Building Trustworthy AI in Healthcare Through Transparency
12 minAudio
In this episode, Halsey Wise, CEO, Lyric, discusses the importance of trust, truth and transparency in healthcare AI, including decision lineage, governed AI and keeping humans in the loop. He also explores how AI can reduce administrative waste, improve decision-making and strengthen trust among payers, providers and patients. This episode is sponsored by Lyric.
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
- Hi, everyone. This is Brian Zimmerman with Becker's - Healthcare. Thank you so much for tuning into - the Becker's Healthcare podcast series. I'm excited to - welcome Halsey Wise to the podcast. - Halsey is the CEO of Lyric, a role - he stepped into after serving on the company's - board since 2022. - Over nearly three decades, he's led six companies - as a CEO or president and served on - corporate boards across health tech software and financial - services, - including board seats at Cerner and Cotivity. - At Lyric, he's focused on scaling the company's
- investments in AI and health care decision intelligence - to bring greater trust, truth, and transparency to - the broader health care ecosystem. - Halsey, welcome to the podcast. - Hey, Brian. Thanks for having me. Appreciate being - here. Great to have you. So you've said - health care cannot improve without trust, truth, and - transparency as I as I noted in our - the the intro there. - After thirty years leading companies across the provider - EHR and payer sectors, - what convinced you that transparency is an operating - requirement rather than a nice to have?
- Yeah. Good question. I think transparency may be, - mean something different to a lot of folks. - But for us, definitely, transparency is an operating - requirement. And after thirty years of being in - health care - as a human being, as a patient, and - and being in the provider world now helping - health plans and seeing a a wide aperture - in health care, I think a lot of - health care decisions - depend on logic that others need to understand - and defend. - And across the whole value chain, decisions about - eligibility, - coverage, payment, care, risk, other things like that
- require a a very clear line - back to the underlying - policy data and reasoning. - I call that the decision lineage. The decision - lineage has to be, - available and trusted. And we're in a regulated - industry, and health care, as you know, is - a lattice work of rules, regulations, - complexity, - and there's a high cost of failure in - our industry. So when a decision - affects a claim, - a provider, - or a member experience, or a patient,
- there's a super high cost of failure here. - And regulators, - clients, partners, and providers, and all expect that - they have a logic that's dependable and reliable - that they can follow it. So transparency is - not a feature layered on top of AI. - It's a prerequisite condition - for the adoption of AI. - And - at Lyric, we believe every recommendation needs to - be, - traceable, - auditable, - and attached to the policy data - and the source data from which it comes. - And that gives organizations not only transparency,
- but the confidence on which to act upon - it. Yeah. The the decision lineage as you - as you put it, you've gotta be able - to to trace that back. And and can - we go even further here, which is you - really I guess, in practical terms, what separates - an AI system that is able to do - that, that that clinicians and executives can trust - from one that simply produces a very confident - answer? - And and why does that distinction become even - more important as AI advances? And it's gonna - continue to move closer to clinical and and - operational and financial decisions as as it becomes
- more - more advanced. Why why does this distinction really - matter between an AI system - that folks can trust and one that, you - know, just looks nice? - Yeah. And that's something that's worth spending a - little time on. - Confidence is not the same as trust. You - know, a confidence answer and a trustworthy answer - can look identical or similar from the start, - but that's part of the problem. The difference - is really whether you can open a recommendation - and see the policy data and logic behind - it. You know, I would I would posit
- that a confident sounding answer that turns out - to be wrong - is more dangerous - than having some uncertainty acknowledged at the beginning. - And so, you know, the health plan clients - that we work with and others across the - ecosystem, - trustworthiness, - you know, needs governance. - And so, you know, at Lyric, we call - it governed AI. It operates in a defined - architecture that combines - policy data - that deterministic - rules, a workflow, - intelligence, and clinical and payment data that's trusted
- and proven at the source. So it's proven, - verifiable, - and trusted. - And so AI is not generating - an answer in any situation in isolation. - It's really structuring evidence and servicing the highest - value action - for, - a human in the loop to review, and - we call it augmenting human decisions. We don't - make any decisions on our own. We have - health care decision intelligence that pushes information - to our, - health plan clients for them to make the - most informed decision at their given their circumstances.
