HIStalk Interviews Kevin Phillips, Business Category Leader, Philips Capsule

Kevin Phillips, MBA is business category leader of acute care informatics with Philips Capsule.

Tell me about yourself and the company.

I joined Philips in 2021 through the acquisition of Capsule. I’ve been with the company for 16 years.

Now more than ever, we have an opportunity to make an impact on improving how clinicians can more effectively care for their patients and reduce the tedious elements of documentation, the steps to find information, and the need to react to details, all to allow making faster and more informed decisions.

How does the former Capsule Technologie business fit into the healthcare strategy of Philips?

It’s clear for us that to support the move toward virtual care, to remove some of the manual repetitive tasks for the frontline care team, we need to find ways to arm clinicians remotely to support and assist that frontline care team. It all starts with how we harness the live data that is connected to all of those devices around that patient bedside,

We started, decades ago, with how we automate data for use in medical device integration to the EMR. That has now expanded to, how do we leverage that same data and transform that into actionable insights in a host of different locations? It could be within the central station at the nursing center station. It could be at centralized virtual care centers. It could even be at the mobile handset of that care team. They can now start to manage their patients really anywhere.

It started with the observation that live streaming data is a critical resource for moving from reactive to proactive care.

Once all of that data became available and the opportunities to analyze it became obvious, how did it shake out whether that would be done a company like yours or the EHR vendors themselves?

If we look back over the past 25 years, medical device integration was once a Class Two medical device. Most EMR companies have historically shied away from moving into the medical device realm. Also, connecting to these devices requires at times hardware connectivity such as hubs and unique cabling that gets deployed in the room.

For those two reasons, while a few EMRs have moved in that direction, it was a logical place for a dedicated entity to focus. Philips, as a leader in patient monitoring, said, we have the capabilities to connect to all these devices in the room. Not just manage them for getting data into the EMR, because less than 1% of that data makes its way into that patient record, but to leverage all of that data for more proactive use cases as well. Other use cases around full disclosure databases, where you can dive into risk initiatives or leverage things for alarming and alerting as well.

Moving more and more into this regulated space is why EMRs haven’t dipped their toes fully into the space today.

What is the value of that other 99% of data that EHR vendors don’t use?

In most charting systems, you’ll typically chart every one to 15 minutes. The key element is that there needs to be a clinical validation step. That’s why they’re not doing it in more routine fashion.

What is missed when you capture only a snapshot of that data is alarms, waveforms, and device settings that provide comprehensive but subtle insights in patient care, such as the physiology of that patient and the status of devices. If you start to understand those subtleties, can new insights be brought? How can we, through partnerships where these same questions are being asked, better leverage all of that data to leverage some insights as well?

What kind of device monitoring do you do?

One example is the different modes of a ventilator. The data is critical for a respiratory therapist to understand the state of the respiratory care for that patient.  Those sorts of elements give additional context to not just the device, such as ventilators, but if we look at all the devices that are surrounding that patient and capture all of that together, you can have a richer view of that patient.

There are also scenarios of failure modes for devices or sensors falling off. Understanding those alarms or states can give additional clues around how somebody who responds to those alerts remotely, or who provides secondary oversight, can tell the bedside care team or the remote care technician how to effectively manage that patient and how to manage the devices in that room. That’s why even the device settings are of critical importance in many situations.

How much of the company’s strategy is driven by data needs that didn’t exist or weren’t possible a few years ago, such as clinical surveillance, virtual care, and real-time analytics?

Most of our investments are toward moving us into that direction. We’ve invested significantly in the data model of all the drivers that are communicating to each of these medical devices to truly support this semantic interoperability where waveforms, alarms, and device settings can be liberated and ubiquitously understood by all endpoints.

Secondly, we are seeing this move towards flexible, centralized monitoring units. Remote virtualization, where patients may not move from one care setting to another, but the technology does and the care team moves and adjusts with them. To do that, we needed to find ways that we deploy this data management backbone so that data can be leveraged. But also tools, applications, holistic viewers, and the ability to alert assigned care teams to changes in that status.

The technology now supports these new care models. But there’s also this collaboration with our clinical services team and clients that help understand their objectives are and how we can help them change how care is delivered from that in-person to remote virtual care location as well.

Is the level of EHR integration as deep as you need? Does the EHR need to follow you along as you come up with new concepts of how the data you capture can be used on their side?

Yes, but I would say that sometimes the EHRs get maligned with “they’re not interoperable” and “they won’t share data.” We share a lot of the data that we capture from devices. The context around the ADT, labs, and patient history is fairly easily obtained from the EMR. 

This relevant context, when paired with live data, that deliver this additional insight. A lot of the research and partnerships that we have with key academic hospitals like MGB are allowing us to identify these observations and then deliver those insights back into production.

I find the EMRs to be actually quite collaborative in this. For the clinicians, these are contextually launched within EMRs. It’s actually a quite collaborative process across the board.

What opportunities does AI present?

We have used machine learning techniques to help us build a variety of our algorithms that are deployed in our solutions today. We’re also seeing that new agentic AI helps us streamline mapping that we use within our drivers, obviously with human validation at the back end.

We also have many different reporting, retrospective reporting solutions. We have surveillance tools where you can see alarm events trends, but sometimes you just want to ask a simple question about what has happened. These are areas where generative AI and assistants can be deployed in these tools. We are continuing to explore that area and drive it into the solutions moving forward.

Agentic AI will allow us to reason with the data and eventually  integrate video over time. We can reduce and streamline unnecessary workflow steps. That is just fascinating. In all aspects of our life, we are evaluating how AI can reduce the number of steps to get certain activities completed. That’s no different than what we are trying to accomplish within Philips as well.

What factors will be most important to your business strategy over the next few years?

We have seen tremendous advancements in technology. But we have to co-create with hospitals to identify how we can support clinical adoption and change workflows. Activating virtual nursing, virtual observation, and virtual surveillance use cases requires a shift in the activities that are done at the bedside, which activities are done virtually, and how that collaboration occurs. There is certainly a big push and a need to move towards that. But we have to figure out how we continue to collaborate with our health systems to maximize and streamline that workforce.

There is also the reimbursement landscape and the regulatory landscape around deploying AI into solutions and medical devices. We expect to see additional guidance from the FDA. How can we identify the best pathway to introduce this new technology in a safe and effective way? That’s always our core focus.

We could focus on a million different use cases. Our focus is to co-create with specific leading health systems and work backwards from highly impactful use cases. 

If we do these things, our investments will have a global impact with the clients who use our solutions.

HIStalk Interviews James Lakes, President, Mednition

James Lakes, MSc is president of Mednition.

Tell me about yourself and the company.

Mednition works with healthcare leaders and providers who are looking to improve the clinical, financial, and operational performance of their emergency departments. The ED is the front door for about 40% and up to 60% of all their inpatient volume.

We believe in the concept of catching things early, whether that be high-risk conditions or sepsis. Catching the problem early, putting people on the right care track, and then leading to better outcomes from clinical, operational, and financial standpoints.

I spent almost 30 years of my career working in health IT, primarily at big platform companies like Microsoft, VMware, and Salesforce. Over the last four or five years, I’ve focused more on startups, which led me to Mednition.

How is AI for nurses and operations philosophically different from physician-focused AI?

No ill intent towards any of our clinical leaders, but when we’re investing in innovation, in the provider space in particular, the focus is typically physicians. These are your highest-cost employees. They have high benefit, whether that be thoracic surgeons or orthopedic surgeons.

We often see that because EDs have traditionally been considered a loss or cost center for most providers, there hasn’t been a lot of investment in innovation for ED nurses and ED departments in general. The nurses in triage ED are the first people to meet and assess the patient. Helping them get more accurate within their acuity setting to send them on a path with providers leads to better outcomes across the board.

Upwards of 55% to 60% of clinical staff are nurses. We believe that helping them is just as important as helping physicians.

How do nurses choose that path or make triage decisions?

We based our software on the ESI model, the Emergency Severity Index, which is sponsored and built by the Emergency Nurses Association. This is a scale from one to five, five being the least sensitive or least urgent, with one being the most urgent. About 85% of all ED nurses in the country are trained on that model. We based our model on that, and we have a deep research partnership with ENA to improve our model to make sure that we’re taking best practice into consideration.

Nurses spend more time with patients and have intuition about outcomes that aren’t found in charts. How do you incorporate that?

A nurse has two to four minutes in that triage moment with the patient. They are interviewing, getting their primary complaint, taking their vitals, and observing that patient to make a decision on what care path this individual should be on. In that time, they interview you and observe you, but they really don’t have time to check all of your clinical history, the accuracy of what you tell them about the medications you take, or what conditions you have had previously.

Our product is called Kate, which is the name of the daughter of one of our co-founders. She showed up in ED with an inaccurate acuity setting and was in serious condition. She survived and is now a young adult. The founders wanted to prevent other families from going through that.

Kate  goes to work in the background. The nurse does their observation. They set their acuity setting. Kate looks at the patient’s health, their history, their medication lists, all those things. She compares that to millions of cases in our model and comes back with an acuity setting only if she differs with the nurse. If not, she doesn’t send anything, which means that she’s not disturbing the nurse unnecessarily.

The only time the acuity does not matter is when Kate suspects sepsis. Then she will automatically send an alert regardless of the acuity setting of the nurse. Sepsis is the biggest killer in hospital settings and testing, getting antibiotics in them, and acting quickly, leads to better outcomes. That time to action is important.

Executives in some health systems influence software decisions more than frontline clinicians. How do you sell the product?

Like any startup, you work early on product-market fit. Next you figure out what your target audience is.

We know, and can see in our engagements, that ED nurses who are on the front line, their managers, and their directors are our biggest champions. However, often they are not the ones who make the final decision or have the budget. We understand the nurses deeply and work with them collaboratively to take that to their leadership. We highlight the issues, how Kate addresses them, and then build the case for the ROI from a clinical, operations, and finance perspective to justify the investment.

We’re getting better and better at it. The nurses are getting better and better at advocating for themselves with our support. But we do see that it’s a broad stakeholder sales engagement. It requires having champions at the front lines, but then making sure that the leaders at the top who are making the decisions where those critical investments go are aware of the impact of Kate.

What is the value proposition that you present to the CEO, CFO, or CIO?

I’ll use an example of a provider that we worked with recently. Within six months of going live, Kate actually paid for herself. They tracked  a number of metrics. One was left without being seen. Patients who show up in the ED will leave if they wait too long, and that can lead to worse outcomes. The might have to come back, or it could be lost revenue because they go someplace else.

Length of stay is also a big issue. This organization was able to not only lower their left without being seen rate by 1%, but also lowered their length of stay by 23 minutes in less than six months.

On the financial side, because they had better accuracy and better documentation, they saw $400,000 of additional financial revenue because they had fewer down charges and fewer denials from payers.

How well do health systems integrate what happens in the ED with opportunities for long-term patient engagement or revenue generation?

The market is getting better at recognizing the value of getting things right at your front door and the downstream impact of that. It’s definitely a shift in mindset and a shift in focus.

How does Kate integrate with other health system platforms?

We integrate with the EHR, specifically Epic and Cerner. We are working on a couple of others, such as Meditech and Medhost.

The good thing about Kate is that she’s working in the background. She alerts or notifies nurses only when she has to. It’s  not obtrusive to them. We talk a lot about alert fatigue for clinicians. We have made it a distinct focus of our company to not be a notification alert problem for our clinicians.

How much evidence or background do you provide along with the recommendation?

Kate provides a message. The nurse says it’s a three, Kate thinks it’s a two, and she delivers why she believes it’s a two. The nurse then has the option. We don’t make the decision for the nurses. They have the option to change the acuity setting to a two.

Typically when they do that, it’s because Kate has identified something they may have missed, or they may have made an error. Then they correct that error or they improve the documentation when they do the up-triage, as we call it, that then drives the decision, which leads to better outcomes downstream.

When they don’t act, they typically document why, which leads to feedback to us. We have a clinical team of physicians and nurses, some of the top in the space, who review those cases and feedback from the nurses. We use that to generate cases and continuously improve our model.

Having come from big tech companies, what are their advantages and disadvantages in their involvement with health technology?

Those companies have incredible R&D teams, incredible market reach, and incredible flexibility in those platforms that provide a tremendous amount of value. But when you start to get down into deeply research-oriented, specific use cases, the specific clinical decision support, they will struggle, because they are trying to build a big platform to then fit into various scenarios in healthcare. Whereas when you’re a startup and you’re focused on that specific problem, you can get very, very good at it.

I’m a former athlete. It’s like thinking about Steph Curry. He’s a marksman, maybe the best marksman we’ve ever seen. He’s a specialist. He’s deeply talented at that one thing. If I tried to make him an all-around player where he was going to be the best defender, the best rebounder, the best passer, and all those other things, he may not be able to be world class across all them all. I might dilute his talent. It doesn’t mean that there aren’t any players that can do that, Michael Jordan being one of them, but there aren’t many, and that’s why they stand out.

If I have a person, system, or solution that is really targeted and can be world class at that, that is the benefit of the startups. The platform companies can bring broader value across broader spectrums, but they may not be as specific as you need for things like sepsis identification or triage acuity.

What is the present state of the healthcare buying market from the viewpoint of a startup that is trying to scale, and how will it look in a year or two?

We’re all concerned about what will happen with changes in Medicaid and any reductions in Medicare expenditures. What does it mean? In our particular space of EDs, a recent Vizient report says that they anticipate higher volumes in the ED because more and more people who lose coverage will use the ED as a form of primary care. When they use that as primary care, they typically wait until their state is dire or even worse. Acuity and severity will become more difficult and intense for our EDs.

