Latest / Becker’s Healthcare Podcast / Healthcare Upside / Down: What Makes AI Work in the Revenue Cycle & What Has to Be True First

Transcript

Healthcare Upside / Down: What Makes AI Work in the Revenue Cycle & What Has to Be True First · Becker’s Healthcare Podcast

• 0:00
  - Hello, everyone. This is Molly Gamm with Becker's
• 0:01
  - Healthcare. Welcome to Healthcare Upside Down, a podcast
• 0:04
  - by Becker's Healthcare and ECG management consultants in
• 0:07
  - which we'll explore the upsides and downsides of
• 0:09
  - health care and the industry's most current trends.
• 0:12
  - So thank you so much for tuning in.
• 0:14
  - Joining me for today's discussion is Chris Ford,
• 0:17
  - associate principal of ECG Management Consultants,
• 0:20
  - and Amy Sebero, chief growth officer with Janus
• 0:23
  - Health.
• 0:24
  - Chris, Amy, thanks for being here today, and
• 0:26
  - thanks for joining me live on-site at our
• 0:29
  - event here in Chicago this week. Thanks for
• 0:31
  - having us. Thank you. Great to be here.
• 0:33
  - Well, I I just shared a little bit
• 0:35
  - about you, just your your name and titles
• 0:36
  - and organizations, but let's hear more. Let's, learn
• 0:39
  - more about your role, your organizations, and the
• 0:41
  - work you're focused on right now. And maybe
• 0:43
  - as you do so, Amy, I'll turn to
• 0:45
  - you first. Can you share one thing
• 0:47
  - about the revenue cycle conversation in 2026
• 0:50
  - that looks and sounds quite different to you
• 0:52
  - than it did even a year ago? Sure,
• 0:54
  - Molly. Thank you again for the opportunity to
• 0:57
  - be here today.
• 0:59
  - Well, I've been in revenue cycle for forty
• 1:02
  - years, and, really, Janice focuses on helping health
• 1:06
  - systems use
• 1:07
  - data and AI
• 1:10
  - to pinpoint patterns in their workflows where processes
• 1:13
  - are breaking down in the revenue cycle.
• 1:16
  - And it's an intelligence
• 1:18
  - layer
• 1:19
  - that's built into revenue cycle operations
• 1:23
  - that provides visibility
• 1:25
  - into
• 1:26
  - where
• 1:27
  - they're losing money and productivity.
• 1:30
  - And then we automate those opportunities
• 1:34
  - and help them continuously
• 1:36
  - measure performance.
• 1:38
  - In terms of the questions
• 1:40
  - that were being asked a few years ago,
• 1:43
  - I think
• 1:44
  - general question was being asked, where can we
• 1:47
  - use AI?
• 1:49
  - Now it's more about how can we put
• 1:52
  - AI to work at scale,
• 1:54
  - embed it into the way our teams are
• 1:57
  - actually operating,
• 1:59
  - And most importantly,
• 2:01
  - how do we prove
• 2:03
  - that it's actually moving the financial needle?
• 2:06
  - And there was an article that came out
• 2:09
  - on Friday
• 2:10
  - in Becker's
• 2:12
  - that 42%
• 2:13
  - of health systems are deploying
• 2:15
  - AI
• 2:16
  - across multiple use cases,
• 2:20
  - but only 4%
• 2:22
  - have achieved
• 2:23
  - scaled implementation
• 2:25
  - with measurable outcomes.
• 2:27
  - And that's a very small proportion. And I
• 2:29
  - found that interesting
• 2:31
  - because,
• 2:33
  - it just shows
• 2:34
  - how challenging it is to actually prove AI
• 2:38
  - is moving the financial needle. Mhmm. And so
• 2:41
  - it is very critical
• 2:43
  - to really be able to
• 2:45
  - build that intelligence layer into the revenue cycle.
• 2:49
  - Mhmm. Thank you so much. I'm
• 2:50
  - it's a concerning gap, but I'm also happy
• 2:53
  - to hear numbers like that in this conversation
• 2:54
  - as we get started. That that 42% down
• 2:57
  - to four, that's quite quite a difference.
