
• 0:13
  - AI powered clinical decision support
• 0:15
  - and what it takes to successfully introduce those
• 0:18
  - tools into clinical workflows
• 0:20
  - while maintaining clinician trust and focusing on better
• 0:23
  - patient care.
• 0:24
  - So joining me for this discussion, we have
• 0:27
  - Allison Schneider, the director of applications
• 0:30
  - at Aspen Valley Health in Colorado.
• 0:32
  - And Allison will share how her team approaches
• 0:35
  - new technologies,
• 0:36
  - the factors that guide their AI adoption decisions,
• 0:39
  - and also what they've learned from implementing
• 0:42
  - clinical key AI to support clinicians at the
• 0:44
  - point of care.
• 0:45
  - Allison, without further ado, I wanna welcome you
• 0:47
  - to the podcast, and thank you so much
• 0:49
  - for being here today.
• 0:51
  - Of course. Thanks for that awesome introduction.
• 0:53
  - As mentioned, my name is Allison Schneider, and
• 0:56
  - I'm the director of IT applications at Aston
• 0:59
  - Valley Health.
• 1:00
  - I joined Aston Valley Health about four years
• 1:02
  - ago as an IT analyst,
• 1:04
  - supporting EPIC's ClinDoc and STORK applications.
• 1:08
  - In that role, I was able to work
• 1:10
  - closely with our clinical teams and understand how
• 1:12
  - technology
• 1:13
  - directly impacts and affects, their day to day
• 1:16
  - experience.
• 1:17
  - From there, I became our clinical applications manager,
• 1:21
  - overseeing the clinical applications within EPIC at Aston
• 1:24
  - Valley Health.
• 1:26
  - And after a little over a year in
• 1:27
  - that position,
• 1:29
  - I transitioned into my current director role where
• 1:31
  - I now oversee all of our applications, including
• 1:34
  - EPIC at Aspen Valley Health.
• 1:37
  - Just a little bit about what I really
• 1:38
  - enjoy,
• 1:39
  - within this work is being at the intersection
• 1:42
  - of technology,
• 1:44
  - clinical care, and operational strategies.
• 1:47
  - Ultimately, our goal, at Aspen Valley Health is
• 1:49
  - to make technology work better for the people
• 1:52
  - providing the care and most importantly,
• 1:55
  - our patients.
• 1:57
  - Such a great intro, Allison. Thanks for telling
• 1:59
  - us a bit more about yourself and,
• 2:02
  - and really your role. It sounds like you've
• 2:04
  - got a lot on your plate, and it's
• 2:06
  - a very dynamic position to be in. But
• 2:08
  - I love your note there about, you know,
• 2:10
  - the the continued focus on excellent patient care.
• 2:13
  - And that's really what where I wanted to
• 2:15
  - start our conversation today.
• 2:17
  - Even in our own reporting at Becker's, we've
• 2:19
  - we've outlined how Aspen Valley Health has this
• 2:22
  - reputation for really embracing innovation
• 2:25
  - while staying deeply committed to patient care. So
• 2:28
  - when a new tool like AI comes across
• 2:31
  - your desk and in your role, how are
• 2:33
  - you deciding whether it's the right fit for
• 2:35
  - your organization?
• 2:38
  - Very hot topic. I'm sure not just at
• 2:40
  - Aston Valley help,
• 2:42
  - but specifically for ABH.
• 2:44
  - We typically
• 2:45
  - take, an epic first approach.
• 2:48
  - So before introducing another platform, we look at
• 2:50
  - what Epic already has to offer or has
• 2:52
  - on its road map for the future.
• 2:54
  - Epic is moving, as we know, incredibly quick
• 2:57
  - in the AI space, and that gives us
• 2:59
  - a strong foundation for evaluating new capabilities.
• 3:03
  - But being available doesn't automatically make a tool
• 3:06
  - the right fit, especially for Aspen Valley Health,
• 3:09
  - being a critical access hospital.
• 3:11
  - We look at whether it solves a clearly
• 3:14
  - defined problem,
• 3:15
  - integrates into existing workflows,
• 3:18
  - protects patient information, and genuinely
• 3:20
  - improves the clinician or patient experience.
• 3:23
  - We also consider how it fits within the
• 3:25
  - direction established by our AI council that we
• 3:28
  - have at Aspen Valley Health and then ultimately
• 3:30
  - as an organization as a whole.
• 3:33
  - For us at Aspen Valley Health,
• 3:35
  - AI should not be technology for technology's sake.
• 3:38
  - Ultimately, we are aiming for it to address
• 3:41
  - a real need and create measurable value without
• 3:43
  - adding unnecessary complexities
• 3:45
  - to our existing workflows.
• 3:48
  - Yeah. Sounds like a really sound process that
• 3:51
  - that your team has outlined, Allison.
