Fixing Healthcare’s Front Door w/ Dr. Ashish Mandavia

The digital front door to healthcare is jammed, and it’s costing patients, providers, and payers alike.

In this episode of CareTalk Executive Features, host David Williams talks with Dr. Ashish Mandavia, CEO and cofounder of Sohar Health, about how AI and automation can transform eligibility and benefits verification from a frustrating bottleneck into a seamless, real-time process.


🎙️⚕️ABOUT DR. ASHISH MANDAVIA
Dr. Ashish Mandavia is the Co-Founder and CEO of Sohar Health, an AI-powered platform transforming front-end revenue cycle management (RCM) for digital health providers. In his role, Dr. Mandavia is responsible for setting the company’s strategic vision, overseeing execution across products, operations, and partnerships, and ensuring Sohar’s solutions address critical gaps in healthcare administration. With a strong background as a practicing psychiatrist and health tech executive, he leads Sohar’s mission to simplify the insurance verification process, an often-overlooked source of friction for patients and providers.

Under his leadership, Sohar Health has developed a proprietary API-based system that delivers real-time insurance eligibility and benefits data with 99% accuracy and sub-30-second processing speeds. This makes the patient intake process almost instant, when it used to take hours or days to complete manually. Dr. Mandavia is deeply involved in client success, innovation, and thought leadership, frequently representing the company at industry summits and guiding its efforts to support digital-first care delivery models. His responsibilities also include talent development and investor engagement, particularly as Sohar scales its partnerships with major behavioral health providers like ZocDoc, Talkiatry, Rula, Cerebral, and UpLift.

Dr. Mandavia’s vision centers on building a smarter, more transparent “digital front door” to healthcare, starting with benefits verification. He believes the patient journey begins with financial clarity and uses AI to eliminate administrative burdens so providers can focus on delivering quality care. Through Sohar, he is pioneering how automation can improve access, reduce denials, and enable more sustainable behavioral health operations.


🎙️⚕️ABOUT CARETALK
CareTalk is a weekly podcast that provides an incisive, no B.S. view of the US healthcare industry. Join co-hosts John Driscoll (President U.S. Healthcare and EVP, Walgreens Boots Alliance) and David Williams (President, Health Business Group) as they debate the latest in US healthcare news, business and policy.

GET IN TOUCH
Follow CareTalk on LinkedIn
Become a CareTalk sponsor
Guest appearance requests
Visit us on the web
Subscribe to the CareTalk Newsletter

Transcript

David: 0:00

Healthcare's digital front door is jammed. Before patients ever see a clinician, they have to wrestle with insurance verification, which is a slow and confusing process that can delay care. But what if eligibility and benefits could be verified instantly in real time? Welcome to Care Talk Executive features a series where we spotlight innovative companies and leaders working to advance the healthcare field. I'm David Williams, president of Health Business Group. My guest today is Dr. Ashish Mandavia, CEO, and co-founder of Sohar Health, which is using AI and automation to make the front end of the revenue cycle as seamless as the rest of digital health. Ashish, welcome to Care Talk. For having me, David? Well, I have to ask, you know, how did your background as a psychiatrist shape your view of administrative barriers to care delivery?

Dr. Mandavia: 0:55

It's a, it's a bit of a journey, so, uh, I appreciate the question. It's often the first one that I get asked. Um, well, I mean, as a psychiatrist in, in the, the European healthcare system, I think, uh, I'd say I very much saw firsthand the power of kind of simple, transparent access to care, no surprises, no barriers. If a patient needed to come in to get some support. They could. Um, coming to the US I was very much struck by how administrative complexity can kind of delay care, frustrate patients and providers, hearing peers and colleagues talk about, uh, you know, endless hours maybe in the ER setting or urgent care setting, kind of having to be on phone to insurer, uh, to kind of ensure that they can give. Good quality care to their patients, the care that their patients deserve. Um, yeah, it gave me a sense of real kind of, uh, um, shock firstly. But then when you get over that, it kind of seems like, okay, this is not so much a clinical issue. This is more of an infrastructure issue. Uh, and so I've seen clinicians burn out. Patients turned away. Or because navigating insurance complexity is a bit of a maze. Um, and I think those experiences of being able to empathize with patients who are really trying to get the care that they need, uh, and are desperate for that, um, and providers who are ultimately wanting to do what they were trained to do, which is deliver good quality care, um, has shaped my belief about trying to fix these kind of invisible bottlenecks that exist, uh, to help improve access and, and, and trust in care.