• 2:59
  - Let's turn to you and hear about the
• 3:01
  - biggest way you've seen the conversation about AI
• 3:04
  - in the revenue cycle shift, Chris, and also
• 3:06
  - just more about you and your role at
• 3:08
  - ECG Management Consultants.
• 3:09
  - Yeah. Sure. Thanks, Molly, and thanks for having
• 3:11
  - us today. ECG is a national health care
• 3:14
  - consulting firm, health care only, been around for
• 3:17
  - fifty plus years.
• 3:18
  - I spend most of my time in the
• 3:20
  - revenue cycle,
• 3:22
  - and really where we've been focused,
• 3:24
  - in the last handful of years has been
• 3:25
  - all things margin. So not just revenue cycle
• 3:28
  - optimization,
• 3:30
  - or operations, but sort of that underlying
• 3:33
  - push for margin, need for margin.
• 3:36
  - So a lot of work, building out things
• 3:38
  - in the revenue,
• 3:39
  - integrity, charge capture areas, really looking for that
• 3:42
  - top line incremental revenue
• 3:44
  - while still doing the same blocking and tackling
• 3:47
  - we've been doing for years, you know, denial
• 3:48
  - prevention,
• 3:50
  - denials management, really trying to find those efficiencies
• 3:53
  - in the workforce.
• 3:54
  - And so, you know, as we think about
• 3:57
  - the last couple of years and how the
• 3:58
  - conversation has changed recently,
• 4:00
  - to build on what Amy had said, you
• 4:03
  - know, I think it's moved very rapidly
• 4:05
  - from
• 4:07
  - what can we automate or how much of
• 4:08
  - this,
• 4:09
  - can be, you know, reliably automated
• 4:12
  - to really shifting that conversation to
• 4:15
  - how much of this actually requires a human
• 4:17
  - touch.
• 4:18
  - Right? And so really sort of refocusing that
• 4:22
  - conversation. You know, I've had a lot of
• 4:23
  - clients who have started to use
• 4:26
  - things like, or words and and sayings like
• 4:29
  - reimagining the workforce
• 4:30
  - or redesigning
• 4:32
  - the org structure. And so really starting to
• 4:35
  - identify
• 4:36
  - where we have opportunities
• 4:38
  - to leverage our our valuable resources while
• 4:42
  - using AI to do some of those more
• 4:44
  - routine
• 4:45
  - tasks that are are repetitive, but also really
• 4:48
  - resource intensive.
• 4:50
  - I stay with you, Chris, on that point
• 4:51
  - because I think Amy had just illustrated those
• 4:53
  - numbers she had shared that although AI is
• 4:56
  - dominating every conversation in health care right now,
• 4:59
  - including at this at this conference that we're
• 5:01
  - we're all at, Where do you see that
• 5:03
  - conversation being the most productive, and where to
• 5:06
  - those numbers is it moving ahead of the
• 5:08
  - operational work that also has to happen alongside
• 5:11
  - it?
• 5:12
  - Yeah. I find the conversation to be most
• 5:15
  - productive
• 5:15
  - when we're talking about AI
• 5:18
  - to reduce
• 5:19
  - the administrative burden or the administrative
• 5:22
  - friction.
• 5:23
  - You know, there's a lot of use cases
• 5:25
  - out there already, whether that's prioritization
• 5:28
  - of of the work,
• 5:29
  - you know, predictive analytics,
• 5:32
  - looking ahead to see what might be denied,
• 5:34
  - you know, agentic AI now being able to
• 5:36
  - do some of those,
• 5:38
  - again, more more monotonous
• 5:41
  - things like
• 5:42
  - calling the insurance
• 5:44
  - plan to sit on the phone and check
• 5:47
  - insurance.
• 5:48
  - So I still think that that's the the
• 5:50
  - value play right now is is where can
• 5:52
  - we use this to reduce administrative burden.
• 5:56
  - Where I do see
• 5:57
  - it getting a little bit ahead
• 5:59
  - is,
• 6:01
  - when when AI is teed up as the
• 6:03
  - solution. Mhmm.
• 6:04
  - Mhmm.