• 3:54
  - And kind of in the same vein, I
• 3:55
  - wanted to also touch on the concept
• 3:57
  - of trust.
• 3:59
  - I know trust comes up as an issue
• 4:01
  - with AI frequently, whether we're talking about clinicians
• 4:04
  - or patients.
• 4:06
  - So from your lens, what are the nonnegotiables
• 4:09
  - for for earning that trust?
• 4:11
  - For sure. That's another great question.
• 4:14
  - Ultimately, the patient has to remain at the
• 4:16
  - center of every conversation. That is the first
• 4:19
  - nonnegotiable
• 4:20
  - for us at Aspen Valley Health. Some other
• 4:22
  - things that we take into consideration
• 4:24
  - is we need appropriate privacy and security protections,
• 4:28
  - transparency about how the technology is being used,
• 4:32
  - reliable clinical content, and clear human oversight is
• 4:35
  - also really important for us and definitely a
• 4:38
  - nonnegotiable.
• 4:40
  - The AI that we are, you know, bringing
• 4:42
  - into our system should ultimately support our clinicians
• 4:45
  - and not replace their judgment or create another
• 4:48
  - disconnected step in their workflow.
• 4:50
  - The integration is also extremely important for us,
• 4:53
  - especially coming from the technology side. We ask
• 4:56
  - whether the tool can complement our existing epic
• 4:59
  - workflows and provide information at the right moment.
• 5:03
  - Trust develops when clinicians
• 5:05
  - understand the tool,
• 5:07
  - know its limitations, and see that it consistently
• 5:09
  - supports safer and more efficient care overall.
• 5:14
  - Mhmm. Yeah. So it sounds like really what's
• 5:16
  - important here is keeping patients at the center
• 5:18
  - of every conversation when it comes to AI,
• 5:21
  - ensuring that it's secure, protected,
• 5:24
  - that the clinical content is reliable, and that
• 5:26
  - it's integrated, whether that's in clinician workflows or
• 5:29
  - on the technology side.
• 5:32
  - So I'm curious. Is there a tool that
• 5:33
  - comes to mind that you've recently brought on
• 5:36
  - that that clears that bar in those expectations?
• 5:39
  - Yeah. That's another great question. A tool that
• 5:41
  - we recently introduced to Aston Valley Health was
• 5:44
  - Clinical Key AI.
• 5:46
  - We were super drawn to Clinical Key, because
• 5:49
  - it addresses
• 5:50
  - very practical challenges,
• 5:52
  - including the clinicians need to find relevant evidence
• 5:55
  - based information quickly,
• 5:57
  - often while caring for a patient, while managing
• 6:00
  - several competing priorities. So super quick and reliable.
• 6:04
  - Mhmm.
• 6:05
  - Absolutely.
• 6:05
  - And I'm hoping to know too. So with
• 6:08
  - Clinical Key AI, what really drew you to
• 6:10
  - the tool, and what were you hoping to
• 6:12
  - or what were you hoping it would solve,
• 6:14
  - I should say?
• 6:15
  - Definitely. So our goal was to give our
• 6:17
  - clinicians faster access to more synthesized clinical information
• 6:21
  - to support their decision making. Right? On the
• 6:23
  - floor, things can happen really quickly, rapidly,
• 6:26
  - specifically in our emergency department. And just as
• 6:29
  - importantly, we wanted that information to be accessible
• 6:31
  - within their existing workflows rather than requiring them
• 6:34
  - to leave Epic and search through multiple external
• 6:37
  - resources.
• 6:38
  - So clinicalkey.ai,
• 6:40
  - really cleared the bar for us because it
• 6:42
  - supported a real clinical need, came from a
• 6:44
  - trusted content provider, and could ultimately be integrated
• 6:48
  - into our environment that our clinicians already use.
• 6:51
  - Great to hear that you're seeing early success
• 6:54
  - with the tool.
• 6:55
  - I'm curious, Allison. How have clinicians
• 6:57
  - taken to it so far? And and is
• 6:59
  - there a moment or even an example that
• 7:02
  - really captures how
• 7:03
  - the staff are finding and using evidence in
• 7:05
  - the middle of a busy clinical workday?
• 7:08
  - Of course. That's another great question.
• 7:10
  - So far, the response has been very positive.
• 7:14
  - Obviously, with NEAI, there's some natural and initial
• 7:17
  - hesitation
• 7:18
  - whenever we do introduce a new tool,
• 7:21
  - especially within health care.
• 7:23
  - Something that has helped with this adoption is
• 7:25
  - having clinical key AI, like I said, integrated
• 7:28
  - directly into our Epic system.
• 7:31
  - So our clinicians can access it without significantly
• 7:33
  - disrupting their workflows, which has been super helpful.
