Expert On Board: A Fireside Chat with Karen Decaran-Voigt

Kaustubh: The very first question that I had was, of course, you know, around your journey.

Given that you've been in healthcare for so long, and you were VP at CommonSpirit and UnitedHealth, could you share a bit about your journey, particularly around how to organize or manage the strategy and network aspect of things?

Karen: As many people who know me, I grew up in rural Ohio, in an area where healthcare was scarce and, our healthcare was the emergency room at the hospital I was born in - preventive care didn't exist back then. By the time I graduated, I had seen enough people pass away or people not seeking care because they couldn't afford it because of pride, right? Fast forward, by the time I graduated, I was determined to study medicine. I started in medicine at Akron, and my boyfriend at the time had an accident - and lost his arm in that process.

Around the same time, I began working at Cigna Healthcare in an administrative role, handling authorization requests and interacting with case management nurses and medical directors. I got to know about healthcare delivery in a way that I hadn't ever even known existed because of where I grew up. With tuition reimbursement from Cigna, I switched from a clinical focus to business administration and IT, since Cigna would fund these but not clinical studies. I completed my undergrad from Baldwin Wallace, funded by Cigna, specializing in business administration with a certificate in IT.

My clinical exposure and new business skills led to a role in network development The people we were serving - they needed the right hospitals, the right doctors, etc. Plus my learning at Cigna allowed me to graduate to learn to negotiate contracts and find the right doctors and the right hospitals to serve, whether it was Medicare or commercial populations.

This attracted UnitedHealth Group, where I handled negotiations with large health systems and later moved to South Florida to develop Medicare programs across the country. I built about 21 of the markets that UnitedHealth Group is in [for Medicare]. I used all that learning to just help me evolve, um, to where I am today. But, it really started with having that scarcity at home.

Kaustubh: Yeah, that's quite a journey, you know, starting from a place with a little health care access to, uh, the onset opportunity and then the transition. I personally don't know if a single person has made, a switch in that short duration. So kudos for taking the call. One particular aspect of this was, uh, I think, uh, you know, that you were, uh, the chair of the value-based contracting committee and, you achieved a hundred per cent compliance rate.

You increased the membership under value-based contracts from, in fact, to a significant 42 per cent. So, uh, in these [like] mammoth tasks, what were the biggest challenges that you faced and how did you navigate your way through them?

Karen: The biggest challenge was that nothing was standardized. There wasn't a common language. We had to do two things, Firstly, within the contracts that the company used [to do value-based contracting] - we had to standardize and make sure that you didn't have to administer a different service fund model for every single type of entity we were contracting with [because you get no scalability].

So we learned to standardize the contracts. And then when you have standardized contracts, guess what? You also, on the other end, become efficient through being able to process what you contracted. Another challenge was aligning physician incentives. Previously, incentives were more volume-driven, with little focus on value or quality performance. We worked to get on the same page strategically, ensuring contracts were compliant and easily administrable. We shifted physician incentives to align with these value-based contracts, fostering a partnership mindset in healthcare.

"The biggest challenge was that nothing was standardized. There wasn't a common language. We had to do two things, firstly, within the contracts that the company used... Another challenge was aligning physician incentives... We shifted physician incentives to align with these value-based contracts, fostering a partnership mindset in healthcare."

We concentrated on Medicare Advantage and commercial populations, as Medicaid data from states was too complex at the time. This focus helped us migrate smoothly and ensure compliance, improving quality scores and risk adjustment codes.

A critical step was narrowing down our electronic health record systems from 57 to just Epic or Cerner. This was necessary for being part of the value-based networks and aided in scalability. It was important to have systems that could communicate effectively, especially when adopting new technologies.

Kaustubh: With the implementation of EHRs, did structuring data across various positions contribute significantly to compliance?

Karen: Yes, you need to be able to exchange data compliantly. And, even before HIPAA existed and PHI had a definition, people consciously knew the importance of communicating accurately and having those systems talk to each other. It's a collective effort and remains so.