If you have increasing volume and increasingly complex cases, it becomes a recipe for potential chaos in our EDs. We are hoping to help them alleviate some of that by being proactive with something like Kate.

What is the company’s strategy over the next three or four years?

We will have a sepsis breakthrough designation for our Kate sepsis model in early 2026. We are submitting for final approval from the FDA. We hope to have that by the end of Q1. I think we will be the only software product with FDA approval for sepsis early detection.

We are including more and more models. We have a partnership with a leading children’s hospital and research center for building triage models for people under one year old. We are working all the way down to infants, both for triage and sepsis.

Another model that we’re working on is continuous monitoring for sepsis post-triage. You’ll see some of that come out in early 2026.

We will continue to add models. Our focus will primarily be early detection of any high-risk patient condition.

The biggest thing is being an advocate for our ED departments and the ED leaders out there, They are sometimes not the first line of thought for for our healthcare leaders. If you talk to a CEO of the health system, they can often tell you who all the cardiac thoracic surgeons are and who their family members are, because that’s a high revenue focus area for them. But if you bring that attention to their ED leaders or ED departments, that’s not their focus and they don’t have that same familiarity. We  try to be an advocate for that space because that front line sometimes gets forgotten.

HIStalk Interviews David Lareau, CEO, Medicomp Systems

David Lareau is CEO of Medicomp Systems.

Tell me about yourself and the company.

I have been with Medicomp for a number of years. Our core competence is that we produce a clinical data engine that we call a Clinical Knowledge Graph. We’ve been building it based on expert input since 1978.

We have been through many technology transitions. Now we’re in the world of AI with clinical applications. We feel that we are well positioned in that area since we have very domain-specific knowledge for training small models to do what we need them to do.  

Will the customer cost of using AI technology that is sold by big tech firms that have to keep investors happy going to inevitably increase?

We think that the Butcher’s Bill is going to come in for these large models that are expensive to use. People already are starting to say, “We are going to use AI to train for specific workflow issues and specific clinical domain issues.” We believe right now that we are well positioned in that space.

We are having great success in terms of performance and lower cost by using a small model rather than a large model. Our engine, and all the power that’s in it, can be run on CPUs rather than GPUs, inside a vendor’s own security environment without going out there. We can do that because we have a target of 400,000 clinical concepts with hundreds of millions of links for diagnostic relevancy and coding relevancy. To be able to do that with a small model is because we have a clinical target that’s been very well-defined over the last 40-some years. 

The vendors that we are working with have seen their costs drop by using a smaller domain-specific model that is trained on our clinical data points that link to our engine. The roadblock to that was the lack of a clear standard for how to communicate between applications using various aspects of AI. The MCP, or Model Context Protocol, developing as a standard has allowed us to expand the number of partners that might be able to take advantage of our unique Clinical Knowledge Graph.

That’s why what’s happening in the industry is an opportunity for us rather than a threat. If I want to know the 250 clinical concepts that might be relevant for somebody with chronic kidney disease, that’s in our engine. The MCP allows us to present a standardized way to request that information from our engine and send it back to the application, whether it’s a language model or somebody else that wants that information. It will be what drives the integration of all these AI agents that people are building.

It was an essential, missing building block for communications between systems that are using AI to do very specific tasks. When Epic, for instance, announces that they’re building hundreds of AI agents, they will be using things like MCP to communicate between various aspects of their system.

Has AI changed your business strategy?

It has. We definitely have had to adapt to it. The conversational AI still captures text. It does a very good job at it. We’re really astonished and pleased at how effective it is at removing the need to enter text into a medical record. But it’s still text.

We have been using AI. We’ve been using language models internally to fine tune our offerings and our tools. We are building a small model, domain-specific, task-specific ways to use our data, extract data from text, and then operate on it to service all the downstream things that you have to do, like quality measures, adequacy of documentation for Medicare’s Hierarchical Condition Categories, that sort of thing. We’ve had to embrace it and figure out how to use it transparently, effectively, and affordably in the clinical domain. 

It is an exciting time as the AI tools have matured, the power has matured, and you have everybody in the industry rowing in the same direction. But they need clinically specific tools to get where they need to to make it affordable and useful at the point of care.

We saw it as more of a threat two or three years ago. A threat being anything that causes people to not need to do business with you today is a competitive threat. When the frenzy over AI started a few years ago and really built lately, it really was a competitive threat to us because it made people sit on the sidelines and wait to let AI figure it out. Now people are realizing that generalized predictive pre-trained transformer is not enough for clinically specific work. That’s where we are hooking it into our Quippe Clinical Knowledge Graph to do very specific things for clinicians. People are realizing the proper uses of AI in clinical medicine and the things that it doesn’t do so well.

We are pleased with the way things have developed over the last 12 months, as the rubber is starting to meet the road with AI in medicine.

Startups and big tech companies might be slow to realize that AI and ambient documentation are table stakes that aren’t much of a business moat. Does your phone ring from companies that have the technology but need help understanding how to integrate it into healthcare workflows?

I get four or five inquiries a week. Most people that call when they hear the specificity of what we’re doing say, “We’re not quite to that point yet. We’re just trying to figure out how to compete with all the other people that are in our space.”

We’re starting to see that people are actually putting these applications into use. Those are the more serious inquiries when calling us. They say, “We’ve got the table stakes working, but now we’re having trouble meeting all the downstream requirements because we just have text, we don’t have data.” They need to get there because when the government puts in very specific requirements for things like quality measures. They are looking for specific data points. That’s what’s in our engine.

They say, “The acquisition of documentation is no longer an issue. Now we need to do something with all the downstream processes that are tied to the information in that text.” When they hit that wall, that’s when they’re calling us.

You wrote something about instafraud, the claim by insurers that some providers are using AI to increase billing, and their intention to use AI to stop it.

We’re in initial conversations with some folks in compliance and regulation. One example where it shows up is in Medicare Advantage, which was supposed to reduce the cost of caring for people people in Medicare. It uses risk adjustment codes, Hierarchical Condition Categories, to identify somebody who has a disease that puts them at risk of poorer outcomes, and then to manage those conditions. But to do that, you have to code a diagnosis to get that risk, and then receive more money put in your risk pool each year. 

AI was algorithms even before AI became a thing. People were using algorithms to say, “This guy has a high creatinine. He probably has chronic kidney disease, so let’s code that.” If you code it and send it, you get a higher risk score, but the documentation has to support it. 

When we published our E&M algorithms when the 1997 guidelines first came out, the most common question we got was, “You guys have all this data that can support a code. Could you use it to tell us the three things we need to do to get a higher level of service to get more money?“ We said that we could, but the government has seen what we’re doing and warned us off and said, “If you do that, we’re going to come after you.” So that feature was disabled. You can’t ask it the minimum you needed to document to get a higher code.

The same thing started to happen about five years ago with risk adjustment. They called it “suspecting.” They wanted to use AI to look at the record and find potential evidence for one of these HCCs that would support a higher code. This would be submitted without necessarily seeing that the documentation supported that the patient actually had that condition.

Suspecting is a valid thing if there is a condition that’s unaddressed and you then address it, but it’s fraud if you look for the possibility that somebody has something and then code that they have it without investigating whether they actually do.

There’s a tug-of-war going on between the payers, the regulators, and the enterprises over the proper handling of patients with chronic conditions in the Medicare Advantage program. The government is starting to pay a lot of attention to that because Medicare Advantage was supposed to cost less per patient than traditional Medicare fee-for-service and that hasn’t turned out to be the case. I think it’s because people are over-coding for risk factors.

How does Epic’s public sharing of its AI roadmap affect innovation?

There are a number of layers to that question. Epic is not the only large vendor that I would call an impediment to innovation just because they’re a dominant in their space. We do quite a bit of business in Asia, and this is not limited to the United States.

Years ago, we had another unnamed vendor in the US, not Epic, whose customers told them, “We really like what this niche vendor is supplying. I want it.“ We had the experience with a different vendor years ago, where they said, “We’re getting a lot of customers that are asking for what you have. We’re probably going to develop that ourselves so we’re not interested, but we are willing to work with you. But since we think of our customers as an asset, you’re going to have to pay us the bulk of your revenue for access to that asset.”

That’s the moment where I realized that dominant vendors, because this vendor was large in the space then, tend to treat their customer as an asset and as turf that they own, not as an obligation to provide a higher level of service. When vendor app store organizations were first set up, the agreement that you had to sign as a niche vendor said, “We need to vet what you have. You need to show it to us. You need to show us your source code so that we can make sure it doesn’t create any vulnerabilities on our system. But if we then decide to do something like that ourselves, you have no recourse to us.”  That scares off the niche vendors.

We’ve also had the situation where a large consulting company that specializes in implementations for the large HISs said, “We have a lot of customers asking for what you have. We have a lot of people asking for the kinds of things that you and other companies like yours provide. But we also have $90 million a year in consulting revenue from this vendor for implementation assistance, and they’ve told us that we’re putting that at risk if we start to introduce these niche best-of-breed vendors into their ecosystem.” So it really does stifle innovation in that sense.

Do companies call you wanting to buy Medicomp rather than try to build complex healthcare technology?

Yes, we get that from investors, private equity, and larger vendors. The issue for us is that we do one thing. We focus on it. We don’t do anything else. Being employee owned, basically, allows us to focus on what we do and not get distracted, and we plan to keep doing that.

As people see how you can leverage our Clinical Knowledge Graph for a very specific thing that has a widespread need, we get a lot of activity. I thank people very politely and explain that we see a model of sustaining what we’re doing for quite a while, even into the next generation of the company.

What will be important to the company’s strategy over the next few years?

We need to make certain that we are using these new AI tools to make ourselves more productive while producing our Clinical Knowledge Graph. With the new Model Context Protocol, MCP, stuff that is coming out, we’ve been API based for a long time. We make it easy and transparent to link to our clinically data specific APIs to accomplish specific tasks. Acquiring documentation, no longer a problem. Acting upon it and doing something with it is the next step.

Our strategy over the next three to four months is that some EHR vendors will start showing the intersection between ambient listening coming in as documentation, then link to our engine that will validate, filter, and present that information and accomplish all the specific things that you have to do with the data, such as getting the right billing codes, meeting the quality measures, and verifying adequacy of documentation for HCCs. Linking our stuff and allowing our engine to be accessed through the MCPs to accomplish specific tasks other than just documentation.

We see great potential in that space. We will have the first few implementations of that hitting the market over the next few months with some specialty-specific EHR vendors.

HIStalk Interviews William Cavanaugh, CEO, Concord Technologies

William Cavanaugh, MBA is CEO of Concord Technologies.

Tell me about yourself and the company.

I’ve been in technology for over 30 years and health tech for 20. I’ve worn just about every hat there is to wear in a healthcare technology company, from making the coffee, developing the software, taking out the trash, closing the deals, and writing the business.

The high-level mission of Concord Technologies is to advance healthcare through universal exchange and intelligent processing of data. We leverage advanced AI to drive a smarter, faster, and more connected healthcare ecosystem.

What kinds of documents do health systems receive and what challenges do they experience in processing them?

The big challenge in healthcare is the exchange of data between disparate healthcare entities. You have to look at the volume of data. There are 2.3 zettabytes of data generated every year across healthcare. What’s a zettabyte? I can tell you that it’s a billion gigabytes, but that doesn’t really represent the challenge.

If you look at one hospital to paint the picture, one hospital creates 50 petabytes every year. Again, that is difficult to comprehend. Picture yourself in an NFL stadium, in the upper bowl. If you printed the physical equivalent of the annual data from just one average hospital, it would fill 750 NFL stadiums to the brim, and it is growing at 36% a year.

Now you need to share that data. You can’t email it to a doctor because it will go to junk or spam. You need a secure, ubiquitous way of sharing that data. Everyone thinks that the big EHR vendors are going to solve the problem, but there are 500 EHR vendors. They are also not the only player when you add in radiology information, PACS, payers, and pharma. There are thousands of different systems.

On the entity side, the US has 6,000 hospitals, but the number blooms over 200,000 disparate entities and growing when you add in post-acute, outpatient, private practices, urgent cares, specialty practices, et cetera. The problem that we are solving spans 200,000 disparate entities, 1,000 software vendors, 2.3 zettabytes of data growing at 36% a year, and you need to share data.

The space that we play in is documents. Think about documents between your payer, pharmacist, EHR, specialty, and primary. Our very large customers do big volumes. We do about 22 million pages a day through our network. Our big customers do over 50 million documents a month. One of our big EHRs does 90 million a month. We bring that data through an exchange protocol, universal protocol, and then we like to say that we bring it to life. We classify the document, extract key pieces of information, and then insert it into the systems that we’re on.

People might think of interoperability as a FHIR-based data exchange. How does that approach coexist with how documents are managed?

I always say that we’re not in the fax business. But at 10,000 feet, we are a fax company, even though we don’t use paper and fax machines. We use the digital fax protocol to exchange these documents.

FHIR has been around for a long time, plus HL7, integration engines, QHINs, and HIEs are trying to create the structured data exchange. We keep it simple. You have a phone number, and from any EHR, you click “send document.” If MD Anderson wants to send a document to Debbie’s Dermatology in Rice Lake, Minnesota from the EHR, they click “send document” and Debbie’s Dermatology, if she has a fax number, receives a document. Then it automatically sends a response back to the referring physician at MD Anderson that the document was received.