David: 2:23

Well, welcome to America. You know, it's, uh, a rude, a rude, a rude awakening. We have a family from Canada and that's, even though it's just across the border, sort of equally shocking. Meanwhile, we, you know, we throw, uh, caste dispersions on those, on those healthcare systems. Alright, so when you, when you, when you started to discover that you know this, there's a lot of different inefficiencies in a lot of administrative ones that we have in American healthcare. So why, you know, the focus on eligibility and benefits verification specifically.

Dr. Mandavia: 2:51

Yeah. Uh, that's a, that's a good question. It, it seemed like a very natural place to start. Um, if you were to think about revenue cycle management as a spectrum or, or a kind of cycle split between front end, mid, and, uh, backend. Um, where my co-founder and I, uh. Uh, my co-founder Lisa and I, were at a, a company called Pego. They used to be known as Quick Genius. I think you may have spoken to our CEO uh, just a few weeks ago. Um, we, we kind of saw firsthand, um, as we kind of made the move over from Europe to US of how, um. You could go and get credentialed with, uh, uh, or contracted in with insurer, but then when you kind of come across the hurdle of having to, uh, handle RCM, you know, it would, you would endlessly get. Wrong data, um, have to be on the phone to payers and ultimately get a lot of delays to your cashflow. Uh, and so we realized that if you, if you don't get eligibility and, and verification of benefits networks status determination right at the front end, you're really setting yourself up for denials later on. And so it's that invisible pain point that, that, that affects every provider, but. Hasn't really got a more kind of modern, slick digital health, I guess, approach towards it. Uh, and thus, you know, it's, it's, it's easy to get wrong and people have kind of lost faith in the credibility of the data that they, they, they try and get back, uh, um, in that sort of space. I think also the patients themselves kind of are, are a little bit, uh, wary of the, the messaging or information that they get, which often leads to 'em either not trying to access care further or, um, or just moving on to another provider organization who's a bit more confident with, uh, with whether or not they're gonna be seen and, and, and, and, and their insurers gonna pay for their care or.

David: 4:37

Got it. So you mentioned RCM. So for those that aren't, you know, a living and breathing RCM, which unfortunately does not stand for Radio City Music Hall or something like that, but revenue cycle management, what, what's broken in insurance verification today? I mean, I, I think about it sort of very simplistically as I have my card and somebody wants to know if I have insurance, I can just verify it by showing them my card or, or the digital equivalent. So what's, what's broken?

Dr. Mandavia: 5:05

Yeah, I mean, it, it's mainly broken because, uh, because the data itself is, is, is fragmented. It's located in many different places and, uh, there's uh, a lot of misalignment of incentives. So payers hold that data ultimately, and providers are chasing it. And so. Um, the systems that connect between both the providers and the payers were built decades ago. I mean, it, it really was the evolution or the natural step from moving away from paper vaccine was to do kind of archaic data transfer, which is sort of like. Digital vaccine. And, um, that means that you get, uh, data points that are, that are kind of inconsistent, incomplete, um, which ultimately mean that you get inaccurate data being pulled back from the payers. Um, and AI is finally kind of giving us a way to clean up, um, validate and interpret that data in real time. Uh, which is something that kind of the legacy sort of technologies that sit between, um, the two kind of sides of the coin, the providers and the payers, um, really kind of, uh, struggle with. Um, so ultimately I think it really does come down to that, that that misa misalignment of incentives that ultimately insurers are. Underwriting risk. They're not so much technology companies, they, it's not their number one priority to surface this sort of data in a timely, fashionable way. And providers, um, want to try and get patients through the door so they want it, uh, delivered in a fast way, and it's frustrating to them when. They have to go onto a portal for an insurer or pick up a phone call and call the insurer. So, um, so I think that's certainly where the, the, the kind of biggest challenges, uh, lie. And like I think I alluded to earlier, if you don't get it right at the beginning, it's very difficult to correct later down the line. So it's all well and good getting to delivering care, giving a great patient experience there and then submitting the claim. But if the payer comes back and turns around and said, well, actually you wanna network. With that patient's plan, then there's not really too much that, that that provider can do. They can either write off that claim, they could go back to the payer and try and argue their case or ultimately, and all too often what happens is that they end up landing the patient with a surprise bill. Um, and that's not a great experience. Ultimately, the patient may end up in, in, in, in debt, or if they can pay off their medical, their, their, their, their medical bills, then they're probably gonna be a lot more wary and mistrusting the next time they come along to try and access care. Um, and that's a vicious cycle that, that, that affects the patients at the core.