Kaustubh: Yeah, as you said, you know, healthcare has quite a diverse set of stakeholders and it requires [I think] varying degrees of collaborative efforts from all of them to make the process a success. So definitely the kind of transformation that you've seen and brought, in fact, lays the groundwork for what AI tools can achieve as it's because of that kind of standardization, and that level of digitization, that AI can truly harness the information that exists in a meaningful manner and then add valuable layers on top of it.

Karen: Yeah. Well, we don't harvest half of the information we collect, right? There's only so much we can see because there's only so much we can ingest. That's why I get excited about what AI is doing to the industry. It helps us to concentrate on what's important, considering the vast amount of data we deal with. It's what we need because, within and beyond our industry, there's so much. And we all just need a focus.

Kaustubh: Yeah. In fact, on this note, there are of course multiple technological innovations happening. Healthcare is amongst the most underdeveloped industries from a technical perspective, and that creates the most exciting opportunity for improvements as well.

So, sitting where you are, from this high-level perspective, where do you see technology playing the biggest role? And what kind of technologies do you feel have, uh, you know, in the next five to 10 years, the most impact lander?

Karen: I believe the most significant impact will come from predictive analytics. The vast amount of data we have, which is still underutilized, can be leveraged more effectively in healthcare. This surprises me because, I mean, we use it for football. Predictive analytics is crucial, yet often overlooked.

Now, when you build out a network, we've traditionally focused on meeting minimum requirements. And most of the time people accessing health care aren't necessarily participating in it on purpose [because we have to or we go because we need to]. So I think AI can help transform this 'adequacy' into optimal network performance.

"I believe the most significant impact will come from predictive analytics. The vast amount of data we have, which is still underutilized, can be leveraged more effectively in healthcare. This surprises me because, I mean, we use it for football. Predictive analytics is crucial, yet often overlooked."

The latest and greatest is accuracy. Most of the data may not be accurate. So there's no reason we shouldn't be leveraging AI to identify the discrepancies or the redundancies. We should just be using it to eliminate admin tasks that just waste time and money for every insurer that exists today. Another area is in addressing availability issues, especially post-COVID, where healthcare provider burnout is prevalent. For example, if a patient needs specialized surgery elsewhere, can AI help streamline the process and facilitate better outcomes at a lower cost?

None of that flexibility exists today. Plus, for clinical trials, you can use all that data [you get]. I do know why it is expensive and then people don't use it. Well, we need to learn.

Kaustubh: Absolutely. I think that makes a lot of sense. All of these points are going to be addressed at some point in the future. I feel, uh, the sooner it is, the better.

Karen: I'm counting on you, Kaustubh. I'm counting on MarianaAI to do this.

Kaustubh: Yeah, that's exactly why we launched MarianaAI. In the end, all of these problems are going to be solved with the significant help of future technologies that emerge - the kind of AI technologies we are building today, the kind of technologies that could exist in the future.

All of those would, you know, fit in seamlessly across all the niches that healthcare has. So if you look at what MarianaAI does right now, we take clinical encounters as a unit and we streamline it from pre- to post-encounter. And that includes a collaborative team of specialized AI agents working with the physician in real-time.

In a manner that's seamless and emulates that typical software. So, within the past 12 to 16 months, since large language models have begun to come into place, the impact has been quite massive, on multiple fronts. So going forward, you can only expect this to increase and the aspects that you mentioned are, you know, reconciling information, ensuring accuracy, and not only of the original information but also of the outcome, depending on the severity of the case in question.

All of these things are going to play a huge role in how AI Medical Scribe is adopted across the ecosystem. Different types of organizations have different types of people, and they have their own preferences and methodologies for going about this. So in that context, how would you recommend a payer or a provider organization to prepare themselves for these future technologies and adopt them in the best manner possible given that you have already driven them to adopt Epic and Cerner in the past?

Karen: To add on to your point, what we all want is time back and we know that time is money. MarianaAI's impact in healthcare is about giving time back to doctors and clinicians, reducing their administrative tasks through automation. You guys offer a system where doctors can work with patients more efficiently, using AI to provide suggestions based on the patient's history and current needs. Like if my doctor actually was listening to me during the visit, we'd probably solve things a little bit faster. This leads to better patient experiences and quicker resolution of health issues. And this efficiency can translate into cost savings, addressing healthcare inflation. Now, this amount of money that's not wasted, it's used somewhere else, right?