That’s what we do very simply, but we don’t stop there. Your big dermatology clinic gets 5,000 documents in a month. What is this document? We classify it. Then a dermatology clinic is looking for different pieces of information in that 50-page chart that just came across and that a urology clinic would be looking at. We extract the pieces of information, leveraging AI, that are relevant to the receiver of the document. That’s where we bring it to life.

Fax gets a bad rep in the market. I almost didn’t take this job as CEO because I heard we were a fax company, but we’re in the digital exchange business, using a universal protocol.

You asked about FHIR, though. There are instances where FHIR comes into play. We use FHIR to do a lookup to find that patient in Debbie’s Dermatology to match it so we can insert into the system a record. Then we use HL7, which has been around for 15 to 20-plus years as well.

The mental picture of faxing is someone watching thermal paper spool off a fax machine that is covered by taped-on “send” numbers. Is healthcare the only industry where faxing is still a viable way to exchange information?

When you say fax, you think of the curly paper, and if you’re as old as I am, the dial tone. That’s not the business we’re in.

We had a third party do some market research and I’m still surprised by the number of fax machines and paper faxing that is still done in healthcare. Anywhere from at least 10% to 15% of the documents still go through that old-fashioned, corded phone protocol.

Other entities also use fax, both digital and old-fashioned fax. Legal still uses it to fax documents. Payers, the FBI, and the IRS still use it. Other big government entities and institutions, along with mortgage companies, use old-fashioned fax. They’re also migrating to digital fax.

There is still that need when you want a secure ubiquitous protocol to send and receive documents where email doesn’t work, and that fax protocol is still used outside of healthcare. But I would say that around 70% of the digital document exchange via that fax protocol is within healthcare.

How does the process change in moving to digital fax, and what technology criticisms does that eliminate?

The biggest criticism of digital fax is that it’s not structured. By structured, I mean that you are mapping specific data fields from one system to the next. Fax comes in as an unstructured document, such as a PDF, Word document, or chart. It’s not broken down into its discrete fields. 

When that document is received, whether it’s a two-page prior authorization or a 500-page patient chart, it’s just a big PDF. What am I going to do with that big, unstructured document? If you stop just with the digital transmission, even through a cloud-based digital fax protocol, that’s the knock on fax. It doesn’t get me to where I need to be. I still need to scan through the document or read it to figure out what it entails.

With the introduction of large language models, which is the generative AI that is permeating all parts of society, I see the ability to grab unstructured data, pieces of information, from a 500-page patient chart through a large language model that can understand the context as well, which large language models are really good at. They extract the key pieces of information that are needed for the recipient. That will transform how digital fax will have higher quality, lower cost, and better efficiencies for healthcare than try to use things that have been around for a long time. I get to be too geeky, but it’s called CCDAs to structure all these fields in HL7 and FHIR to map all these discrete fields from one system to the other.

Why don’t we just do this mapping and do all this structured data exchange? Again, you just have to look at the volume. Epic has anywhere from 50,000 to 150,000 discrete data elements, based on the configuration, and every configuration of Epic alone is different. Doing that mapping isn’t rocket science, but it takes a lot of one-time work and ongoing effort to keep that up versus just sending the whole document through a secure, ubiquitous protocol that everybody has. You don’t need FHIR, HL7, a QHIN, or HIE. You have a phone number, so you can leverage the telecommunication backbone and security that is already there. Now let technology do the work to bring that unstructured document to life.

That’s relatively new even for our company, and within the overall digital fax industry. But it’s a way to transform interoperability within healthcare.

How much of the information in those documents needs to be integrated into the EHR and other systems?

The unstructured document that comes into the hospital, usually through digital fax protocol, is still probably at least 80% of the transmissions in healthcare. We’re seeing Direct Secure Messaging, and think of that as secure email. Maybe it’s about 10% of the transmissions right now. When you do it through a Direct Secure Message, it comes in through structured, but the challenge is that it doesn’t represent all of the data.

You can’t put an image in there, obviously. You’re not going to structure clinical notes. You still have to provide some unstructured data, which gives context to the recipient, the physician who needs to review the patient who was just imaged at a facility or gone to an emergency room, to get the whole context of the patient.

You call your AI approach “Practical AI.” What does that mean?

We call it Practical AI because it’s exactly what it is. A lot of AI doesn’t add much value. Ours is practical because it’s pretty straightforward and we’re focused on solving real, practical problems. So with 10,000 documents coming into a payer, hospital, or pharmacy, is it a purchase order that goes to finance? Is it a prior authorization with high priority that needs to be responded to within the next 30 minutes because there’s a patient in an ER waiting for that prior authorization? Or is it a claim that needs to be processed in the next 30 days? The first part of our Practical AI is that we’re going to look at this document that just came in and identify its type. 

The other part of the practical side is that in healthcare, nine times out of 10, there’s a patient associated with it, and probably a provider and a record number. We have to extract the patient and identify them by date of birth and address so we can find that patient in the recipient system. That’s a practical use of AI to classify, extract, and then decide what the system needs out of this 50-page document. Sometimes 20 pages and sometimes only three fields. We will make it practical in terms of what’s needed for this incoming transmission for that hospital provider or payer.

How does AI fit into the hype cycle and your company’s business strategy?

It is definitely advancing along the hype cycle and finding some real practical uses. We who use ChatGPT or any of the tools see its ability to digest information in human speech, synthesize information, and create really nice clinical summaries. If the meeting you’re in has three action items, you don’t have to take notes, because it’s going to find it for you. That’s the practical side of how AI is being used.

In our world, we’ve been doing machine learning for over 10 years. It requires a lot of training and use. It gets more challenging and specific with the introduction of large language models. Now you can throw large pieces of information at a large language model, especially when it’s been fine tuned with customized prompts for healthcare, to add real advantages of efficiency, accuracy, and clinical efficacy in the delivery of care.

HIStalk Interviews Peter Bonis, MD, Chief Medical Officer, Wolters Kluwer Health

Peter Bonis, MD is chief medical officer of Wolters Kluwer Health and an adjunct professor of medicine at Tufts University School of Medicine.

Tell me about yourself and the company.

I’m a gastroenterologist. I was at Yale on my first faculty job when I was recruited to join UpToDate as a startup. I joined the company, and along with many other capable people, I was able to lead it to grow, scale, and become a very important information resource that is used by healthcare professionals around the world.

Wolters Kluwer acquired UpToDate in 2008. We became part of a portfolio of information services across different verticals. Those verticals include health, tax and accounting, finance and corporate compliance, legal and regulatory, and corporate performance and ESG.

What are the advantages of presenting clinician-authored or clinician-supervised content at the point of care rather than using the literature search engine approach of some of your competitors?

Let’s frame the issue. Patients expect their doctors to give them the best possible advice. It’s a covenant that doctors would be seeking to counsel their patients with the best possible information.

As it turns out, clinicians have regular questions. When they get answers to those questions, they in fact change decisions about 30% of the time. As readers are out there doing the thought experiment of being with their clinician, imagine that they would change their plan if they had a particular piece of information. Those are the stakes.

We decided to address that information need, which has been well documented, by recruiting a faculty of the best people in the world who are clinically active and who are contributing to the body of knowledge in the area that they are writing about.

We framed the approach by understanding the types of questions at an extremely granular level, having an evidentiary way to look at the body of evidence, make that transparent, rate the level as a recommendation so that it’s highly transparent, and infuse into that the wisdom of these people who are some of the most deeply experienced clinicians in the world.

Human curation not only can summarize the body of evidence, but also can add to that the clinical wisdom and experience of considering factors that are important, such as patients’ values and preferences, to issue recommendations that are granular enough to be used at, or near, the point of care.

Doing that purely as a matter of information retrieval, even with advanced technology, is complicated. The expectation is that that technology can ingest all of that material, present it, prioritize it, and consider all of those factors that I just mentioned to make that experience transparent for both clinicians, and ultimately the patients that they’re serving.

Clinical decision support in its early days pushed guidance indiscriminately on physicians, with the assumption that they should digest it all and also to avoid malpractice issues from not offering complete advice. How do AI-focused tools address that, and could AI itself tailor the content to what an individual physician sees and how they react to the information, such as measuring overrides?

That is the frontier and the challenge, and indeed it’s the opportunity. We have plenty of opportunities to inject knowledge at or near the point of care, both for matters that might be more operationally focused, but also in this high-stakes domain of clinical care. Doing that well can improve care, remove friction, and help to ensure that every patient gets the best possible care, no matter who they are seeing and where they are being seen.

Doing that well is extremely challenging. It requires an enormous commitment to be sure that the experience is as accurate and usable as possible. And where feasible, to include information that is relevant to specific patients and make that experience transparent enough so that the clinician who is ultimately making those decisions can feel confident in the accuracy of that decision, or at least to be sure that they can serve as an interpreter when applying it to the patient in front of them.

To do this well, particularly in this area of decision support, requires a enormous commitment. You have to be sure that all of the different components of that which can break down are done as well as they possibly can be, and to provide an experience to clinicians that is as transparent and as effective as possible.

The business model of massively funded OpenEvidence appears to be running drug company ads that are targeted to the retrieved medical information of the patient. Will clinicians see the ad-supported model as a conflict of interest?

We focus on what we do and have always done well. We have been entirely supported through subscriptions. We have extremely strict policies related to conflicts of interest, particularly among our internal staff, but also all of our 7,500 external contributors, the external faculty and peer reviewers who contribute to UpToDate. We have found that important for maintaining integrity, increasing transparency, reducing bias, and ensuring that our sole purpose is to deliver care recommendations that are clear, unbiased, and free of any commercial taint.

Whether that can be done with a different business model remains to be determined. Ultimately, the market will let us know where the cracks are in that type of a model.

We will continue to do what we do and do well, which is to have a commitment to deliver an effective and easy-to-use experience, focusing on making it easy to do the right thing wherever frontline healthcare professionals are working in their EHR in an enterprise environment or on their mobile devices, Making that experience as free from bias as possible to ensure safety to the best of our capabilities. Providing transparency so that the entire experience is grounded in information that has been curated by humans, and in fact some of the most experienced clinicians in the world.

Will standards of care change as enterprise-associated physicians are provided access to sophisticated knowledge tools while others are financially forced to do without or to use free resources such as ChatGPT?

That’s an excellent point. It really comes down to the matter of how widely governance can be established across healthcare enterprises and small institutions as well. Obviously the governance involved in advanced technology such as AI requires a multidisciplinary approach. It’s not clear that that is going to be available widely for all of the different types of institutions that could take advantage of these technologies.

I do think there is a potential for creating a digital divide, or at least to have some institutions which have governance processes in place and others which may be relying on third parties such as their electronic medical record systems to do that governance process for them.

It ultimately comes down to the safety and effectiveness of the information services, particularly in the high-stakes domain of clinical decision support. For an institution that employs doctors, it’s not just the doctors, but it’s the institution itself that has risk involved, along with the potential benefits of helping to achieve high quality, consistent, and safe care. Having the right information available is certainly a fundamental piece of that equation.

Everybody cites the supposed fact that it takes 17 years to incorporate research findings into frontline care. Will that go away as point-of-care tools can put fresh information right on the screen of the person who is making a clinical decision?

It’s interesting you mention that. The 17-year statement has been cited often, to the point where I decided to hunt down one day the original source of that. In fact, there is documentation, but it’s much more nuanced than that. And in fact, it is not 17 years.

A lot of the adoption of new technologies and new approaches is related not just to having the information available, but also other factors, such as financial incentives, convenience, and superiority over alternatives. But there is a process of information diffusion. 

UpToDate since its origins has done very well to accelerate that process. We have, for many years, showcased some of the newer concepts in a specific feature within UpToDate called Practice Changing Updates. It describes what is new to ensure that our subscribers have an efficient way to know when practice has changed because of new studies, new guidelines, or simply new knowledge that has accrued.

Now with more tools available at or near the point of care, including Gen AI, that process will continue. Ideally, as new technologies evolve and new knowledge evolves, we as a system will have an easier time at implementing them for the right patients.

The physician who is making decisions from the EHR may be presented with patient summaries or suggestions, information they already know but might miss, and new information that they are seeing for the first time. How do you present that without overloading them data they don’t need?

It’s an excellent point. Doctors are overloaded, and that fact is critical to consider. 

Studies have looked at the number of tasks that clinicians have to perform to fulfill all of the requirements that are expected of them. Primary care, for example, would have to have about 26.7 hours per day to complete all the tasks that are required. That is impossible to achieve, obviously, so there’s always a matter of triage. Designing systems that do not produce a cognitive overload is a critical part of the overall design process, and also the approaches of who should be doing what. It doesn’t always have to be clinician facing.

The potential for overloading clinicians is absolutely there. Many organizations are seeking to have that mindshare and to inject knowledge in front of clinicians, and all of it can’t be done. It has to be prioritized and it has to be effective. How that will look is still a work in progress. There are many efforts to do this using advanced technologies, but there’s also a long track record of what works and what doesn’t work.

I’m optimistic that we can do better and that these advanced technologies will have an important role, but the devil is in the details. How will this work within workflow systems? What will the interaction look like with the data that are available within the clinical record, and perhaps even from other sources, to create an experience that helps frontline providers and their patients? That will be the journey that we’re on.

If I can digress for a moment, what is happening to the patients in all of this? All of what we are talking about is taking place in the background, when there is an enormous erosion of trust in healthcare services and healthcare professionals taking place in the backdrop. Patients are increasingly fed up. They are looking for alternatives. The healthcare system is increasingly unaffordable, and it delivers variable quality of care depending on where you are, your level of insurance, and other factors as well.