David: 7:37

So it sounds like, you know, the in incentives are out of alignment, which is, uh, fairly typical. I think there's also, you may be alluding to some technology issues because these payers are, you know, these are big kind of insurance companies and they literally use. Mainframe computers with some of these databases were just not designed for. Someone shows up online and instantly wants an appointment. So I think that, you know, you're probably leveraging things like APIs and AI in order to actually go from it and make it look like, Hey, these systems were actually set up for this kind of use and differently, I think from a legacy clearinghouse system. So, so what's the Sohar approach compared with, you know, how others have tried to tackle this problem?

Dr. Mandavia: 8:17

Yeah, so our approach has always been to take, uh, an API first approach. And, and, uh, for anyone listening who's not entirely sure what that means, it's, it's a form of, uh, interface or exchange of data that is, uh, particularly robust. It's quite simple, uh, to, to integrate, but it means that a lot of like. Various different transactions can be happening during that API call being made. And then once the, uh, an an organization or a technology platform like Soha has got that information from the insurers, then it can be called upon in order to then surface up to the patient and. A key component to that is actually can we do that in a timeframe that makes sense as a patient is going through an intake flow, um, as you alluded to there, coming onto a website and actually trying to access care. Then there and then patients now are so used to in a kind of retail environment. Uh, choosing the items that they want to add to their cart, then getting to the checkout and instantaneously being able to make a payment. Um, it's so different than that in healthcare. It's endless questions to go through, endless information to give, and then typically you'll met with a screen at the end of it to say, okay, we'll get back to you in 24 hours or in 48 hours to let you know if we can, we can take this on board. So what what Sohar really aims to do is, is um, do a lot of that kind of, uh, those processes. Very quickly under the hood and then surface that up, uh, for the provider to then display to their patient. It's very much trying to be that invisible infrastructure layer, um, that sits between the provider, the clearing houses, and the, and the, uh, the payers. And that allows us to, to really work with. Many different vendors in the space, not just being beholden to the insurers themselves. So we go where the data is and we really build a huge amount of, uh, robust resilience under the hood there, uh, in order to find that data if we can't find it from the kind of regular sources.

David: 10:14

Got it. So I love your marketing claim, which is 99% accuracy and verification under 30 seconds. So I love the claim. Yep. Is that actually doable? You described a lot of complexity. Uh, is that an aspiration or is that you achieving that?

Dr. Mandavia: 10:30

We're achieving that. So, so when it comes to determining if someone's eligible or not, we're achieving 99%. Um, it's, it's hard work. It's a lot of monitoring and scraping of information, but it's, uh, but, but it's achievable. Um. The harder part is actually making sure that the numbers that return that come back from the insurer, so the actual benefits, information, copay, co-insurance, deductibles, whether that's individual or family deductibles, are as accurate as they possibly can be. And the reason for that is that this. There's so many different numbers that are surfaced up from the insurer. Like if you are going for a, uh, a telehealth, uh, psychiatry appointment or a psychotherapy appointment, there may be different benefits associated with that. Well, that would be different from going for a dermatology appointment, and so making sure that you can pull those numbers accurately, that's where the challenge comes in and that's often where solutions kind of very much fall short. They're really designed to try. And just, uh, focus on general medical benefits and not so much honing into specialty care or into areas such as ophthalm, uh, ophthalmology or, or, or, or dentistry. So, um, this is kind of the, the sohar difference. We do apply a lot of that technology that you're referring to earlier, the use of AI machine learning to be able to kind of pick that data out, uh, accurately. And when it comes to surfacing benefits, we can accurately determine that we, we can be. Between 93 to 95%, um, accurate with surfacing those benefits, um, on the speed front? Well, that's a really interesting one. Uh, our average speed for running one of these checks typically sits at around seven seconds. So it's not quite stripe, milliseconds, uh, you know, uh, the, the ultimate kind of consumer-like experience, but it's getting pretty close and that's way better than. 20 minutes on hold with a, with a, with a front end, uh, operations team member who has to phone up the payer or, you know, waiting that 48 hour period to hear back from the, uh, the provider organization while someone in their back office is making the call or compiling all the information data. So, um, yeah, I very much kind of stand by those marketing claims. It's, uh. It's important that we can stand by them and we validate them in, in quite, um, quite unique ways. So we, we look at, we analyze not only the eligibility information for our provider organizations, but we also analyze their claims data. We don't do much with their claims data. We don't submit their claims for them. That's, that's something that they already have figured out and are very efficient at doing. But what the claims data really unlocks for us is telling us whether or not that data we surfaced up during the intake flow. Matched what was on record, uh, and on file with the insurer. And if you complete that feedback loop over and over again for a certain plan, certain pair in a certain setting, then you can get to really high levels of accuracy. And that's exactly the, the secret source that that drives so hard. Great.