• 6:05
  - AI is great at scaling processes,
• 6:09
  - finding those efficiencies,
• 6:11
  - but it's not going to fix
• 6:14
  - poor processes.
• 6:16
  - If you're making
• 6:17
  - decisions based on bad data,
• 6:20
  - if you don't have tight operational accountabilities,
• 6:23
  - AI is not going to fix those things
• 6:25
  - for you.
• 6:26
  - And so it's not quite yet the solution,
• 6:29
  - although it is a really big component
• 6:31
  - and does have the opportunity to really supercharge
• 6:34
  - your your operation.
• 6:36
  - Amy, if there's more you wanna add to
• 6:37
  - to that response from Chris, I'd welcome your
• 6:39
  - thoughts on that one too. Sure. I think
• 6:42
  - there are a lot
• 6:43
  - of leaders that are looking to AI to
• 6:46
  - fix some of the fragmentation
• 6:48
  - out there. We've operated in silos for years
• 6:51
  - and
• 6:52
  - applied a lot of point solutions into various
• 6:55
  - areas of the revenue cycle,
• 6:57
  - and some are looking to AI to fix
• 7:00
  - some of that fragmentation.
• 7:02
  - And in fact,
• 7:03
  - if you're not careful, it will actually reinforce
• 7:06
  - it. So I completely
• 7:08
  - agree with Chris
• 7:10
  - that it's not going to fix a broken
• 7:13
  - process.
• 7:14
  - Where I believe
• 7:15
  - AI truly becomes effective
• 7:18
  - is when you have that connected view of
• 7:20
  - the revenue cycle from front to back end.
• 7:24
  - It gives you a holistic
• 7:26
  - understanding of your revenue cycle operations,
• 7:29
  - and then you can build that intelligence
• 7:32
  - layer
• 7:33
  - into your revenue cycle.
• 7:35
  - You end up with a smarter revenue cycle,
• 7:38
  - which
• 7:39
  - adapts. And then, ultimately, what you want is
• 7:42
  - a revenue cycle that performs better over time.
• 7:46
  - Mhmm. Let's talk about that a little bit.
• 7:48
  - I I whilst we're thinking about the the
• 7:50
  - front to back end, denials are can be
• 7:52
  - a really useful test case here because this
• 7:55
  - is an area where health systems have invested
• 7:57
  - heavily over the years in both people and
• 8:00
  - technology.
• 8:01
  - And there's some real wins in some places,
• 8:03
  - and then there's a lot of stalled progress
• 8:05
  - in in others.
• 8:06
  - What do you see distinguishing the two?
• 8:09
  - What what does the operational discipline
• 8:11
  - behind true denial prevention actually look like?
• 8:14
  - So
• 8:15
  - denial prevention
• 8:17
  - has been
• 8:18
  - an ongoing question
• 8:20
  - as long as I've been in revenue cycle
• 8:23
  - and it's never going to go away.
• 8:26
  - Payer behavior continues to evolve and it's happening
• 8:30
  - more and more rapidly.
• 8:32
  - Payers are
• 8:34
  - very effectively
• 8:36
  - using AI.
• 8:37
  - And I think forward thinking
• 8:40
  - organizations
• 8:41
  - are using denials as intelligence
• 8:45
  - to continuously
• 8:47
  - adapt
• 8:48
  - as changes in behavior happen.
• 8:52
  - So where I think AI
• 8:55
  - really becomes powerful,
• 8:57
  - and this is one of the things Janice
• 8:59
  - is doing, is identifying
• 9:01
  - process variations,
• 9:04
  - For example, in working certain types of denials,
• 9:07
  - then measuring
• 9:09
  - those variations
• 9:10
  - against KPIs,
• 9:12
  - developing the optimal path
• 9:14
  - for the work for that given process
• 9:17
  - using AI
• 9:19
  - that will yield the best financial outcomes.
• 9:23
  - And then, of course,
• 9:24
  - with any type of new process,
• 9:28
  - consistently measuring the team's performance,
• 9:31
  - against
• 9:32
  - that optimal path.