In more recent surveys this year, 15% of consumers don’t trust their doctors, which is up from 7% in 2023. Only 24% believe that their healthcare systems are focused on caring for patients, down from 77% in 2020. Instead, about three-quarters believe their hospitals are mostly focused on making money.

This process of busyness and the business of medicine is having a fundamental effect, not only on clinician burnout and the actual care delivery, but in a very fundamental way around trust and the experience that patients are having. Ideally, technology will help this problem, both for frontline providers and for patients who are seeking to have a better, more affordable experience.

We are in that potentially awkward phase where some physicians aren’t interested in technology for technology’s sake, but digital natives are coming out of medical school who can’t wait to do everything electronically. How will that change the way that physicians are educated and then trained?

There has already been an organic adoption of technologies, particularly by younger clinicians and those who are trainees. That has been going on for a very long time. It’s really no different that an adoption cycle occurring with Gen AI as well. Although it’s not uniform, clinicians of all ages and career statuses are facile at adopting technologies for it.

But I do think it will change education in many ways and we’re on that journey as well. One is where AI fits into traditional education and the awarding of continuing medical education credits. Is an AI experience and AI-generated content sufficient and trustworthy, for example, to award continuing education or CME credits?

For students, can you adapt these technologies to support a more effective learning journey and a lifelong learning journey? Certainly AI has been applied for adaptive learning. We at Wolters Kluwer have had a lot of experience in this area, and there are opportunities there.

There’s also training around healthcare professionals being an effective consumer of information services. And particularly now, to understand the limitations of Gen AI and how its convincing and compelling answers can make us falsely believe that they are accurate when they clearly need more interrogation.

A final point is that there is an emerging literature about the degradation of learning from overreliance on Gen AI tools. There is some empirical data that reliance on Gen AI tools might lead to a decreased ability to retain and then to apply that knowledge in other settings. That’s a fundamental pedagogical change. Where this comes out and how educators will approach all this remains to be determined.

For the moment, clinicians at all levels, including trainees, are adopting Gen AI tools. It’s important that the tools that they are adopting to lead to their training and to patient care will be effective, safe, and reliable over an extended period of time.

What about AI governance?

Governance is important. It is tempting to use tools that are expedient. In fact, they are so compelling that there’s a tradeoff that I think clinicians are willing to take around expediency when they haven’t really taken a sharp look at what’s being traded off for accuracy, reliability, and some of the other dimensions of challenges related to the core technology.

The word that I’d like to get out is the emphasis on adequate governance. That can be by a third party, such as the electronic medical record vendor who is forwarding and embedding these tools, or the governance committees themselves at institutions. They need to be sure that all the tools that they are onboarding that are provider-facing, or that take advantage of advanced technologies, are properly vetted, scrutinized against important benchmarks, and transparent. If there are deficiencies, you  have the tools necessary to understand those deficiencies over time in domains like we operate such as decision support, where a right and wrong answer to an untrained eye or even to a trained eye can look equally good.

You need a gold standard to be sure that each answer is complete, accurate, and contemporary. That’s hard to do, but nonetheless, that’s the work that needs to be done to be sure that we’re helping all the healthcare professionals live up to their covenant and deliver the best possible care for their patients.

How do you choose a company strategy when AI and other technologies change literally every day?

Across Wolters Kluwer, we have a lot of experience with adopting advanced technologies. Across our verticals, we have already released more than 20 Gen AI related products and services. We are reinvesting constantly into advanced technologies and innovation, including AI, SaaS, blockchain, and other emerging technologies.

In the area of clinical decision support, such as what UpToDate provides, we have to really live up to our own standards in this high-stakes domain. There’s an evolving regulatory framework, but we understand our North Star. We understand in constructing this content that we are part of a medical community. We adhere to those standards. We have 55 physicians who work for UpToDate as deputy editors. Many of them are still in practice, mainly in academic medical centers. So the culture is one of patient safety, of seriousness, of understanding that there is a live patient somewhere behind all of our computer screens.

We have taken our time, as we have looked at the advances and particularly in Gen AI and how they can be applied, so that we adhere to our own standards and the standards that have been expected for our more than 3 million users out there. That means very, very careful product development and extensive testing. We’ve had a lot of innovation around ways to ensure reliability, accuracy, and validity, including not having the known pitfalls of Gen AI solutions like the degradation of context.

These things are very important. Generic Gen AI tools, for example, may recommend drugs that can be unsafe because they don’t ask contextual questions such as, is the patient pregnant? We have found examples of generic Gen AI tools that recommend drugs that are potentially perfectly suitable for the condition, but not if the patient is pregnant or they could be harmful to the fetus.

There have been many examples like that, so we have to understand the limitations of the technology and understand where the technology is going. We grounded it in this database that we have built over 30 years, which is not only summarizing the evidence, but infusing it with the clinical wisdom of deep experts drawn from a faculty around the world.

It’s our own commitment, our own standards, that are deferential to what is expected of us from our customers and the responsibility to take our time to test, release slowly, develop feedback mechanisms, and ground exclusively in UpToDate not the chaos of the internet, and in my view, create one of the most effective Gen AI solutions for decision support that currently exists.

HIStalk Interviews Dan Dodson, CEO, Fortified Health Security

Dan Dodson, MBA is CEO of Fortified Health Security.

Tell me about yourself and the company.

I have been in cybersecurity since about 2014, and in healthcare IT for 20 years. Fortified Health Security is a cybersecurity company that is focused exclusively on healthcare. We provide two kinds of services, advisory services and our managed security service provider business, or MSSP, for 24×7 monitoring and management of cyber technologies.

How does a healthcare-focused cybersecurity firm work differently than a more generalized company?

The attacks, adversaries, and the vectors they use are similar to other industries. The difference is how you respond to those threats and adversaries and risk reduction.

We believe strongly in having a knowledge base and an understanding of how healthcare organizations work, not only from a governance and regulatory perspective, but regarding infrastructure, legacy applications, mixed environments, EHRs, and medical devices. We build our playbooks and recommendations to take those elements into consideration. Our clients get more actionable intelligence so their teams can respond and take actions faster with the intelligence that is infused into our recommendations.

The top things organizations are trying to work through are AI, third-party risk, and training and awareness. Those three things are what organizations are talking mostly about with us.

What findings have surprised you in performing security risk assessments?

One surprise that we see is that everybody is at a different spot, and the weaknesses and the opportunities to improve are pretty vast. We’ve seen a lot of organizations make investments in different areas, some of which are reducing the risks that they set out to do. Sometimes they have opportunities for improvement. But as they’ve built their program over years, some areas tend to have significant gaps.

Third-party risk is a big area where organizations are struggling to tackle those challenges. Obviously with the rise of AI, we are in the early innings of understanding that from a risk perspective at the client side.

A lot of conversations are happening around end-user training and development. It’s a big challenge to actually drive better utilization of the tools to combat phishing, et cetera.

Are easily guessed or shared passwords still a big problem?

That certainly is still a challenge. The vast majority of compromises that could lead to a breach of data involve the end-user clicking on an email and giving up their credentials into a phishing email. Then the adversary comes in, moves laterally across the environment, and ultimately causes havoc. That’s still the number one entry point, so organizations are focused on combating that.

It seems like tools should have gotten sophisticated enough to block the clicking of suspicious links.

Tools are out there, and not having a tool would certainly increase your exposure. But this is an area where the adversaries are good. They are able to navigate around those tools and ultimately end in the inbox.

We see organizations thinking about how to reduce that attack surface. Do I have employees within the healthcare organization that maybe don’t need external email to execute their job? That’s a little bit of a culture challenge, because in the US, people associate their employment with having email. No one really talks about that. It’s the norm.

We are seeing some creative designs around that to make sure that we are limiting the attack surface. There are actually some cost benefits as well, such as fewer licenses for whichever email that you may use.

The other approach is training end users. A recent development is that most people are familiar with someone who was compromised personally for some type of phishing attack. Or, they have been impacted by breaches at Target or Nordstrom’s. One part of training is whether to focus more on the personal side and helping users understand how to protect themselves at the individual level. That would ultimately increase the level of protection for the organization.

What about users logging into their company email from personal devices?

That is still an issue. BYOD is prevalent. We have a lot of contract labor. If you live in a metro area, physicians have multiple privileges at multiple facilities. 

Who is winning the AI war between hackers and organizations?

I think the data would would tell us that the adversaries are being more successful. Breaches are continuing to occur. If you look at the Office for Civil Rights, the number of breaches year over year is stabilizing, but the impacts are getting larger. So I would say that, unfortunately, the adversaries are probably winning that fight. The adversaries are also using AI to launch more sophisticated attacks, both via email and help desk voice impersonations. They are definitely leveraging AI to hit us on all fronts.

How is the government’s role in healthcare cybersecurity changing?

Our view is that we are in a little bit of a standstill. There was a lot of energy at the end of the Biden administration. Senator Warner was leading that charge. Frameworks were put in place for programs that would provide clear expectations, along with some monetary support in a carrot-and-stick model to adopt said frameworks.

But a lot of that has stalled. The current view is that we may see tweaks to frameworks and expectations, but monetary support coming alongside that is probably off the table, at least in the near term.

Hackers have threatened to report their breach to HHS or have contacted individual health system executives, board members, media outlets, and even patients to threaten to expose breach information in hopes of getting a ransom payment. How do you address that dynamic, especially knowing that you wouldn’t be paying the most of honorable people with no recourse if they don’t deliver?

That’s the biggest challenge if you have a ransomware event or active breach that ends up in some type of negotiation. Thinking about adversarial intent, bad actors come after us to begin with because it’s monetary. They will pull all the strings that they can to create as much leverage against that organization to increase the likelihood of payment.

Also driving that behavior is class action lawsuits. Attorneys who used to chase car wrecks and malpractice cases have turned their eyes to cyberattack class action lawsuits. The adversaries know that, so they will weaponize that against the victim that is under attack. They will pull the strings on anything they can do to increase the likelihood of payment.

What are the advantages of organizations moving from point tools that are monitored by understaffed internal security groups to moving to a more centralized approach?

In most healthcare delivery organizations, teams are quite small. A lot of those individuals have been at that healthcare organization for a number of years and have made their way to the cybersecurity team. Health systems in general are not the best at training and having dollars available to train resources.

How do we make those individuals who have institutional knowledge about the networks, environment, and culture of the organization as effective as cyber warriors as possible?  We partner with those organizations to bring high-fidelity, actionable information to that team so that they can take quick and swift action.

As far as which service or what opportunity, I would just tell you that every healthcare organization is at a different point in their cybersecurity journey. They have made prior investments. Can our organization plug in, leverage existing investments, and operationalize that in a more efficient way to ultimately drive down risk?

One of your reports about downtime preparation quoted a chief nursing officer whose hospital experience an unanticipated problem because young nurses couldn’t read the cursive handwriting that doctors used to write paper orders. Is it common to find problems during downtime that weren’t anticipated in the plan?

Almost every time. Organizations do their best to prepare for downtimes that are short in duration. Hospitals go on diversion a lot for various reasons that have nothing to do with cybersecurity. They have downtime when they have to patch a system, implement a system, or upgrade a machine. We are relatively good at doing that for a short period of time. The challenge arises when you are down for a long duration and you don’t really know how to manage through days or weeks of not having access to the systems.

That’s driven by a couple of things. One, we are heavily reliant on systems when delivering care, whether that’s the EHR or the hundreds of other applications that power these health systems. So when they are down to some degree, the clinicians are frozen in their normal work habits. Anxiety and nervousness sets in because they want to take care of the patients, but they don’t have the technical controls in place to ensure that they provide swift, quality care. It slows down the care delivery model significantly.

Calculating is another issue we see. How am I calculating if I’m making an order for a particular medication? Med reconciliation is another thing that drives a lot of nervousness, making sure that I’m giving the right meds at the right dose to the right patient. Most of that at scale is done electronically, and that becomes an issue.

Communication is also another big challenge that we see. How are we communicating as a team if we’re using some type of a pager system or a walkie-talkie-system like Vocera and it’s down? That’s how we are used to communicating.

Lastly, a lot of the younger physicians have never operated in a world where they haven’t had technology. They were trained on an EHR at med school and they’ve been delivering care for years while being guided by electronic systems.

How do you advise organizations to deploy resources to protect their ever-increasing reliance on external technology vendors?

Step one is understanding how you interact with those third parties technically, so that if they have an event, you can take quick action to sever ties to limit the disruption to your organization from an adversarial perspective. But then comes the challenge that you need that system to deliver it, but the reality is that for the hundreds of systems that are that are in these healthcare delivery organizations, there’s not enough dollars to have backup systems for every single one of them. It’s unrealistic, both monetarily and operationally. That would also double your attack surface, so it’s not necessarily recommended. The first step is getting your arms around all of your third parties.

Step two is determining what the interaction is between your organization and those third parties. 

Step three is putting in some contractual language and some compensating controls on your side to try to limit the downtime.

Step four is that as you think about the disaster recovery plan, work with your clinical teams to understand how they would operate with certain critical systems down. Start with the ones that are most useful clinically and are most widespread so that you have some type of backup plan in place in the unlikely event that it’s unavailable.

What is the company’s strategy over the next few years?

Our strategy is to continue to work with healthcare organizations to increase their cybersecurity posture. We believe very strongly that a coordinated, programmatic approach through various elements of their cyber program can help minimize that risk. We are going to invest in our central command platform, which is our service delivery platform that provides actionable information and drives results across their entire organization to reduce risk.

HIStalk Interviews David Bates, CEO, Linus Health

David Bates, PhD is co-founder and CEO of Linus Health.

Tell me about yourself and the company.