David: 13:29

Excellent. And you can disclose what it is 'cause it's hard to do. So, uh, so that makes sense. So, so I think it's, maybe it's self-evident from the discussion we've already had, but maybe you can make it tangible about how having improvements like this could translate into better outcomes, you know, for providers or for patients. I understand Less friction, less burnout, you know, quicker time to care. Can you make it tangible what that might mean for you, a prototypical patient or, or in some other way?

Dr. Mandavia: 13:57

Time and time again. What I can see, what we can see on our side is that, um, a patient will access a provider, and when I say access, I mean they will go onto their website or give them a call and try and book an appointment. And if they can't get a response within. 20 to 30 seconds. They're already Googling. The next, um, the next provider organization. And here I'm really referring to the, the kind of direct to consumer, uh, digital health providers that will advertise on television or Facebook or Instagram. And you can really see it happening. They will try with one, uh, one provider organization. They won't be able to get through, or it's taking them too long. So they'll go to the next one and then they'll go to the next one. So in terms of, uh, tangible difference for the patient, the sooner that they can get a response and an answer, the more likely they are to one, put their credit card on file, uh, or pay upfront for, for, for a consultation or, or, or a provider visit. Um, and two, they're, they're much more likely to then attend that initial appointment that they book. How does that translate on the other side, like on the provider side? Well, it's very, very tangible. The business case really, uh, really, really sits there. So from an operations perspective, it's. 1000 hours of manual work hours per, uh, per per week being saved. It's 90% of your operational costs associated with your front end, uh, um, administrative staff, uh, potentially being reduced or redirected in, into another, uh, another area of that revenue cycle, uh, continuum. Um, it's. Increase of conversion. And ultimately what we are seeing with our customers that we work with those direct to consumer digital first, uh, providers such as the likes of psychiatry ruler, they are seeing upticks of 20% in patient conversion. Um, and their definition of conversion is does the patient actually make it an attend that first uh, appointment? That's what matters to them. Um, and so. The, ultimately what I feel is that we are broadening access. We're creating a much more enjoyable, um, frictionless sort of patient experience, but ultimately driving top line revenue generation for provider organizations. Making their cashflow worries also, uh, a a little bit easier in some, in some cases we've seen that because of using our network status feature, so determination of net, uh, network status based on the, uh, the patient's plan, uh, providers have completely zeroed down any claims denials associated with, with being out of network. Um, and those are particularly challenging to be able to kind of, uh, combat with the, with the insurer.

David: 16:39

Well, it's a very good value proposition because you're not just talking about, you know, saving a few cents on a transaction, but if it's done right, it's the difference between having that patient in the first place and then even if you do accept them about whether you're actually gonna get paid or not. So it's, uh, it's a good position to be in. You mentioned a couple of well-known, um. Names. I was gonna actually ask you if, uh, if you can name, uh, customers. I love the name Tory, by the way. That's an, that's an awesome name. Good. Any, any other, any other cool ones you can name or, or other ones that, bland

Dr. Mandavia: 17:08

names. So the organizations that we work with, uh, include Tory Ruler, um, better Help, uh, cerebral, uh, and two chairs. I mean, you could see a trend developing there. They, they're, they're all direct to consumer. They, they're all typically in the behavioral health space and that's where we've certainly, uh, carved out a bit of a niche. Um, but beyond that, we've, uh, we've, we've started working with, uh, with larger group practices who, who deliver kind of, um, uh, care physically, um, and are basically powering a lot of the, uh, the tooling that they already have and integrating directly into their practice management software or electronic health record so that their existing teams who are on the ground picking up the phone or, or seeing patients come through the door, can, can, can use this technology to make the experience better for a patient.