• 9:35
  - Thanks for talking or speaking to that. And
• 9:37
  - I I wanna turn to you, Chris, because
• 9:39
  - I think as we think about measuring the
• 9:40
  - team's performance and, Amy, you kind of painted
• 9:43
  - a picture of how historical this challenge has
• 9:45
  - been for health systems and how AI is
• 9:47
  - giving
• 9:48
  - bigger payers quite quite an edge.
• 9:51
  - My question for you, Chris, is if if
• 9:53
  - is denial prevention realistically
• 9:56
  - a revenue cycle responsibility,
• 9:58
  - or do you see it requiring ownership well
• 10:00
  - beyond the revenue cycle team?
• 10:02
  - I love
• 10:03
  - this question because it's one of the things
• 10:05
  - I preach to my clients regularly when I
• 10:07
  - come in and they say, oh, it's those
• 10:08
  - people in the revenue cycle.
• 10:10
  - The revenue cycle starts the moment that patient
• 10:13
  - interacts with the health system.
• 10:15
  - And so while the revenue cycle certainly owns
• 10:18
  - the scorecard,
• 10:19
  - we don't own the outcomes
• 10:21
  - by by ourselves. Right? And so
• 10:24
  - appreciating the fact that revenue cycle isn't end
• 10:27
  - to end,
• 10:28
  - Soon as that patient picks up the phone
• 10:30
  - and and calls or their provider sends a
• 10:32
  - referral through the front door, the revenue cycle
• 10:35
  - has started. We need to collect really good
• 10:37
  - demographics. We need to get that insurance plan
• 10:40
  - on file and do the initial query. We
• 10:42
  - need to determine if the services are gonna
• 10:44
  - require authorization.
• 10:46
  - Do we need to have a referral, and
• 10:47
  - is it linked? So the revenue cycle has
• 10:49
  - started. We're off to the races. The denials
• 10:53
  - primarily are generated upstream
• 10:55
  - in those sort of front patient facing processes
• 10:58
  - through to the provider documentation.
• 11:00
  - And so
• 11:02
  - figuring out how to collaborate
• 11:04
  - cross functionally
• 11:05
  - is critical for the success of of any
• 11:09
  - revenue cycle operation and and is especially true
• 11:12
  - when we think about denials. There are so
• 11:14
  - many opportunities
• 11:15
  - throughout that continuum
• 11:17
  - where we can drop the ball. There can
• 11:19
  - be a gap, and it results in a
• 11:21
  - denial that we end up spending more dollars
• 11:23
  - on the back end to collect.
• 11:26
  - Let's add another another variable into the mix
• 11:28
  - here. We're talking about denial
• 11:30
  - management. We're talking about the ownership of this
• 11:32
  - work, and I think that I love that
• 11:34
  - revenue cycle might own the scorecard, but there's
• 11:36
  - so much other
• 11:38
  - efforts and oversight and hands that touch this
• 11:41
  - this ball.
• 11:43
  - Let's talk about budget because for a health
• 11:45
  - system facing significant denial volume or margin pressure,
• 11:48
  - an operational redesign like this and AI investment
• 11:51
  - are both competing for some of the same
• 11:53
  - dollars.
• 11:54
  - How should leaders think about balancing and sequencing
• 11:57
  - those two things? What have you seen separate
• 11:59
  - the organizations that are consistently improving financial performance?
• 12:04
  - Absolutely. And and, you know, I appreciate the
• 12:06
  - way you asked the question because I do
• 12:08
  - think it's a balance. It's it's not an
• 12:10
  - either or question.
• 12:11
  - You know, when I think about this,
• 12:13
  - my recommendation would be, you know, first start
• 12:15
  - with identifying where the value is leaking. So
• 12:18
  - where's the leakage? Where's the gap? Where's the
• 12:20
  - breakdown?
• 12:21
  - Fix what you can fix operationally first, going
• 12:24
  - back to our earlier discussion that AI isn't
• 12:27
  - gonna solve bad processes.
• 12:29
  - Fix what you can fix operationally,
• 12:31
  - and then identify technology that you can bring
• 12:34
  - in to scale those now strong and standard
• 12:37
  - processes.
• 12:38
  - So when I think about organizations that are
• 12:40
  - successful,
• 12:41
  - they go to market
• 12:43
  - looking for technology,
• 12:44
  - understanding the problem they're trying to solve first.