I’m a scientist, engineer, and entrepreneur. I spent 16 years in venture capital. I’ve also started a few companies along the way. I’m passionate about brain health, its importance in the world, and how much suffering can be alleviated by preventative, proactive care, equipping people with agency over their brain health.

That speaks to the introduction of Linus Health. We have a brain health enablement and dementia prevention platform that spans the care continuum. It is clinically integrated, but it’s not just in the clinic. We want to care for people all the way through their life journey to optimize their brain health and prevent disability.

How does the diagnostic process differentiate between occasional forgetfulness versus true cognitive impairment?

The approach we take is unique, novel, and well validated. We mimic what specialists have done for decades by analyzing the process by which a person carries out a task. We leverage multiple sensors in a tablet, smartphone, or laptop to analyze the characteristics of their behavior. 

Behavior is the primary observable output of the brain. We deconstruct that behavior to understand any kind of emergent brain dysfunction. There are many kinds of brain dysfunction, not just a catch-all memory. There is executive function, language, visuo spatial, and of course memory. For each of those, it’s important to understand the type of impairment and what is likely giving rise.

The treatment, the intervention, can then be specific to work on that particular aspect of brain function, and as much as possible, improve health, improve function in their daily life, and equip them with the tools and capabilities to compensate for whatever disability is emerging.

That’s really important. It’s not just a binary thing, impaired or not impaired. We need to understand what is actually going on with this individual and how we can help them optimize their daily life and promote health in their brain.

What is the trigger for performing the test? Is it a one-time diagnosis event, or does regular screening have value?

I’m a big believer in preventative health. We should not wait for disease. We should be proactive in our brain health. It’s important to have a baseline. Everyone should be their own metric of how they’re doing, especially when it comes to brain health, so you can catch things early.

I believe in doing a brain health screen wherever healthcare is delivered. Wherever you would check blood pressure, you should check brain health. It’s important because with these new tools, you can catch things years before they would show up as symptoms. When you catch it early like that, you can intervene early.

It’s akin to oncology. You don’t want to wait until you have a tumor bulging out of your neck or they’re all over your body. You want to find, as early as possible, any kind of emergent illness and then intervene during that window of intervention to preserve function and health. With the brain, it’s incredibly important, because neurons that are lost are not recoverable.

To your point of should we screen? Absolutely. We need to change the way that we think about brain health. We need to understand that there is something that can be done. There is incredible hope for people. Many of them don’t have to get dementia. Up to 45% or more of cases can be prevented through lifestyle modification. Treatments are coming to market and getting approved by the FDA, so the earlier the right people can get started on them, the better the prognosis.

I’ve talked to people with Alzheimer’s disease who are living, in their words, their best life. They have known they’ve had it for a number of years. They are on one of these disease-modifying therapies and are still living their life, traveling, doing things with their kids.

Alzheimer’s disease is not dementia. It can result in dementia, but it can also be slowed down. Not every person with Alzheimer’s disease will get dementia. It’s important that we have this education and understanding in the market so there’s not a fatalistic view.

People are proactive in assessing their brain health and proactive in doing the lifestyle modifications. Those who need it will seek treatment to preserve brain function. With dementia, every single person can benefit from brain health assessment, especially with good platforms, even if they have dementia. Equipping care partners with knowledge and resources, equipping the individual. It’s a dyad, the two together that the patient and the care partner know how to optimize health and quality of life. Not only the life of the afflicted person, the patient, but the life of the care partner. If they’re not equipped, informed, and supported, their health suffers tremendously in the majority of cases.

That’s on the dementia side. Back to Alzheimer’s, detecting it early, even so-called pre-symptomatic. Platforms like Linus Health can find them before symptoms are apparent. Intervening and preserving function is incredibly important. People can live a much higher quality life than if there’s not intervention.

Intervention is not just “take this medication.” It’s holistic, and they need to be engaged in many ways, including for some the disease-modifying therapies. For everyone, it’s a discussion among themselves, their doctor and the family, and figuring out what is the best treatment course for this individual. It stands and it is irrefutable that the sooner you engage, the better the outcomes, and the better the quality of life for the individual and for the family.

Can that diagnosis and management be scaled by using non-specialist clinicians instead of less-available specialists?

With the right tooling, primary care can handle the majority of cases. They’re rightly positioned to. It is similar to how they now manage diabetes when it used to be an endocrinologist. Managing hypertension used to be a cardiologist. Those specialties are still needed and are used on the more difficult cases as their training warrants.

It’s using the right resources for the right people. The majority can be managed in primary care. Platforms like ours provide those primary care physicians with the capabilities to diagnose and the decision support to triage, to guide them to care for those individuals right there in the primary care setting. Especially addressing their modifiable risks, addressing reversible causes. That can all be done in primary care.

For cases that need specialist care, they should be referred right away. This helps streamline referrals, reduce wait times, and get people to the right resource as expeditiously as possible. That optimizes healthcare.

What is the role of biomarkers, which are in essence a blood test for dementia? How does that fit with cognitive testing and could they be applied to a population?

The approval of Fujirebio’s blood test, Lumipulse, back in May was a huge step forward in Alzheimer’s care. You need to first establish a cognitive impairment. Some of the key opinion leaders in the field have shown that with these blood biomarkers, they need the establishment of some kind of functional impairment to make their predictive accuracy appropriate and meaningful.

It’s a great addition to what I would call the emerging service line for brain healthcare, especially cognitive care. The blood biomarker helps, once you’ve established that there’s a concern with memory or thinking, what is the likely etiology? What is causing that impairment? If the p-tau blood biomarker is positive, it is very likely that the individual has Alzheimer’s disease. They should be triaged, if appropriate, to disease-modifying therapy or one of the many drugs and intervention methods that are coming out that can deal with that etiology. 

It’s equally important to know that it’s not Alzheimer’s disease, because you want to work on other contributed factors to find out what is giving rise. That could be a co-morbid condition like unmanaged hypertension, unmanaged diabetes, undiagnosed sleep apnea, and the many other things it could be. Get those things treated and then retest. For those who are appropriate, getting them to a neurospecialist since it could be Parkinson’s, Lewy bodies, or a variety of things.

That blood biomarker is incredibly important to know how to triage people following a cognitive assessment.

How do payers approach cognitive conditions?

Unfortunately, we’re seeing a number of payers that are not, in my opinion, assigning appropriate value to brain health. I don’t know if it’s broadly appreciated yet how important the brain is and how important it is to the quality of life and health of the individual. The brain is that organ that you can’t transplant. It’s important to who we are. I don’t know if the health system and the willingness of payers fully reflects the value of brain health and function. 

Members should demand more brain health focused resources to preserve their quality of life. Dementia is the number one health fear of middle-aged and older people, yet it’s not standard to assess cognition to try to prevent dementia. CMS does reimburse. There are CPT codes for the digital cognitive assessment. There are CPT codes for brain health visits. There’s a reimbursable pathway. 

With value-based care, CMS has done a great job, especially Medicare Advantage. They have risk adjustment factors. They have certain HEDIS exclusions. They have the incentives aligned with identifying and caring for cognitive issues and dementia care.

That’s at the CMS level. Different insurers take different stances on cognitive assessment. Maybe they don’t want to pay for the expensive disease-modifying therapies, so they don’t want to screen.They don’t see that the patient will be a member that long. Some of them don’t take an active role to support brain health care, and that’s really a shame. We need to do better than that.

You could say, “That’s because you have a company that is associated with finding emergent illness with the brain.” No, it’s because I’m a human being. I care a lot about brain health and I hate to see people suffering from it. If your primary goal is to insure people for health, you need to make sure that the things that are most important to their health are being looked after. If you do the right thing, everyone will win. It will pay off in the long run. The total cost of care will go down. People’s quality of life will go up, and there will be a lot less suffering.

Have studies looked at the age of onset and the insurer at the time of initial diagnosis? I’m wondering how much of that happens before people reach Medicare age.

The studies are emerging. It’s early days. There is the empirical evidence, which we need more of and I’m sure is underway. We are tracking people and we want to make sure that we have those cases.

There is precedent in all other chronic conditions that early intervention, early management, leads to reduce total cost of care. The best thing that we could do, and we need our system aligned, is to incentivize prevention. But our system is aligned to treat sickness, and so all of the incentives are around treating sickness. This is a policy matter, but how do we incentivize prevention? How do we enable people to take action over their brain health and reduce significantly the disease burden?

Neurological disorders are the number one disease burden in the world. Neurology in psychiatry is still an emerging area, especially with new tools and capabilities. But there’s a real opportunity here. Most health stems from the brain, and every other function is to support the brain. It’s good to start on first principles. How do we optimize brain health, and from that comes total health.

To your point on the insurers, I do appreciate that they have a business to run, and you can’t paint with broad brushes. Some of insurers are proactive in promoting brain health, even some of the largest ones. Some have taken a stance, while others have decided to put their head in the sand. That has never been a good strategy. You’re saving some dollars on the front end, but you’re causing irreparable harm to so many families.

How has last year’s acquisition of Together Senior Health changed your capabilities?

It has given us another step forward to enabling people to have a higher quality of life, even those who are living with disease, and their care partner. We’ve taken the RADAR tool (Risk of Alzheimer’s and Dementia Algorithm) to a validated capability that is now in production. We can risk stratify entire populations for the risk of undiagnosed disease. That enables insurers and health systems to know who is likely to be suffering from illness, the presence of emerging illness, and the risk of it. That’s where we should focus our care resources to minimize disease impact as much as possible.

You take it all the way through our platform and through the clinic. Then, back at home, how do we care for those individuals and optimize their brain health trajectory? The acquisition of Together Senior Health was a big step on those bookends, risk stratification on the front end and the engagement and care on the back end. That has helped complete our platform to span the continuum of care, all the way from identification, diagnosis, treatment planning, post-clinic engagement, and health coaching and monitoring.

How do you expect the company’s business to change over the next few years?

We are seeing the engagement of health systems across the country. They are leaning in increasingly. So I expect to see in the next two years that the standard of care will be set, and the service line will emerge for cognitive care. It will be such a good thing for the world, for the population. I see it spreading into Europe, the UK, throughout North America, Asia, and beyond. Get to Africa, get everywhere, and promote brain health.

We need to work on the stigma that is associated with dementia. We should not have a stigma. People should realize that it is a new day. There is incredible hope. The drugs that exist today on the market, and especially the ones that are coming, are tremendous. They are showing such benefit, and it’s still early. Everyone needs to know that there’s tremendous hope for the future.

We need to treasure the senior population. They are pillars of the community. They have so much life experience. We need to honor them by caring for their brain health and getting them the attention and resources that are needed to prevent dementia. That is super important.

HIStalk Interviews Kevin Healy, CEO, ReferWell

Kevin Healy is CEO of ReferWell.

Tell me about yourself and the company.

I started in the CEO position at ReferWell in April of this year. Before that, I worked with a private equity group called Chicago Pacific Founders, where I was an operating partner and CEO for one of the portfolio companies that I helped build and get moving. Prior to that, I was with Optum for 12 years, where I led the sales and growth teams in Optum Insight at one point, Optum Health at one point, and Optum Rx at one point. I had the pleasure of sitting within all three pillars of Optum. Before that, I had several startups, build-ups that led to successful exits.

It has been a whirlwind of healthcare over the last 25 years. Before that, I was in the golf business, so it was an transformation, obviously, from golf to healthcare [laughs]. It has been a wonderful experience.

I love this business. I love ReferWell. What attracted me to ReferWell and pulled me out of retirement was the product itself, which was incredible. But just as important were the people in the company. The team was amazing, dedicated, smart, young, ambitious, and ready to go conquer the world and fix the healthcare system. I loved the energy and the product itself.

ReferWell is a simple company. We do one thing, and we do one thing very, very well. That is, we get people to the doctor. Companies, health plans, and hospitals spend thousands if not millions of dollars to find out who they need to outreach to. Then they spend lots of money to find out what happened to those patients after the outreach. But very little, if any, is spent on actually getting them to the doctor.

I liken us to a light bulb. The light bulb is big and bright, but without the filament in between, it doesn’t work very well. ReferWell is the filament that connects the two entities. Who do we have to reach out to and engage, and what has happened with those individuals? The big part is about what happens to them when they get to the doctor. We do that better than anybody else.

What are the patient barriers to making and keeping appointments and following through on referrals?

It’s an overwhelming issue. Close to 40% of all appointments go missed. Sometimes when an individual is at the doctor, the doctor says, “Mrs. Jones, you need to go to see a cardiologist” and hands her a list of cardiologists to call. Or you have to go see multiple doctors, so you get a list of multiple doctors to call. Sometimes they happen, but 40% of the time, they don’t happen. That’s a barrier of understanding who to call, not being able to reach them, not being able to find an appointment, or not having transportation to get there. Also, not really understanding why they need to go.

Part of our unique positioning in the marketplace is not only about technology, but about people. We believe that people, our care navigators, are an integral part of this process. It’s not just using an AI or an AI agent. We have real people talking to real people about real problems and real issues. That’s extremely important. We can never take that human touch out of health care, and we never want to. But we can support it with advanced technology.

Our proprietary scheduling platform allows us to schedule with our care navigators, or for a provider or a payer to schedule, at the time of communication, or at the touch point with the patient or the member of the health plan to be able to set an appointment at the time of engagement. It sounds so simple, and the amazing part is that it is a simple idea that is hard to execute. That’s where ReferWell comes in.

How can patients be helped to choose a specialist from that list that their doctor has provided?