David: 17:57

Let's talk about why you started in behavioral health. I mean, there's, you know, your clinical background would obviously lead you there. Yeah. But anything particular about behavioral health as a starting point?

Dr. Mandavia: 18:08

It's really hard as a, as, as a, as a space goes. Um, a lot of the rules that apply to, uh, other, uh, medical specialties really don't apply in the case of behavioral health. Uh, and on top of that, um, because the, um, the supply of providers within the behavioral health space is very limited. We see a lot of, um, uh, carve out payers existing or carve out plans existing where essentially large provider networks will, uh, offer their services to various different insurers. So then those insurers can say that they do cover behavioral healthcare. Ultimately, that creates a bit of a spider's web in terms of, um, okay, if I'm going to, to go to Blue Cross Blue Shield, California. Um, uh, to go and see a psychotherapist, then actually we probably need to be checking insurance benefits with carve out density like Carolon or Magellan, because they're likely gonna be the provider network who's delivering, um, delivering care. And when you have that sort of level of nuance, uh, in there, it becomes very difficult to ensure that benefits are accurate, that network status is accurate. So I think where we, where we really, um, kind of felt we could, we could develop like a stronghold. With t uh, going after a space that was very, very complex, um, but really needed help, uh, and, and therefore saw high volume of patients coming in, uh, a lot being turned away or churning because of, um, uh, uh, insurance related challenges. Um, and a lot of claims being written, uh, uh, being denied and therefore written off by those companies so they could really see the benefit of our technology quite quickly. Um, as we move on, of course, we, we, we look to expand into, in, into other areas. And when we've done it to this degree and this level of sort of, uh, depth in, in, in this area of behavioral health. The other ones kind of naturally, um, um, kind of layer on. Uh, and we've built our infrastructure and as such that it, that it can kind of move into other specialties. So now we also offer, uh, our services to primary care practices, uh, to weight management facilities. Particularly hot topic, uh, physical therapy, occupational therapy, and speech and language therapy.

David: 20:24

Got it. And are there differences in working in a different, uh, care setting? Yes. A physical space or hybrid compared to the fully digital?

Dr. Mandavia: 20:33

Yeah, exactly. And then, and then where you, what you see there is, is then like a, a bit of a. Blurring or, or, or change in kind of terminology as to what's eligible and what and, and what's, um, what's not. As soon as you get into physical settings, like treatment centers, for example, where someone may be, um, in, for a prolonged inpatient stay. Um. The definition of eligible is, okay, yes, does that patient have active insurance? And will it cover them for this service? But then it goes all the way into the depths of, uh, do we have prior authorization as well for, for delivering this care? Um, and that's why we, our process has lent itself really to the kind of outpatient setting. Um, high volume, high turnover of patient, sort of, um, coming through the, uh, coming through the front door, uh, settings. Um. The challenge, I guess, in digital health is that care can happen anywhere. It could be in the patient's home, it could be via telehealth. It could be that hybrid model that you referred to, um, and those. Encounters don't fit kind of, uh, neatly into that kind of legacy insurer logic, which is that either you're seen in hospital, a care setting of some sort, uh, or, or an outpatient or office visit. Um, and so that definitely brings in a layer of nuance. The other challenge is that with digital health companies, they have newer CPT codes that they bill on. Um, and those aren't particularly well mapped. Uh, so often kind of verification breaks, uh, breaks there. Um, so we certainly see there being subtle differences, uh, and like kind of wide differences, uh, uh, in, in different parts of the healthcare system. But ultimately where the most volume of these sorts of checks take place, it really does lend itself to speed, uh, and, and, and, and being done quickly and in a timely manner. Do

David: 22:23

you have a vision of what a fully transparent digital front door, uh, might look like? In healthcare.