- That distinction really moves AI closer to decisions - that affect payment and patient care, and that's - super important because, - you know, health care is now almost 20% - of GDP or $5,000,000,000,000 - and 20% of that's waste. Every dollar that - waste is dollars that have a big opportunity - cost that could be redirected back to patient - care, and that's what we're trying to be - a part of. So a confident answer that - can't be traced is a liability. - A trustworthy answer is something a health plan - can rely on, act on, defend
- to a patient, provider, or a regulator. - Yeah. - The and to your point too earlier, the - stakes here are high. Right? The stakes are - high in clinical care, but the stakes are - high financially too. Right? This is a these - are these big swings of money moving, and - those have implications for direct patient care and - the health of these organizations. - Correct? - A 100%. - And, you know, it just a lot of - times health care has been opaque in the - past. - Patients, - you know, you, - our family, my family, no one accepts a - black box answer. Physicians don't accept a black
- box answer. - Health plans don't accept black box answers. You - have to have the traceability, the auditability, - the grounding in the context of the decision - so people can understand - where the model or AI - source the data that is used to augment - human decisioning. - Yeah. Yeah. I wanna get to something too - that's been - sort of a a a truth in the - industry for for a long time, which is - that - providers, payers, even technology companies as well are
- really trying to attack many of the same - problems. - But but the challenges there is often the - incentives don't align, and and, of course, that - will create an environment of limited trust. - Where can AI help, I guess, here, sort - of help solve this perennial challenge in the - industry and create greater transparency between those groups? - And and I guess also the flip side - of that coin, where could it do the - opposite? Where could it if if leveraged inappropriately - perhaps, where could it have deepened that this - challenge worsen it? - Yeah. Brian, AI can make decisions easier to
- understand. - The biggest opportunity some of the biggest opportunity - we see in workflows where providers, payers, and - other firms already interact, but they don't see - each other's reasoning or logic. - So payment and coverage decisions are maybe one - of the clearest examples. - You have AI showing the policy data and - logic behind a recommendation. - A provider then has something to respond to - instead of a black box denial. So we're - very overt. And, again, that traceability and the - logic and that I've used the word decisioning
- decision lineage. - So it's - all transparent, available. There's sunlight on it. There's - no attempt to, - to hide or have any opacity in in - the process. It's all, you know, fully straightforward. - But an AI makes it helps, you know, - pick up pattern recognition, - anomaly detection, - and signals - that are otherwise - harder to - determine. But we - use that again to augment human decision making - and never making a decision only in an
- autonomous manner. We just simply don't do that. - But the speed that AI - delivers - is really not as helpful without the true - governance because without the governance, the mistrust exists. - So, - AI has to be moved faster, but it - has to be explainable - and at scale. And so a human in - the loop design is critical, - and technology should inform judgment, not replace it. - And the people that are affected by a - decision, you, me, - all of us as patients need to understand - why a particular decision was made, and that's
- what we do on behalf of, our health - plan clients and providers. - Yeah. And - showing how that decision was made is Right. - Is what is what can bring back that - trust and perhaps even do you do you - believe it's possible align incentives, - even more so? - Yeah. I mean, I think I think to - have it everyone wants to reduce waste in - the system. I don't think there's any entity - that that, you know, is pleased about the - amount of waste. And so I think everyone's - aligned to try to get the waste out - of the system and redirect it back at
- patient care. - And and I think AI can be a - tool in that. And I think that the - there's such a significant amount of money that - could be re redirected to patient care given - the size, scale, and complexity of the problem. - Excellent. Well, closing here by maybe - reasserting some of the stakes involved in in - this transformation. So what is at stake if - health care gets trust and transparency and AI - wrong? - And then, - again, another flip side question, what could getting - it right make possible that's never been possible
- in health care? - Yeah. I mean, there's a tremendous amount at - stake, you know. Right? Health care, as I - say, is an enormous - market. It's 20% of GDP and 25% - of that's waste. It's incredibly important - to get it right. If it gets it - wrong, - AI adoption stalls, regulators, and providers, and patients - don't accept decisions that they can't question - or they can't audit and have traceable data. - And worse, really, the industry risk entrenching the - estimated 300,000,000,000 in annual administrative waste that we're
- already fighting. - And so AI can help us - address that, but if AI gets it wrong, - that deadweight loss in the system and the - lack of care is certainly a penalty that - our society shouldn't be willing to accept. - If health care really gets AI AI right, - we can improve the speed of decisions and - the accuracy of decisions at scale. - Decisions that once took days to review - and appeal can happen correctly on the first - time. We at Lyric are processing - over 10,000,000,000
- annual claims a year, and we're doing it - in milliseconds. - So we provide deterministic - information that's traceable, auditable, and grounded in context. - And so AI can really provide the efficiency. - It's a foundation for an industry that spends - more resources - on care and and fewer litigating its own - decisions of the past. It's a super important - tool. It can help propel - degrees of freedom in patient care that were - never possible before. - Yeah. And and and to your point, getting
- it wrong means a lot of that transformation, - the positive transformation stalls out. Right? - 100%. - 100%. And, - we use AI aggressively and confidently at the - request of our health plan clients, but we - also - use it, you know, as I said, in - a very responsible, ethical manner, - very focused on, again, providing transparency and traceability - so that there's this concept of trust, truth, - and transparency where we started is is earned
- every day. - Well, Halsey, I think that's a perfect place - for us to end. Thank you so much - for coming on the podcast. - Hey, Brian. Thanks for having me. Appreciate it. - We also wanna thank our podcast sponsor, Lyric. - You can tune in to more podcasts from - Becker's Healthcare by visiting our podcast page at - beckershospitalreview.com.
In this episode, Jon Grandstaff, Senior Vice President of Payment Integrity at Zelis, discusses how payers can reduce payment friction, use AI to support more accurate claims reviews and improve payer-provider...
In this episode, Amy Conlon, MPH, Population Health Sr. Manager at Exact Sciences, and Jeff Huebner, MD, FAAFP, Executive Medical Officer, Division of Medicaid Services, Wisconsin Department of Health Services, discuss...
In this episode, Denis Mrkva, Senior Vice President of Analytics Solutions at Data Axle, discusses how health plans can combine consumer, employer, geographic and healthcare data to identify growth opportunities, reduce...
In this episode, Bob Tavernier, Solutions Executive at Quest Analytics®, discusses how health plans can leverage AI, provider data and strategic partnerships to improve network accuracy, strengthen regulatory compliance...
In this episode, Martina Lee Strickland, Head of Marketing, PR and Brand at Clever Care Health Plan, joins the podcast to discuss the importance of building trust when serving underserved communities. She also shares...