• 7:36
  - Overall, we've seen particular value in our emergency
• 7:40
  - department
• 7:41
  - where our physicians often need to evaluate a
• 7:43
  - broad range of possibilities very quickly.
• 7:46
  - Clinical key AI can help them explore and
• 7:50
  - develop a differential diagnosis by some synthesizing
• 7:53
  - relevant clinical information
• 7:55
  - at the point of care.
• 7:57
  - Instead of spending valuable time reviewing extensive literature
• 8:00
  - and, searching several different resources,
• 8:03
  - clinicians can get the pertinent information more efficiently.
• 8:06
  - It isn't making the clinical decision for them,
• 8:09
  - but instead, it's helping them find and evaluate
• 8:12
  - evidence more quickly so that they can apply
• 8:14
  - their own expertise.
• 8:17
  - Yeah. And such an important distinction, Allison. I
• 8:19
  - appreciate you making that as I know, that
• 8:22
  - goes right back into the trust factor that
• 8:24
  - we were talking about and, you know, still
• 8:26
  - having a human in the loop here.
• 8:28
  - And, you know, this has been so helpful
• 8:30
  - to hear what your journey has looked like.
• 8:33
  - For another leader or peer who might just
• 8:35
  - be beginning to explore clinical decision support tools,
• 8:39
  - What's one lesson that you would share from
• 8:41
  - Aspen Valley Health that you'd want them to
• 8:43
  - hear or to know?
• 8:45
  - For sure. This has been a big journey
• 8:47
  - for us. And as we know, AI is
• 8:49
  - coming out rapidly.
• 8:51
  - So a lot of quick decisions on our
• 8:53
  - end and how we wanna roll it out.
• 8:55
  - My biggest recommendation would be to begin with
• 8:57
  - the problem, not the product.
• 9:00
  - Be very clear about what challenge you're trying
• 9:02
  - to solve and then carefully research which tool
• 9:04
  - is the best fit for your organization.
• 9:07
  - It's also important to involve the people who
• 9:09
  - will actually be using the technology.
• 9:11
  - IT can make recommendations, and we can try
• 9:13
  - to support our clinicians, but, ultimately,
• 9:16
  - having that clinician engagement
• 9:18
  - and then education and ongoing support are very
• 9:21
  - essential.
• 9:22
  - And like I said earlier, there may be
• 9:24
  - some hesitation initially, but adoption can happen relatively
• 9:28
  - quickly when clinicians understand the purpose of the
• 9:31
  - tool, trust its underlying content, and see how
• 9:33
  - it can help them within their daily work
• 9:35
  - and their workflows.
• 9:38
  - Well, such helpful perspective, Allison. Thanks for sharing
• 9:41
  - what's worked at Aspen Valley Health and,
• 9:44
  - some of the, you know, challenges you may
• 9:46
  - have run into and how you've navigated them.
• 9:48
  - I think it's really helpful for listeners to
• 9:50
  - to hear. But before we wrap, is there
• 9:53
  - anything else that we haven't touched on or
• 9:55
  - anything else you wanted to leave listeners with?
• 9:57
  - For sure. I think my last piece for
• 9:59
  - you is that, you know, we know AI
• 10:01
  - is advancing incredibly quick, and it has tremendous
• 10:04
  - potential to improve health care workflows
• 10:07
  - while also reducing some of the burden of
• 10:09
  - the clinician experience.
• 10:11
  - At the same time, speech should not replace
• 10:13
  - thoughtful decision making. Health care organizations
• 10:16
  - still need to be strategic about the tools
• 10:18
  - they select,
• 10:19
  - intentional about governance, and clear about the outcomes
• 10:22
  - that they hope to ultimately achieve.
• 10:24
  - So when AI is implemented responsibly with patients
• 10:27
  - at the center and clinicians actively involved, I
• 10:30
  - do think it can be super meaningful in
• 10:33
  - the way that they support their care and
• 10:35
  - their delivery for the patients at the center.
• 10:38
  - Something else I just wanna throw out that
• 10:40
  - vendor partnership is another critical piece, you know,
• 10:44
  - with AI. And I just want to add
• 10:46
  - that we've received excellent support from Elsevier
• 10:49
  - and that their partnership has been very valuable
• 10:51
  - throughout our implementation and rollout specifically for clinical
• 10:54
  - key AI. So thank you for that.
• 10:57
  - Well, Allison, thank you again for spending time
• 11:00
  - with us today and sharing all of your
• 11:02
  - great thoughts and your insights.
• 11:04
  - Of course, we also wanna thank today's podcast
• 11:06
  - sponsor, Elsevier.
• 11:08
  - Listeners, be sure to tune in to more
• 11:10
  - podcasts from Becker's by visiting our podcast page
• 11:13
  - at beckershospitalreview.com.