The unique factor is that the federal government has been kind enough to rate health plans, and health plans have been kind enough to rate physicians and practices for quality of service and quality of outcome. It’s kind of a cost-quality equation that health plans, for example, apply to doctors. They a four- or five-star rating, just like health plans have a five-star rating. 

It allows us to filter based upon location, so the closest to the office of the physician that they’re visiting or closest to their home or their place of work. Then also by quality. The highest-quality cardiologist within a mile from my house, or two miles from my house, and these are the doctors that have available appointments in the next week or two weeks. Quality and location filtering has to be taken into account.

We have multiple sources of information that we absorb, so we can triangulate that information into who would be the best for this individual to go to. It gives them options.

We’re not making the clinical decisions. We’re letting the patient, the doctor’s office, or the health plan help them with those clinical decisions. We’re just offering up the information and telling them what’s available in terms of spots, schedule appointments, etc.

Provider directories have always been a challenge to maintain, so the patient calls down the list and finds doctors whose practice information and insurance acceptance isn’t current. Can that be automated, or does it always end up with someone making a phone call?

The answer is not as simple as one might think, but the progress has been significant. The feds have been all over provider data management, the information that is available to members of health plans of all types, Medicare Advantage, Medicaid, commercial, ACA lives, etc. Several organizations are out there that maintain correct information, and we contract with three of them.

We then have to decide which of the information is most accurate, and sometimes our team needs to make an outreach to find out which is correct if we have conflicting information. But that’s our job. That’s what we offer. Then we update the systems so that everybody has the correct information.

Provider data management has come a long way. It’s not perfect, but on the health plan side, there are fines for not having your provider data management up to date. We use some of the same companies that they use to inform our decision-making process.

It’s impossible, really. Doctors work in several offices, different times of day, different days of the week, different days of the month, across multiple communities, and all of that changes. It’s hard to have it 100% correct, but technology is helping more with that. The groups that we partner with are very, very good at keeping up their data, and that helps inform our decision-making as well.

Aligned incentives would occur if providers benefitted from keeping their schedules full, but if they are employed, they may not see the value of being busier, or maybe their schedule’s already full so they don’t really care. How do practices view the idea of having the schedule availability their providers visible outside?

We look at it in a way that may be a bit hopeful, but I think that most physicians want to give great service to the people that they can provide service to. There is a sense of control with having their own schedules and opening it up seems like a little bit of a loss of control. But they also know that they have the opportunity to serve and work with more individuals. It always behooves them to keep their schedules full, even on the employed side, because they are incentivized to do so and they intuitively want to.

Getting access to schedules for providers has been a difficult process. One of the reasons is that we have so many different electronic health records out there. We have to integrate with them so that we can see what’s open and what schedules are available.

It’s difficult to get doctors to agree to allow people to see their schedules. But as more and more groups start to look at accountable care organizations and are going at risk for the care, care management, and the health and wellbeing of their patients, they are incentivized through financial rewards if they provide good service and have great quality outcomes. They are raising their star levels and want to get that word out there that they are a high-quality care center and can be counted on to have access and will provide quality care. 

It is an opportunity and a change of a mindset for providers as much as anything else, a little bit of relinquishing control. But for example, UnitedHealthcare has a gold card program that stack ranks people by quality of outcomes, and with that comes rewards. As part of that, I can imagine a day when they have ReferWell as their scheduler, and tell providers that we have to have access to your schedules to get the gold card program, either complete access or partial access via a ReferWell platform. Other health plans have the same type of program as well.

Our North Star at ReferWell, that Holy Grail for us, is having organizations recognize that engaging the providers and rewarding the providers for good behavior and good outcomes means that they will have a great partner. It changes the healthcare structure from fragmented to more of a synthesized, hospitality-like structure.

That’s where we’ve lost faith in our healthcare system. It’s fragmented. They don’t talk to each other. All of a sudden we can start talking to each other. Providers, payers, and hospitals are all talking to each other via scheduling mechanism.

It seems like a benign way to do this, but it’s amazing what it would mean in terms of how easy it would be to synthesize all three entities into a much more cohesive care management journey for an individual who has just left the hospital. How do I make my next appointment? Who do I make my next appointment with? Does my health plan know that I have my next appointment with them? That’s the dynamic that has to change. 

Is it more common that the clinician who makes the referral knows that the visit actually happened, or wants to know, or what its result was?

It is becoming more common for them to want to know. But it’s also more common that the patient understands that their primary care physician knows that they had an appointment at another facility or doctor, and maybe even what the outcome was of that particular appointment. It has always seemed odd to me that if something would happen to me and I end up in the hospital, my primary care physician, who has been looking over me for many, many years, has no idea that I’m in the hospital, because I’m not able to talk to them. Connectivity is needed that has not existed in the past. 

That becomes a comforting factor for a patient to understand that their doctor knows that these things have happened. They know what meds I’m on, no matter what doctor I go to. They know what services I’ve had. It makes it feel like the whole system is talking with each other. They can schedule the appointments and have the data from that appointment in their electronic medical record. They can talk to me about what transpired and how I feel since then. Or get in a better understanding about my overall health and wellbeing.

The patient is going to drive this. The patient is going to want their providers to have this information, and they will have a better overall experience when they do. Imagine if you called Marriott’s hotel reservation line to ask if they have a hotel in New York City. They say, “Yes we do, thanks for calling” and they hang up. Wait a minute, I’d like to make a reservation. That’s how healthcare is. You need to make the appointment and get scheduled. It has been barrier after barrier to do so. It’s a less cohesive an experience when they don’t have the information at the provider level. That just needs to end.

We aren’t saying that we are curing all the issues with healthcare, only that an integral part that has never existed is this connection point of getting people to the doctor, finding out what happened while they were there, and then providing the referring doctor with the information about what took place at the visit. It doesn’t seem like that big of a deal, but it is a meaningful and impactful overall consumer experience that is going to change.

How do you expect your business to be affected by changes in federal policy or in federal payment policies?

The government is doing a very good job in several areas. I know that’s not a popular statement to make. But when it comes to healthcare, the government is trying to drive hospitals and doctors towards this accountable care model. The accountable care model is all about payments, all about follow the money. But what they’re trying to do is follow the money, but also follow the outcomes. For us, that changes the dynamic of what’s required. The federal government is forcing our healthcare system to go in that direction. It has been tried and tried and tried. For provider practices with the ACO model, hospitals are getting in with the CMS TEAM model — Transforming Episode Accountability Model — under five different categories of care. 

It’s hard to say that the government has a heart sometimes [laughs], but they have a heart, it’s in the right place, and their heads are in the right place. We just have to put it in action, and I think we’re on the right path.

My six months at ReferWell has not disappointed. It’s such a wonderful little organization and been around for 10 years. It is finding its feet right now and I’m excited to be part of it.

HIStalk Interviews Steve Cagle, Board Advisor, Clearwater

Steve Cagle, MBA was CEO of Clearwater at the time of this interview. He transitioned to board advisor on September 30.

Tell me about yourself and the company.

Clearwater is a healthcare-focused solutions firm that provides cybersecurity compliance and managed security services to hospitals, health systems, physician practice management groups, digital health, and health IT companies. Really all types of organizations in the healthcare ecosystem. We help those organizations to be more secure, be more compliant, and be more resilient so that they can achieve their missions.

I’ve been CEO of Clearwater since May 2018. My background is in healthcare. I started my career in a software company that provided quality management software to help pharmaceutical companies comply with FDA regulations, such as good manufacturing practices. I then spent some time in the pharma industry in consumer healthcare products, running a business before returning back to technology and compliance here at Clearwater.

How do health systems decide how much effort and money to invest in cybersecurity?

Unfortunately in healthcare, most organizations have been historically underinvested in cybersecurity. However, we have seen over the last five years or so an increased focus, especially following the pandemic, when we saw a wave of ransomware attacks on healthcare organizations. Then we had the Change Healthcare incident a year and a half ago, which affected about 70% of the providers and caused very extensive damage.

As healthcare organizations have continued to adopt new technology, technology has become critical to operating their businesses or providing care to patients. They have realized that cybersecurity mission critical and requires them to have the appropriate protections in place to reduce risks.

That’s really the key word. It’s about understanding your organization’s risks beyond the high level. A lot of organizations have done high-level risk assessments. They may be helpful as a starting point. But we need to go much deeper in today’s environment, where attack techniques have evolved to become difficult to defend and protect against.

Organizations have had significant impacts from ransomware attacks and breaches. That’s why the Office for Civil Rights of HHS, which enforces HIPAA regulations, has been focused on risk analysis and their risk analysis initiative. Risk analysis in healthcare requires that organizations understand where they have electronic protected health information, where they have those critical systems that support their operations or are connected to those systems with EPHI, and that they evaluate the vulnerabilities and threats, assess the controls that are in place, and determine the level of risk that exists with each system.

By doing that, organizations will be better informed as to where those high risks are. Based on their risk threshold, they can then identify those risks that fall above that threshold and put specific risk remediation or risk management plans in place to address those risks.

That’s a business-focused way of approaching cybersecurity. It’s not checking boxes. It’s not trying to have the best security program in the world. It’s really understanding your risk at a level that is appropriate. Then, taking actions to bring those risks to an acceptable level.

What were the most important lessons learned from the Change Healthcare incident?

Risk analysis. Clearly there’s been a lot of uptick in organizations really understanding, “I need to get to that next level. I’ve been doing the same type of assessment for many years. I’m going to invest more money into doing that risk analysis so that I can have better information about my security program.“

We’re seeing a lot of attention on cybersecurity and risk from the board of directors and the executive teams. From a cultural perspective, there has been a change in healthcare where this has become a priority that organizations need to focus on.

We’ve seen big changes in resiliency, where organizations have plans in place to not only respond to a security incident, but also to contain it to operate under duress through a business continuity plan. Having updated disaster recovery plans and testing those to make sure that they are effective.

As we look at all the solutions out there that are based on artificial intelligence, we have new concerns. There was a big rush to implement a lot of these new technologies that are based on AI. Unfortunately, many organizations did not take the time to establish policies and procedures about how they will use them and to assess the risks around these technologies. 

It is still risk analysis, but it’s a different set of risks and different set of controls. We are seeing a lot of interest from our clients in helping them to establish governance around artificial intelligence, cybersecurity, and privacy, or to assess their risks of those programs and to help make sure that they are implementing these technologies in a responsible way.

The mainstream press loves headlines about the devastating impact to patients of a local provider that has gone down from a cyberattack. How much do we not hear about providers who are successful in preventing that kind of attack?

That’s a very important point that you’re making. We hear about the bad news, but we don’t hear about the good things that are happening.

We’ve done over 650 NIST Cybersecurity Framework assessments for our clients over the last 10 years. We track and trend maturity levels over time. We see that the industry is becoming more mature. We track over time the organizations that adopt the NIST Cybersecurity Framework, which is a commonly accepted and used framework in healthcare, and we see that they are improving above the bar of the rest of the industry. There’s really good data that we can point to that demonstrates that we are making progress.

The challenge is that the bar keeps getting higher. You have more vulnerabilities, more threat actors. Threat actors have been very successful in obtaining ransomware payments from healthcare. They pay more often than any other industry. When it’s easier to attack a certain sector that is more willing to pay and pay more, that’s going to attract more threat actors.

You don’t hear about organizations that are being responsible. They are assessing risks, maturing their security programs, and not having those attacks. Or if they do have a security incident, they are able to address it quickly and with minimal impact. They have network segmentation and other types of controls in place that make it difficult for threat actors to exfiltrate the data or to do damage.

We will continue to see that maturity improve over time. But we have to realize that unless we stop developing and implementing new technologies and increasing the attack surface, it’s not going to stand still. The bar is always going to become higher.

How often do providers pay a ransom, and if they do, what is a typical outcome?

Fewer providers are paying than in the past. A few years ago, it was 67% of the time, and that number has gone down probably closer to 50%.

You really can’t trust criminals. A lot of them will try to uphold their end of the bargain because they want people to continue paying, but that’s not always the case.

There’s also double extortion. You get the encryption keys to unlock your systems. Maybe some of these organizations have good backups in place and are willing to take the downtime that it takes to restore those systems, which could take days or weeks, or longer. In some cases, those encryption keys do not work. They’ve done so much damage that it doesn’t really help them.

Then the second extortion is to get the data back. Often the data will end up somewhere else in the future. Paying the ransom doesn’t give you any guarantees. You’re really taking your chances. That’s why you are seeing fewer organizations making that payment.

How do organizations allocate their spending across prevention, detection, and rapid recovery?

We always recommend starting with a baseline set of controls and adopting industry standard best practices. We can point to the NIST Cybersecurity Framework. We can also point to the 405(d) health industry cybersecurity practices. Those are both recognized security practices in healthcare based on an amendment to the HITECH Act in January 2021.

The 405(d) HICP is a great place to start because it is provided in different volumes for small, medium, and large organizations. It was developed through collaboration with over 600 firms in healthcare — providers, vendors, and the government. It’s a practical way of setting up those baseline controls. 

Once you’ve picked a framework and standard, you go back to how much more you need beyond that. That comes down to the other requirements that you have. Do you have compliance requirements that you need to meet? Maybe even ones outside of HIPAA. Do you have clients, partners, or payers that require you to meet certain security standards, maybe a SOC 2 audit or HITRUST certification? What’s your risk profile? What kind of risk as an organization are you willing to accept?

Then you do that risk analysis to see where you have gaps between your current level of risk and what’s acceptable. Using all that information, we create a target profile. It’s a long-term roadmap of where we want to focus. That will help determine where to make those additional investments. We know the minimum requirements for standards and practices, but going beyond that, what is the organization’s specific situation? 