"The key to effective healthcare is [in] optimizing the total cost of care, identifying the most cost-effective treatments, and ensuring the right follow-up services.”

We see healthcare inflation. Everybody just keeps raising their rates. We can't afford healthcare. That's why the government intervenes and makes up another program. They're doing the right thing because they're trying to change it. The key to effective healthcare is [in] optimizing the total cost of care, identifying the most cost-effective treatments, and ensuring the right follow-up services. We're [leveraging,] giving back the time to allow people to understand, that they are part of this. I envision the patient being part of their healthcare, whether they know it or not. AI has the potential to activate many existing tools more effectively, connecting them seamlessly to enhance patient care and involvement.

Kaustubh: Yeah, absolutely. You could imagine in the future, these AI calls, with the voice of your primary care physician coming to you every day checking up on whether you did your daily exercises.

Karen: Well the therapist, you know, they're a little less daunting. But I think the other thing is, for these big systems, or any system and even the industry in general, they need to hire people that also understand change and how to communicate change.

Making things more efficient and getting back time and allowing basically money to be saved or money to be made, clinicians will change, insurance companies will change [for money]. Everybody will right? Everybody likes money. But the patient doesn't necessarily have the same financial impact.

They may, they may not. So how you get them involved, that really comes down to how the whole thing changes - you need to have almost like a 'change champion' who can communicate the benefits and manage expectations. It's crucial to have a clear roadmap and consistent communication throughout the process. This ensures everyone, including physicians and patients, is aligned and prepared for the changes, avoiding situations where either are unprepared.

Kaustubh: That's actually a really good point, right? That even within a given organization, oftentimes change towards the better is hampered because the bottom 20 percentile just find it hard to change their existing ways.

How have you seen organizations responding to change, to new methodologies or technologies?

Karen: So I've seen good and bad. I'm going to focus on good. From my observations, successful change management starts with top-down leadership understanding and commitment, ensuring that there is accountability throughout their organization to do so. Most of us work remote now, So communication is crucial; it should be frequent and clear, and align with familiar processes to ease transition.

"So I've seen good and bad… From my observations, successful change management starts with top-down leadership understanding and commitment, ensuring that there is accountability throughout their organization to do so."

The presence of support personnel, or 'boots on the ground', is vital, especially in remote working environments. These individuals assist in resolving issues as they arise, akin to a help desk system. This approach makes the change less daunting and more user-friendly, similar to customer-friendly services we use daily.

However, there will always be individuals who resist change until they face direct consequences, like payroll issues. Therefore, having support readily available to guide these individuals is essential. In healthcare, this resistance to change is more pronounced. Tools like Mariana AI, which enhance patient engagement, are underutilized. I believe that leveraging such technologies can transform patient care by making it more engaging and effective, which is crucial for the healthcare industry's evolution.

Kaustubh: With Mariana AI automating post-encounter communications and the growth of remote patient monitoring, what other technological changes or processes do you think will significantly impact patient care?

Karen: Integrating reminders for essential tasks like medication pick-up and involving caregivers in the process is crucial. All pharmacies have systems that can notify when medication is ready, and incorporating these reminders into Mariana AI's system could be very effective.

Another aspect is leveraging remote patient monitoring more comprehensively. For instance, if a patient, like my mother, has been sitting for too long, technology can notify her or suggest chair exercises, especially after her therapy benefits run out. This illustrates how technology can connect various aspects of a patient's healthcare.

The key is to use health information technology to encompass the entire spectrum of patient care. While we might successfully address the clinical and health insurance aspects, involving the patient in their own healthcare journey is vital. The industry often overlooks this, but it's essential for effective healthcare. That's why, again, I say my vision includes the patient, the patient being part of their health experience, whether they know it or not.