• 12:48
  - Typically, they've got those really strong accountabilities.
• 12:51
  - They've got aligned metrics and and know what
• 12:54
  - they're measuring to.
• 12:56
  - They have standard workflows.
• 12:58
  - They're not
• 12:59
  - putting the technology in place to try to
• 13:02
  - size or scale the problem so that they
• 13:04
  - can then tackle it. So understanding what you're
• 13:07
  - solving for first, I think, is helps helps
• 13:10
  - clarify that investment proposition a little bit.
• 13:14
  - And then, Amy, my question for you to
• 13:16
  - build on, Chris, is, you know, fix what
• 13:18
  - you can operationally, then look for opportunities to
• 13:20
  - scale and standardize,
• 13:21
  - and in some cases, automate.
• 13:24
  - Looking ahead over the next three to five
• 13:26
  - years, what do you what parts of revenue
• 13:28
  - cycle do you expect to be genuinely automated?
• 13:31
  - And then where do you see still strong
• 13:34
  - need for operational leadership?
• 13:36
  - Sure. That high volume, low complexity
• 13:39
  - work is where I believe we have the
• 13:41
  - greatest opportunity
• 13:43
  - for broad automation.
• 13:46
  - So eligibility,
• 13:48
  - authorization,
• 13:49
  - claim status,
• 13:51
  - lot of denial follow-up.
• 13:53
  - But I believe leadership
• 13:55
  - is going to be more important than ever.
• 13:59
  - As transactional work
• 14:02
  - becomes more and more automated, leaders have to
• 14:05
  - step back
• 14:06
  - and ask bigger, more strategic questions.
• 14:10
  - Where do we apply intelligence,
• 14:13
  - and how do we continuously
• 14:15
  - improve
• 14:17
  - when so much is changing?
• 14:20
  - Payer behavior and policies are constantly changing,
• 14:25
  - and
• 14:26
  - that will
• 14:27
  - happen even more rapidly as payers aggressively
• 14:31
  - embed AI into their
• 14:34
  - processes.
• 14:35
  - Technology
• 14:37
  - is changing because
• 14:39
  - AI is continuously embedded, and then even their
• 14:42
  - own organizational,
• 14:45
  - goals are challenged.
• 14:46
  - So the best leaders,
• 14:48
  - I think, will
• 14:49
  - move from just managing people and process
• 14:53
  - to really focusing on
• 14:55
  - optimizing the revenue cycle
• 14:57
  - through
• 14:59
  - applying true intelligence
• 15:01
  - to find that
• 15:03
  - revenue cycle that truly
• 15:05
  - performs better over time. Mhmm.
• 15:08
  - Yeah. And I I just double down on
• 15:10
  - I I would agree. I think
• 15:12
  - looking ahead,
• 15:14
  - we will get pretty rapidly to a point
• 15:16
  - where all of the transactional
• 15:18
  - elements are are
• 15:20
  - pretty
• 15:21
  - automated.
• 15:22
  - So to Amy's point, leaders are going to,
• 15:26
  - be required to to shift their skill set
• 15:29
  - a little bit,
• 15:30
  - and I think it will be welcome. You
• 15:32
  - know, a lot of the leaders that I
• 15:33
  - interact with do spend most of their time
• 15:35
  - trying to put out the fires on, you
• 15:37
  - know, whether it's the the operational or or
• 15:39
  - the technology side. And so starting to
• 15:43
  - shift that focus
• 15:45
  - to things like stakeholder alignment. Right? AI isn't
• 15:47
  - gonna get your faculty providers on the same
• 15:50
  - page, you know, with the operational leaders who
• 15:53
  - are on the same page with the revenue
• 15:54
  - cycle folks.
• 15:55
  - Focusing on performance management. So now we've got
• 15:58
  - our valuable resources doing the the most high
• 16:01
  - value work. How do we double down on
• 16:03
  - performance management
• 16:05
  - and make sure that,
• 16:06
  - we're holding ourselves accountable to those really strong
• 16:10
  - outcomes?