What is the value of health systems communicating regularly with their boards about cybersecurity, and what metrics are most useful for board members to understand the situation?

We speak to a lot more boards now than we did maybe five years ago. It’s pretty frequent. One of the key functions of a board is risk management. If the board is being informed of the other types of risks across the organization, cybersecurity has become an important area of risk, and one that they need to be informed about.

Typical things that we will talk to boards about are trends, particularly across the sector, and the higher-level concerns or risks that they need to think about.  

The board should be putting the governance in place. What higher-level policies do we want to have as an organization? What is the level of risk we are willing to accept?

Sometimes, but not as much any more, we see risk tolerance levels being set by more at the operating level, the IT department. The IT department is not the risk owner. If a security incident renders a hospital in a position where it can’t see patients, that’s a board level issue. That’s all the way up to the board. So the board needs to decide how much risk we are willing to take. How many resources are we willing to apply? And then put the management team to work with the mandate and the support to implement a program that will ensure that the organization is in line with those policies and is on a path to meet that risk threshold.

We have to keep in mind that risk changes over time. Just because we are below our risk threshold today doesn’t mean that tomorrow we’re not. We do M&A, acquire a new part of the business, partner with somebody else that includes new third-party risk, changing the threat landscape. It’s constantly changing, so the board needs to make sure that that risk management program is prioritized and resourced. Then getting information to know that it’s actually being executed appropriately.

What changes do you expect to see in HHS OCR’s enforcement of HIPAA and security?

The Office for Civil Rights has been focused a lot this year on its risk analysis initiative, where it’s making sure that organizations are prioritizing that risk analysis that I spoke about earlier. The notice of proposed rulemaking was released at the beginning of the year. Part of that rule contains updates to the risk analysis requirement that reflect its current enforcement actions and guidance.

A lot of other requirements are more specific and are required under the rule. I don’t think that rule in its current form will necessarily be the one that is eventually published. I do think, however, there will be an update to the rule or at least some additional standards that organizations will need to meet. The HIPAA security rule was last updated in 2013. The world has changed a lot since that time.

Most of the industry is looking for something specific we can point to, not overwhelming, but addressable. Ideally with some support and help from the government, especially for those smaller organizations or rural health organizations that don’t have the resources or the money to improve the programs the way that they would like.

What does the company’s strategy look like over the next 3-4 years?

Our strategy is to be a market leader in healthcare cybersecurity and compliance. To do that, we need to have a full set of capabilities that are relevant to healthcare organizations. Not just today, but over the next several years. Our strategy is to continue to ensure that we can provide those services to our clients in a way that helps them reduce costs, become more efficient, and focus more on their mission, whether it’s treating patients or driving their business. Being a partner and extension of the organization to help them address cybersecurity compliance.

We are excited about our growth at Clearwater. We are grateful to have dedicated professionals in the organization, as well as a growing list of clients that we collaborate closely with. We are dedicated to this industry and looking forward to continuing to serve this industry and help make a difference in healthcare.

We are thrilled to announce a growth investment from Sunstone Partners, which is a private equity firm that focuses on tech-enabled services with a particular focus in cybersecurity and healthcare. That makes them a great partner for Clearwater going forward. We are excited to have a great partner that can help us better serve our clients. We will be investing in more technology, as well as continuing to scale the organization.

HIStalk Interviews Michael Raymer, CEO, Vitalchat

Michael Raymer is CEO of Vitalchat.

Tell me about yourself and the company.

First, I just want to thank you for saving a lot of lives during COVID. Your post about Co-Vents, the not-for-profit that I set up to get refurbished ventilators into the field early in the pandemic, allowed us to connect with the Stryker family, which funded our efforts. We estimate that we saved 20,000 lives, and that all started at HIStalk.

I think you call me the timeless veteran of healthcare. I’ve been involved in large companies driving big innovations, such as GE Healthcare and Microsoft. I spent the early days of my career working on ventilators at Nellcor Puritan Bennett. Since I closed the days of Microsoft, I’ve been focused more on earlier-stage companies. I sold my last business, Pro-ficiency, to a publicly-traded company last summer and joined Vitalchat in October of last year.

Vitalchat is a flexible, AI-driven audio and video platform that enables nurses to practice virtually, and supports procedural telehealth. I joined because video is probably the most underutilized signal in healthcare. We have all sorts of data, more data than we can use. But when you combine video with data, it can enhance the care delivery process.

I’ll also mention that I write a weekly blog post on LinkedIn for those who would like to follow my thoughts about healthcare, strategy, business, and people.

How does a health system use ambient, AI-driven monitoring to enhance nursing capacity?

We have automated nearly 70 workflows. Our implementation teams identify the workflows that are the easiest to implement with the greatest ROI, which may vary by health system. Common uses could be the safety check where you match the IV bag to the patient, which typically involves two nurses being in the room. When you use virtual technology, you can have the nurse at the bedside complemented by the virtual nurse remotely, who is able to zoom in on the IV bag and also the patient ID band to make sure right patient, right medication, right time. That’s just one of 70 workflows that we support with health systems today. 

One of our key accounts is University Hospitals in Cleveland. Their first 150-bed pilot cost them $1 million to implement and they generated a $10 million savings over the first year. Now they’re on the pathway to roll it out across the entire healthcare system. Every day they are encountering a new potential use case for the product. That’s what’s so exciting about bringing in under-penetrated technology and to see all the possibilities for improving care for patients. That’s what it’s all about.

Can the technology bring nurses back into the workforce who left due to physical or geographic limitations or frustrations with floor nursing?

Absolutely. We see that the most successful virtual nurse is one who came from the floor and knows some of the other caregivers on that floor.  But most importantly, they’ve seen almost everything that could happen in the care of patients. They are a steadying hand supporting the nurse on the floor. 

We have heard many stories in our customer base about nurses who likely would have retired or pursued a different career, but took advantage of virtual nursing to continue to provide their unique skill set to health systems. That’s very gratifying. We have a projected shortage of almost a million nurses by the year 2030. If this can help stem the outflow of nurses outside of the hospital, that’s a great thing. We are seeing lower turnover rates and higher nurse satisfaction with having our virtual platform in place.

Do hospitals usually rotate floor staff, or are the virtual nurses people who don’t work regularly in the facility?

I don’t want to generalize, but these are not roles that are being filled by traveling nurses. They are clinicians who have worked in the healthcare system and are well trusted. Some of our sites have a rotation, where they may be three weeks in the virtual care operations center and another week on the floor. That keeps the connection with the clinician on the floor and allows them to understand how they could even be more effective in the virtual command center. That’s a good model for those who are physically able to go back to the floor.

It seems that skilled nursing facilities would be a good use case, although they have limited financial and technological capabilities.

I spent time in my career in the skilled nursing space, which is certainly economically challenged. A variety of use cases can be positive, including medication administration. You typically don’t have an onsite pharmacist, so the ability to have a remote pharmacist interacting with the patient and the physician onsite. You have issues of patients who are worried about falls, such as a movement in the room that might indicate that the patient is trying to get out of bed when they have been deemed a fall risk.

There is significant opportunity in skilled nursing. The issue is the price point. Because we’re agnostic to the hardware platform, we believe there will be an opening for us ultimately in the skilled nursing space, but we’ll probably follow that through the channel of our customers as hospitals. In fact, this week we had a great discussion with one of our customers going out into the home with that subsidiary. We can scale the hardware appropriate to the particular situation, so it could be a great fit.

The Big, Beautiful Bill will allow more money to be funneled into the rural healthcare system, where 25%-plus of the patients in this country are taken care of. The bill allocated $50 billion to technology like Vitalchat provides. Being able to allow that patient to remain in that remote facility supported by a specialist in the tertiary care academic centers was brilliant in the bill. The patient wins. The remote facility wins, because they’re able to keep that patient. The patient’s family wins, because they aren’t having to travel long distances in support of a loved one.

At the same time, the healthcare system can get reimbursed for the specialist care without them physically having to be on site. If necessary, the patient can be transitioned to that academic medical center. Fully implemented, the vision of that is particularly powerful, and you can extend that analogy to skilled nursing.

Is your technology’s footprint light enough to avoid a rip-and-replace of existing technical and physical infrastructure?

One of the technology constraints is bandwidth consumption. We have patents in video compression that allow us to be a very quiet signal on the hospital network. We move the AI to the edge and do not use cloud resources to process that video. Response time is outstanding. We wind up not being traffic on the network that’s going up to the cloud to implement AI or ambient learning from that video signal. Reusing hardware that’s already in the facility.

We recently created a solution for customers of ProConnections, a tele-ICU vendor that closed late last year and left their installed base abandoned. Because our platform is flexible and hardware agnostic, we created a solution where we provide software on their existing hardware platform so their customers can continue to use that product in monitoring patients.

What opportunities does AI offer in analyzing video without requiring human eyes?

The future of care is ambient. You see a lot of announcements in the ambient speech space. It will be more and more assistive over time. We were talking with one of our customers today about a dietary use case. Looking at the plate that’s delivered versus what it looks like when returned, then being able to assist nursing with caloric calculations and the I&O workflow sheets that nurses manually fill out today.

There is such tremendous opportunity to look at workflow as the technology improves. The AI engine, in our case, is easily trainable. I can’t even conceive of all the use cases. Today the dietary one came up and I never even thought of that as a potential use case in the acute care setting. It’s a big time-consumer for nursing. It would give them more time to spend on patient care and not documentation.

Today, it’s 70 use cases. A year from now, it will be double that number that will be proven to add value.

Are the companies working in this area of video analysis specific to healthcare? Also, how do you sort through those many use cases and decide which ones to go after first?

The platform that we built is easily trainable. In fact, we began using NLP as a test case for allowing a clinician to build the kinds of things they want to observe in the room. Our ambient AI engine will process requests by the clinician and provide them immediate feedback.  There’s not a lot of software work that we have to do to enable new workflows. 

A health system has super complex workflows. If you are shifting something from an in-room activity to virtual-based, everyone on the care team needs to understand that. There has to be the appropriate accountability, both for the nurse in the room and the centralized command center.

Most of those workflow challenges are not software issues for us. They are workflow challenges for the health system that are not unlike originally implementing the modern EHR. A tremendous amount of workflow design was done. The beauty of our product is we don’t have to build flowsheets. We don’t have to build order sets. We are insulated from the inherent complexity of an EHR.

We don’t see a lot of big companies in the video space. We see companies much like Vitalchat. A lot of people saw the headlines last year in Stryker’s acquisition of Care.ai and the vision of merging the device company with visual insights. We believe that being Switzerland is a better strategy, not being tied to just one vendor. It allows us to work with Epic, Oracle Cerner and devices in the patient room and in the OR.

The space is interesting. It is not just the AI ambient technology, but also the wrapper around it. The complexity of running video and audio from a patient room is non-trivial. We have been able to bulletproof our platform. I call it a self-healing technology, where our uptime is up from three nines to four nines now for our in-room cameras. If you’re going to rely upon this technology 7×24, it has to have that amazing reliability. That may have been the most difficult challenge technologically, getting the uptime to be there that is necessary to deploy at scale.

We started out talking about COVID and Co-Vents, which brought back unpleasant memories of patients dying in locked-down hospitals with families having their last moments with their loved ones via an IPad. Can technology like yours improve the experience of patients and families and not just that of clinicians?

That is a very personal issue for my wife and me. We lost my mother-in-law during the dark days of COVID. They would not allow my wife to come in the hospital to see her. There was not any kind of technology that allowed us to even have an interaction with her. Fast forward to today, that has softened the landing of this technology in patient rooms, because patients understand video consults now. FaceTiming with family members and similar tools are mainstream. 

I heard a great story two weeks ago where one of our customers was able to have a patient in the hospital be a participant in a wedding. Not just viewing it, but actually having two-way audio communication with family members in the wedding while the patient was in the hospital. We hear story after story where our technology has allowed that patient to connect not just to caregivers in a health system, but also loved ones, and to have loved ones join consults at the bedside from remote across the US or across the world.

What are your plans for the company over the next two or three years?

My primary job as CEO is making sure that the rate we’re growing will allow us to have both the monetary and people capital that are required to support our customers. The good news is that I have a very big Rolodex. When we have a need inside the business, it’s easy for me to reach out to my network and plug individuals in immediately who can make a difference for our customers.

This is an exciting space. A strategist would call the inpatient virtual nursing setting a $2 billion total addressable market. You have less than 5% of the beds in the US that are penetrated with video and audio. That’s a significant opportunity for us to make a difference. It’s exciting to be early and exciting to have unique capabilities in our platform that will allow us to sustain impressive growth over time.

HIStalk Interviews Thomas Thatapudi, CIO, AGS Health

Thomas Thatapudi, MBA is CIO of AGS Health.

Tell me about yourself and the company.

We are primarily a revenue cycle management company. We work with pretty large enterprises, such as Mayo Clinic, Cleveland Clinic, and Baylor Scott & White Health. We offer services on the front end, which is usually scheduling, patient access-related functions, mid-cycle coding, and in the back end, AR and denials.  We are about 15,000 people. We are what I would call a tech-first services company.

My career over the past 20-odd years has been primarily in technology. I’ve been focused on building data-intensive apps. In the last two or three years, I’ve been pretty intrigued with AI and its applications.

What are the biggest pain points in RCM that technology may help solve?

I’ve been working on the provider side only since I’ve started working with AGS Health in the last four years, so I’ve seen a fair bit of insurance. When I say insurance, that’s  auto, home, and health insurance. I’ve worked a lot with payers, I’ve worked with credit card companies.