Dr. Mandavia: 22:32

Absolutely. And I, that's like one of my favorite questions to, to answer. Well, imagine you, uh, you jump onto a website, you go onto an app, or you attend a, a clinic and you just simply. Tell them a bit about yourself. Your first name, last name, uh, how old you are and, and, and where you live. And they were immediately able to bring up, okay, here's your insurance information. Does this sound about right? Great. Well, this is, these are all the, uh, the, uh, uh, types of visits or types of care that you would be eligible for. And this is exactly how much it's gonna cost you. Would you like to proceed? Yes or no? If yes. Uh, you make your transactional payment, uh, and if no, then you, then you look elsewhere. But, uh, uh, what's that really sounding like? Well, it's almost sounding identical to a more consumer-like retail experience, right? It's like the Amazon application of, of healthcare access. And this is something I really strongly believe in. Why should it be so challenging? Why should it be so difficult to be able to work this one out?'cause ultimately, both sides, both the patient and the provider, want to be able to kind of come together to be able to deliver care. And this is really, uh, a friction point and bottleneck in that process.

David: 23:47

I, it's helpful. I mean, the, the Amazon analogy is a good one, and since everyone has that experience, they kind of know what's possible. So you say, look, I used to have to go out to the store and wait and put something on, right. You know, lay away or whatever. And now I have all these other ways to do it. You can't just tell me that healthcare's all gonna be, you know, the same. And you have some younger patients that never actually had that type of experience. They're like, what is this? Like, am I visiting right? Nevermind, like visiting, uh, the US from Europe, but just like on my different planet, it's not acceptable.

Dr. Mandavia: 24:19

Right. Right. Exactly. And I think particularly with some of the customers that we have who've particularly focus in on the, um, on, um, students delivering care, particularly mental health care for, for, for students, whether they're, they're at school level or at college level. Yeah, that's the experience that they want to give their, their customers. They, they, they know that their customers really have never had to kind of go and look up their insurance information. One of the big challenges actually we see with those customers is that, um, it's very hard to map their addresses because their address on the, with their insurance on file tends to be back home with their parents. Um, the address that they quote when they're trying to access care is their college address or, or like, uh, where they're living at the time. And, and, you know, subtle things like mapping that on to make sure that we can get the right data is, is, is really important. And a lot of what we do on the, uh, on, on the backend in, in, within, so hard. So. Yeah. Um, trying to bring about this new, new approach, uh, towards the digital front door. I mean, it just seems like a natural progression and also in keeping with, um, with, you know, movement towards pricing, transparency, ending surprise bills, um, uh, and, and preventing kind of medical debt. So, um, yeah, certainly, uh, a place that I really enjoy and, and am passionate about kind of, uh, leaning into.'cause that's where I really feel like we give value to, to, to, to the actual end user, the, the customer or the patient.

David: 25:44

I mean, I guess if I'm thinking about it from a patient outcome standpoint, and I'm, I'm dealing with the, the population, let's say of, uh, student level behavioral health. Even if I was no good as a clinician, uh, if people are coming for, you know, anxiety. Uh, stress oriented, stress related, uh, things. If you could just make it the process of dealing with the counselor less stressful, you're probably gonna help cure 'em that way.

Dr. Mandavia: 26:09

Big time and yeah, just imagine the scenario that, um, that you, you go through the process, maybe it is seamless, but maybe actually the data that came back was not correct. And in that time period between the you accessing, booking your appointment, attending your appointment, a claim being submitted, and a claim denial coming back, you could have actually in that three month period, seen that counselor maybe six times. So those, each of those six times you are potentially out of network. Um, and therefore you owe the money for each of those six successions. And maybe at the beginning your expectations were, were mismanaged to say you're just gonna pay $10 or $30 every time you come in. So suddenly a bill that you thought of around $180 over a three month period becomes more like $600 over a, over a three month period. And that in itself causes the endless amount of stress for, for patients. And imagine, and, and often my, some of our customers will say. Look, I just want someone to be able to tell my cust, uh, my, my patients that we've got them, you know, we, we, you know, you are okay from an insurance standpoint. Let's not think about that. Let's focus in on your care. Um, and, and often a lot of the time that they spend on the phones or, or having to counsel their, their customers is around their insurance or the bills that they're receiving as opposed to the actual care that they wanna deliver.