• 16:11
  - The payer, you know, negotiations and liaisoning, the
• 16:14
  - regulatory compliance. Right? All of these things are
• 16:16
  - gonna be still are will still be very
• 16:19
  - critical,
• 16:20
  - to the overall operation of the revenue cycle,
• 16:23
  - and I think leaders will have much more
• 16:25
  - time to focus on those and be more
• 16:27
  - strategic about how they move their revenue cycle
• 16:29
  - forward.
• 16:30
  - This is an exciting time for leaders in
• 16:31
  - those space. I mean, just hearing everything you
• 16:33
  - just named there, in addition to what Amy
• 16:35
  - had shared, and then I think about things
• 16:37
  - like talent development and Right. Maybe some of
• 16:38
  - these big cornerstones of leadership that because of
• 16:41
  - the firefighting
• 16:42
  - and the day to day transactional work that
• 16:44
  - was not unimportant,
• 16:46
  - but there wasn't other means to necessarily execute
• 16:48
  - it the way there is now.
• 16:50
  - This is an exciting time for those leaders
• 16:52
  - to be evolving in that space. It is.
• 16:55
  - Before we wind down here, Chris and Amy,
• 16:57
  - is there anything I haven't asked you or
• 16:59
  - or that you wanna touch on or any
• 17:00
  - final takeaways you like to leave our listeners
• 17:02
  - with?
• 17:03
  - Sure.
• 17:04
  - I think
• 17:06
  - we have an exciting opportunity
• 17:08
  - to use AI to rethink
• 17:12
  - what the revenue cycle can look like. And
• 17:14
  - I've talked about building that intelligence layer into
• 17:18
  - the revenue cycle operations,
• 17:21
  - what that allows us to do is
• 17:24
  - uniquely design an optimal path for a given
• 17:27
  - process with AI and then measure team performance
• 17:31
  - against metrics for that optimal scenario. And I
• 17:33
  - talked about that a couple of times
• 17:35
  - because the goal isn't just to automate more
• 17:39
  - work and that's a, you know, that's a
• 17:41
  - problem we can fall into and and Chris
• 17:43
  - has talked about that a couple of times.
• 17:47
  - It's really about building a revenue cycle that
• 17:50
  - learns from what's happening,
• 17:52
  - that adapts and gets better over time because
• 17:55
  - that's ultimately
• 17:57
  - what we want to achieve, and that's what
• 17:59
  - makes
• 18:00
  - being in revenue cycle right now very exciting.
• 18:04
  - So that's where Janus is focused for the
• 18:06
  - future,
• 18:07
  - and we're seeing great results.
• 18:10
  - Chris?
• 18:11
  - Yeah. You know, as I I think about
• 18:13
  - how rapidly things have evolved just in the
• 18:15
  - last two to three years,
• 18:17
  - you know, I think everyone's gonna have access
• 18:19
  - to the same AI
• 18:21
  - tools
• 18:22
  - and and sort of thinking about how do
• 18:24
  - you differentiate
• 18:25
  - the the winners from the the laggards, if
• 18:27
  - you will.
• 18:28
  - And I think the winners are gonna be
• 18:30
  - the ones that are thoughtful about how they
• 18:33
  - bring AI and automation
• 18:35
  - into play,
• 18:36
  - really weaving in what we talked about around
• 18:39
  - using it to strengthen the operation, but not
• 18:42
  - using it as a a one size fits
• 18:44
  - all solution.
• 18:45
  - Right? And so not who automates the most,
• 18:48
  - but who has the operational discipline combined with
• 18:51
  - the technology
• 18:52
  - to make it all work most effectively.
• 18:55
  - That orchestration that I think one of you
• 18:57
  - had mentioned earlier in the conversations will be
• 18:59
  - coming from more critical.
• 19:01
  - I wanna thank you, Chris, and thank you,
• 19:03
  - Amy, for your time and your insights today.
• 19:05
  - We also want to thank our podcast sponsor,
• 19:07
  - ECG Management Consultants.
• 19:09
  - You can tune in to more podcasts from
• 19:11
  - Becker's Healthcare by visiting our podcast page at
• 19:13
  - beckershospitalreview.com.