Kaustubh: I think all the stakeholders need to be brought in and be focused upon as equals, not leaving any particular one behind. As we wrap up, I'd love to hear your final thoughts and any insights on Mariana AI's roadmap and how it can connect with different parts of the industry.

Karen: I'm enthusiastic about AI's potential in healthcare, particularly MarianaAI's evolving roadmap. The industry's future seems promising with such innovations. I'm keen to explore ways to integrate Mariana AI into various industry segments. And there's some of that you might not be able to share just yet. But, I'm excited to see where power, where the industry is going.

And I don't want to put you on the spot, but essentially, you know, how, how do people access MarianaAI? And what are the things you've found already that are your differentiators?

Kaustubh: Yeah, so, of course, there are two facets that we are focused on. One is the comprehensive nature of the platform. We focus on creating a seamless experience, automating non-caregiving tasks for physicians.

But in general, it's in the individual features that we have. As a seamless platform, we are able to utilize the synergy that exists between each of the steps. So if you look at a patient, you know, the first time they come in, a doctor needs to go through their entire background to understand them well enough. Now, typically this process again relies a lot on human intervention. It misses out on a lot of details. So with MarianaAI, they are able to use that entire history. we can create that full view of the patient - enhancing physicians' understanding and decision-making during encounters.

As the encounter proceeds, we get the HCC codes from their background as well, and that, of course, takes care of the Medicare and Medicaid reimbursements, which often comprise the bulk of the volume that the provider is getting. This is further supported by our real-time configurable decision support engine, which documents differential diagnoses, medication options, and more, tailored to individual patient cases and the clinicians themselves. So if you look at how physicians spend their time today, we are aiming to handle around 80 to 90 percent of non-caregiving tasks right off the bat.

The goal is to create a team of collaborative AI agents that are deeply integrated with the healthcare providers, offering seamless care. That end-to-end spectrum of caregiving, including the financial aspects is, of course, a fundamental differentiator because you wouldn't find anything remotely like this anywhere else. But even along with this, the AI models that power this are highly specialized. Fundamentally, we look at ourselves as a healthcare AI research lab because, both me and my co-founder, Vishal come from a technical background. So the expertise that we have is on how to build AI models for nuanced domains.

Our vision is to fundamentally transform how healthcare will be delivered in the future, making it more efficient and effective. The thought process is very simple, right? 10 to 15 years down the line, almost certainly healthcare would be delivered in a significantly different manner than the way it is today, and we want to create that future or bridge the gap to that future with the platforms that we're building and rolling out. So the biggest challenge we face now is reaching the right stakeholders to showcase the potential of Mariana AI in revolutionizing healthcare.

And it's people like you who are helping us expedite that journey and get us to where we need to be to be in front of the right people.

Karen: Starting small is key, especially given AI's rapid evolution. Large integrated health systems may find adapting to change challenging, but beginning with one service area and gradually integrating technology into existing workflows can be effective. This gradual approach allows for assessing what processes can be streamlined or eliminated. However, it's important to acknowledge that each health system and individual patient is unique.

Smaller practices could significantly benefit from the communication tools AI offers, as they often lack extensive administrative support. The U.S. government recognizes the need for change in healthcare due to rising costs and inflation, and AI technology like yours could play a crucial role in this transformation. While starting with government initiatives is an option, focusing on smaller systems in need of immediate assistance could be more impactful.

"The next generation will likely be more receptive to such technological advancements, fully embracing AI's potential in healthcare."

Large systems and academic medical centers might have the resources to implement widespread changes, but a gradual, targeted approach is often more successful. The next generation will likely be more receptive to such technological advancements, fully embracing AI's potential in healthcare.

Kaustubh: Thank you, Karen, for this insightful discussion. Your background and perspective are invaluable, and it's clear that change in healthcare delivery is imminent. We look forward to collaborating and being part of this transformative journey in healthcare.

Karen: Working together is an honor, and I'm excited about our collaborative potential.

Kaustubh: It's been a pleasure speaking with you. Let's stay in touch and continue our efforts to innovate in healthcare.

Karen: Looking forward to it. Best of health to you.

Kaustubh: Have a great day, and let's keep pushing for positive change in healthcare.