David: 27:40

So if I look at the RCM industry, they, they certainly have put an emphasis on front end operations. These issues about wanting to make sure that the provider's gonna get paid for a patient they're seeing and so on. I mean, these are standard bread and butter kind of issues, but what is the traditional part of the RCM industry get wrong about front end operations? Well, first they.

Dr. Mandavia: 28:03

They know it's a problem, but they tend to neglect it. They, they focus so much. And, and, and this is really a, a, a kind of a slightly top-down, um, challenge where the CFO or, or the head of finance within these provider organizations will be looking at the RCM team to, to kind of tell them, okay, what is our um, uh, claim denial rate? What is, what are the main causes for our claim denials? How fast are we, are we getting paid for any denials that take place? What, you know, how successful are we with the appeals process? As you can see, it kind of focuses purely on, um, on the backend of, of the revenue cycle management spectrum, and. Certainly an envo trend at the moment is focusing in a lot around kind of the administrative burden of coding and ascribing and coding of an encounter. And that certainly has a place, uh, and, and, and is why, you know, the big, big organizations in this space like Epic are, are, are producing features in order to, to kind of solve for that. Um, but if, if ultimately. The things that were done within the first few minutes of that patient encounter were done incorrectly. It's all well and good coding up really accurately or really fast. It's all well and good submitting that claim quickly, but it's still not gonna result in, in better collections. It's still gonna result in delay to cash flow. So I think that, um, where we see kind of 50% of claim denials being associated with front end RCM errors, I think ultimately RCM teams are really, um, really struggle to be able to. Do any about, uh, do anything about it often front end, uh, kind of like del uh, front end revenue cycle management sits with, with the operations team within, within a provider organization or if they're a, uh, digital co health company with a, with a product and engineering team that typically sits with marketing and a product team. People who don't really know. What revenue cycle management is or don't, don't really understand the importance of getting these numbers, uh, to be as accurate as they possibly can be. So I think that the fract, the, the, the kind of fractured nature or fragmented nature of the various different functions in the organization mean that. People know that front end RCM is a challenge, but dunno how to solve it. Um, and secondly, I, uh, I, I think that a big part of it is that a lot of the emphasis goes onto how much money are we making? Uh, and, and then it gets forgotten about, okay, how do we go about trying to improve things over time to stop the revenue leakage later on, or increase that, improve our cash flow. Um, and then my final view and thought on it is that, um. Front end revenue cycle management's important, but it tends to be something that's done by, in, in bulk. Uh, it's legacy systems that move in batches. It doesn't move in real time. And, and that's exactly like the, the, the space where our technology kind of bridges the gap, um, moving the, this kind of, these processes that you would expect to happen ticking over overnight if you're, if you're working on an active patient roster in a large healthcare system. Um, and like bringing that into real time and, and, and that's where the change really needs to take place.

David: 31:16

We talked about your use of automation and AI upfront to do this, you know, less than 30 seconds, I guess, you know, seven seconds. Uh, on average, uh, timeframe, if we broaden it a little bit and ask how can automation and AI reduce denials and improve financial transparency? Maybe you can expand on what you've already described.

Dr. Mandavia: 31:36

Yeah. Uh, well, absolutely. In terms of where automation and AI in revenue cycle management can, can improve denials. Well, it's all, all across the, the, the, the RCM spectrum and really with varying degrees of kind of impact, but. Overall, it's a positive thing for revenue cycle management. Um, because so many of the processes are manual, so many of the processes are repeatable and therefore so many of the processes are scalable through using ai. So we've already alluded to AI coding or AI scribing, and then AI coding up from an account. Then you've got tooling out there that allows you to submit claims, uh, really quickly monitor those claims, predict denials, uh, from certain payers that often deny claims associated with that sort of encounter, uh, or for that specific, uh, rendering provider, and are able to kind of preemptively prepare the appeal ready for it in, in our space where, where AI really helps us is. By completing that feedback loop between claims data and the information that we are getting back from the insurers. So the more data that we receive back, the more claims data that we review and analyze, the more accurate we can make the, um, uh, the, the kind of assessment or the, or the, the discovery process in the intake. Um, and that can get very personalized and granular down to that customer level, which is fantastic, but also. Every customer benefits from the, the economies of scale and the learning from all the others. So over time we can really crank up that, that accuracy rate from say, 75% as a baseline up to 95%. And those claims that I was kind of talking about earlier. Um, the other kind of like unlock that we've seen particularly over the last 12 months is the use of, um, large language models. So where you have these really large, and when I say large, you're scrolling and scrolling and scrolling through this, this machine readable data, uh, that gets sent back to us by, by the insurers and holds all the information in it. The in information is there. Being able to pick it out in a consistent and repeatable way has often been a machine learning problem and certainly is the foundation and core of what Sohar does. But for that added refinement, especially as every now and then, the insurer will say, actually, you know what, we don't want to take deductible if, uh, into account if it's a telehealth consultation or we do. Um, then. Adding on a layer of LLMs really, really helps us to fine tune the accuracy there. Uh, and again, using large, large amounts of data to be able to train those LLMs and keep them on good guardrails, uh, means that we can get very accurate, consistent

David: 34:21

results. So we've been talking a a lot about speed and accuracy, and now I wanna talk a little bit about simplicity and clarity, and I'd like to discuss it. I'd like to hear your thoughts about it in the context of, uh, benefits. You know, when I look at my. Benefits book is literally a book and say, you know, is this covered? Is that covered? It is very difficult to understand that. So how do you think about simplicity and clarity and their role in engendering trust? Um, and, and also, uh, offering access.

Dr. Mandavia: 34:54

You're not alone. I think a, a lot of folks over in the US um, really struggle to interpret what their benefits are, what they're specifically covered for. I think I've read a, a remarkable statistic, um, when we first embarked on this journey that. Only 4% of Americans, um, know the, the definition for copay, co-insurance and deductible. It doesn't surprise me. It's taken me ages to fi to figure out exactly what those things mean and how they interact with one another. And I've been in it, like deeply in it for such a long time. Um, so, so. A big part of this is trying to simplify and decant those huge, uh, summary of benefits that you get. And by the way, aren't there so many of them? I mean, I come from a system where in the private medical space here in the uk, there's only four insurers and about a dozen plans each. In the US you've got a thousand insurers and you've got unlimited number of plans under each insurer. Could have different types of benefits for different types of services. Um, the computations really make the mind bogle, which means that, uh, I'm not surprised that patients find it really difficult to struggle this and so to do the providers. So for us, simplicity and transparency is really at the core we take. All of that information, those, those, those, you know, 30 page long sort of documentations, and we need to count them down to the very specific information that's needed at the time and point of care. Um, and in doing so, as long as folks trust the numbers that we're surfacing up. We just make people's lives easier. Um, then they don't have to think about all the other things that, that are included in that pack. They don't have to determine, am I going for this type of behavioral care visit or am I going for this one? And why is it slightly different between those? Um, that's exactly where we, where kind of sohar looks to kind of bridge the gap, the layer needed on top of that. And that's not something that we offer actually. We try and educate our, our. Provider organizations to, to, to, to deliver this, uh, to, to their customers and, and to provide that ultimate kind of patient experiences to is to explain the numbers that they're looking at, where that good faith estimate has come from, what those numbers actually mean. And, um, and really like put their, put, put their money where their mouth is, so to speak with that. So. If we are saying to you that you are only gonna have to pay, you know, $30 for this consultation, um, well, and they're, they're comfortable with the numbers that they, they've surfaced up, uh, surfaced up there, then they should stand by that and they shouldn't then come round later down the line to then collect more, uh, more money. And we see often providers in that kind of paradox. Should we, should we go back, uh, to. To get more money from our patients when we've said one thing and, and the payers have said another. Um, and, uh, or, or, or, or should we just, uh, you know, even in the case of where a patient may have overpaid initially, should we refund them that difference? These are often kind of decision points that provide organizations are making, um, and ultimately result in a good or not so good patient experience.

David: 37:57

Well, I don't know how if the podcast hosts in healthcare would score any better on the than the copay coinsurance than deductible, but I'll say even if only 4% can define it. 94% don't like it, so Oh yeah. It's better to have a good understanding of it. Well, that's it for another episode of Care Talk Executive Features. My guest today has been to Dr. Ashish Mandavia. He's CEO and co-founder of Sohar Health. We've been discussing how to transform the front end of the digital health revenue cycle. I'm David Williams, president of Health Business Group. If you like what you heard, please subscribe on your favorite podcast platform. Ash, thanks so much for joining me today. Thanks for having me, David.