My take on providers is that healthcare has always been a laggard in terms of adoption of technology, and more recently, AI. Providers, even more. Even within providers, revenue cycle management is probably at the bottom of the totem pole when it comes to infusion of either the technology or capital that is required for technology.

RCM is primarily a labor-intensive enterprise. Because there are no unlimited resources for providers, it means that we need another toolset, or at least part of the toolset has to be technology or AI, to address some of the issues.

For example, the denial rates have only been going up in the last two to four years. The payers are making denials more complex. There is no way that providers can throw unlimited resources at it, and neither can the RCM providers like AGS Health. Therefore, each and every portion of the RCM life cycle, from when the patient has completed his or her interaction at the point of care to when the interaction is closed, whether it’s collected, denied, partially collected, or whatever. Through that whole function, it is important that there is some tech infusion happening, or else some of these things will fall through the cracks because there are only a limited number of humans that you can throw at some of these problems.

Are providers thinking about technology and AI for immediate cost reduction or revenue enhancement, are they looking at it strategically, or both?

I see a combination of both, at least in the last 12 months that I have been talking to customers. Last week I was with a chief revenue cycle officer who was progressive and wanted to get ahead of the curve in terms of adoption of AI. The reason is that the CFO comes back and says, can you squeeze more dollars from this? Can you do this? Instead of spending 7 cents collecting a dollar, can you do this using 4.5 cents? The bottom line always has to be, can I collect the dollars faster and more economically?

Others don’t want to miss the AI boom, so they make all the right noises, but actually don’t know how to wrangle with AI. You see both ends of the spectrum here.

How do RCM and consumerism intersect from a technology standpoint?

I’ll take something very simple. A patient needs to get a scan and the prior auth has been denied. Therefore, all it requires is informing the patient that his or her medical procedure has been denied and they need to go back to the clinician for an alternative clinical pathway. The question is, how exactly do you reach the patient to be able to inform them?

One of our customers has 50-odd people sitting in some town in Wisconsin making these calls. But half the time, nobody’s picking up those calls, because they don’t recognize the number.  You cannot even inform them that their procedure has been denied. If you leave them a voicemail or a message, it almost always triggers a call back into the contact center saying, “You left me a message. I have no clue. What am I supposed to do?”

These are patients who most probably have been waiting for that particular procedure for a long time. How do you actually reach out to the patient and make sure that their whole interaction with the healthcare system — getting the procedure done, making sure that they know how much they’re paying, making sure that their schedule is on time, and getting the right approvals from the payers — how do you make that interaction more seamless without making it burdensome? It’s a gnarly problem even now.

With mobile applications since 2010 and people being on social media and attuned to how they work on social media, we would have assumed that by 2025, some of these problems would have been more elegantly solved, but that doesn’t seem to be the case. This is an ongoing problem, so there’s a lot more opportunities than what it might seem.

How will healthcare use agentic AI? Is it too early to ask people if they are seeing results?

There’s been a lot of buzz about agentic AI, especially because of OpenAI and others. The VC-funded firms have been hyping up that word quite a bit. My own hypothesis is that it won’t solve world hunger, where all the humans disappear and there are just AI agents doing everything.  But it also doesn’t mean that the world will remain what it is. There will be some changes on that front.

With payers, when there is pressure in terms of claim loss and medical loss ratios going up, the first thing that they always go after is the provider contact center. One of the largest payers that I worked for had about 12,000 people in the contact center, with 7,000 of them addressing members and 5,000 working in the provider contact center. If the claim loss ratios are going up, the first thing that the CFO does is cut the number of people handling the provider contact center because as you can imagine, they’re not dying to answer questions about, where is my bill or is my prior auth approved? 

As I’ve talked to CTOs and CIOs on the payer side, they would like to deploy agentic AI to answer some of these provider questions. If it’s not already there, we should expect in the next 12 to 24 months that the payers will start fielding some of these agentic AI to answer questions either, if not to the members, at least to the provider community.

My own interaction with AI agents has been interesting. I suffered a home claim loss. I had to call on a Saturday because that’s when it happened. The insurance carrier was shut down, so they had a TPA taking that first notice of loss. It was an unpleasant interaction. It was almost like the lady was like, “How dare you have a claim loss on a Saturday?” I got the claim number, so the first thing that I did on Monday morning was to call them back to make sure that it was logged correctly.

For the first six or seven minutes, it was a very pleasant interaction. The other person was empathetic, saying all the right words, making sure that we were doing well, blah, blah. It took me a good eight or nine minutes to figure out that I was talking to an AI agent. Lo and behold, it was a good interaction. I got my details. I knew who I had to call as my next steps. I knew what to expect.

My assumption is that as the AI agents cross the uncanny valley of completely being unrecognizable as AI agents, patients and even payer contact centers might actually be comfortable talking to these AI agents. Going back to my example of calling up the patient to tell them that their prior auth has been denied, and they need to go back to the clinician. In my mind, there is no reason to do this using a human. We are piloting an AI agent to make these calls as we speak. 

We will start scratching the surface in terms of how many of these interactions can be done by AI agents versus humans. It’s a matter of time when it will happen, not whether it will happen.

Will companies treat AI agents as a feature, not something to hide, because many people would rather not talk to an actual human?

I was reading an article that the Gen Z’ers apparently don’t like calling at all. If they know they aren’t calling a human, they will be more open to calling.

I was at AWS last year and the CTO of Rocket Mortgage was presenting. He made an interesting observation that their mortgage conversion ratios are 3x when person who might take a loan talks to an AI agent rather than a human. There’s more empathy and understanding.

It will be an interesting phenomenon. My own assumption is that we as humans will most probably get attuned to it. When we are booking travel or ordering food on Uber Eats, many of our interactions will most probably be with AI agents. These AI agents in healthcare may not be such a curveball to patients or members. They might actually welcome it versus talking to a human.

How do you program AI to use the human knowledge, judgment, and intuition that a good employee develops and then teach it to apply it in a human-like fashion?

I simply don’t believe that all the human interactions will disappear and it will all be AI. Work will get delivered as a combination of humans and AI. Sometimes AI work being audited by humans and vice versa. Humans and AI are constantly interacting with each other in a seamless workflow. They are correcting each other, learning from each other, and auditing each other. They are passing work back and forth seamlessly.

We’re building a denial workflow as we speak. Right now the way that we do it is brute force. The denial reason that is being presented back to the payer, we’re going to use AI to present the denial letter back. We’re going to use AI to do the doc prep, which is supporting that denial letter. Then it goes to the doctor in Mexico, who says, I disagree with it , or I agree with it, and this is how I would audit it or edit it. Now that is being sent to the payer, but also being presented back to the AI. 

They are learning from each other. The human could learn from AI, oops, I didn’t think that this was like a credible reason or I didn’t think of this combination of CPT and ICD code. That’s a really good reason. AI and humans will constantly reinforce each other, learn from each other, and in my mind, work will get delivered as a combination of humans and AI.

If you think about autonomous coding and radiology, it could very well be that AI becomes 85 or 90%. But if it’s a complex denial more than $100,000, the AI could be just 20% or 25%. The ratio could differ, but it will always get delivered as a combination of human plus AI.

How do companies decide when to make a big AI bet, and if they are wrong, are switching costs so low that they will just take a different direction?

One of our customers told me that they need a full-time person to just monitor all the AI inquiries or propositions that they’re getting from startups. Everybody’s trying to solve for everything.

AGS Health was acquired by Blackstone just a couple of weeks ago. The whole investment hypothesis was, what do you think the scope is for AI? The way that I am approaching it within AGS Health is that we’re taking some very clear cut bets between four to five product lines. I’m looking at denial management. I’m looking at contact centers being up for disruption. I’m looking at how we can do more denials through AI and obviously autonomous coding. 

The question is, can we limit ourselves, fence ourselves, to four to five product lines, or four to five problem statements, and double down and triple down on them and make sure that we are working through them? It’s easy to look at 20 different problems. Each of them looks amenable to AI. The burnout ratio could be high if you end up chasing 20 of them.

The way that I’ve presented to Blackstone is that I’m picking five bets. Be ready for the fact that only three may work out and two may fail. But when the three work out, we will take a larger than reasonable market share. Therefore, we will be well off in the future.

It’s a little bit of change management, whether it’s to the customers or to my own investors, to tell them not to assume that every AI bet will pass the test and be ready for a 30 to 50% failure rate. But let’s take limited bets and see which ones pay off.

How will technology fit into the company’s strategy over the next few years?

The way that I always think about it, and the way that I talk to my own product and technology teams, is that it doesn’t actually matter how fancy the tech is. It could be the fanciest mousetrap in the world, but if it doesn’t solve the customer’s problem … can I collect the dollars faster and much more economically? Can I keep up with the denial claims ratios? Can I keep up with all the regulatory issues? Can I keep up with the payer whims and fancies? If I don’t solve for any of those, then it doesn’t actually matter.

Let’s take autonomous coding as an example. Whatever tech I put in place, if I cannot beat the offshore coder rate, then it doesn’t matter. Am I solving the customer’s problems and am I solving them at an economical rate?  If I have those two questions answered every time I build a mousetrap — whether it’s tech, AI, or a combination of tech, AI, and humans — then we have a winner on our hands.

HIStalk Interviews David Howard, CEO, TeamBuilder

David Howard, MPH, MBA is founder and CEO of TeamBuilder.

Tell me about yourself and the company.

We started TeamBuilder in 2021. It grew out of almost 20 years of healthcare consulting. Much of that involved performance improvement, hospital and health system strategy, and the growth of health system employed physician practices.

Over the past 10 to 15 years, small private practices grew into large medical groups that health systems acquired. A lot of our work was focused around supporting that growth. The early stage of the TeamBuilder concept was studying how staff and staff management within those groups drive cost-effective access to care.

The reality then, as it is now, is that staff scheduling for these health system medical groups and smaller standalone groups is generally done on paper and spreadsheets. TeamBuilder was born out of that prior work to digitize the schedule process through a digital staff schedule that’s built for that care setting. We also apply data science to patient visit volume to align the team by hour and by day to drive cost effective access to care.

As for me, I spent 20 years in healthcare consulting in various aspects, from financial distress to performance improvement.

What challenges of paper scheduling can technology improve?

The scheduling of staff, and even providers, is more complicated than it seems. Dr. Smith might work Monday through Thursday, while Dr. Jones might work Tuesday to Friday. But when you’re talking about the staff – clinical, non-clinical, nurses, and front desk workers — every day is different, because different providers are in the office different days. Each provider sees a different number of patients for an array of reasons, such as the type of their patient panel and how busy their practice panel is. Monday to Tuesday to Wednesday to Thursday can be very different, and very different each week.

There’s a lot of turnover and callouts. A lot of mental gymnastics goes into setting the schedule a month or a couple of weeks out. I’ve got these callouts in the morning. My only front desk person called out. How I find the right person to backfill?

Schedules seem static and stable to an outsider, but a lot goes into it. If it’s a static piece of paper or spreadsheet, it’s hard to make changes and send them back out. Nobody has the right system of record or source of truth for what that schedule is on that given day.

Second is that provider practices, independent or not, don’t have a good way to understand the work that is needed to support that care. It involves a lot of heuristics. A rule of thumb might say that I need two nurses per doctor, but any benchmarks that are out there aren’t grounded in fact. How long does it take to check patients in, check patients out, room them, and come back in and give the injection or support a procedure in the office? It can be eye-opening for what is actually needed versus what managers, providers, physicians, and executives think might be needed .

TeamBuilder does both of those things.

What does your market look like?

When you think about the world of staff and staff scheduling, minds go towards existing legacy scheduling providers. Some great great solutions are out there, such as UKG Kronos, Symplr, Smart Square, and ShiftWizard. They focus on inpatient nurse scheduling solutions and provide the highest value there. It’s very different from the outpatient side of clinical practice and operations.

The outpatient ambulatory side of the house has been neglected over the years. That’s often surprising to people when we talk about TeamBuilder. Many health system executives don’t recognize the differences of staffing across the two.

What variables can be used to prevent overstaffing?

A lot of this is driven by visit volume by hour and by day for the office. In many cases, folks are just thinking, we’ll do 70 visits on this day, so I’ll need this number of people to work these shifts. But what does that look like over the course of the day? Is it 70 visits from eight until noon, and then nobody comes in from noon until 4:00? Folks often anchor by staff or provider, but we believe it needs to anchor on the visit volume and the visit volume throughout the day, not just in total. That is hugely important.

The other variable is how work occurs by specialty. We work with clients to understand their workflow. We have significant client cohorts, so we can say that within neurology, here’s how work is done and here’s how that team can be best aligned.

Do most organizations track productivity and staffing levels using external benchmarks, their own history, or nothing at all?

Some benchmark sets are out there, but the sample sizes are quite low and the questions are simple. The accuracy of the respondents to these benchmark surveys is not very high. The benchmarks that prevail most are the number of staff, which could be clinical or non-clinical, as a function of the number of providers. That becomes a problem, because providers could see 10 patients a day or they could see 35 patients a day. Why would you allocate staff the same if that’s the case?

Another common one is the number of staff per 10,000 RVUs. Relative Value Units is a metric that quantifies the amount of work effort. But it’s a billing designation that becomes a function of the acuity of the visit, how long it took and the complexity of the medical decision-making. But you don’t know any of these when scheduling a patient. So while it’s nice to be able to quantify using RVUs, it’s Monday morning quarterbacking. You won’t know the level of work effort until after it happens.

We anchor on visits. That’s what’s on the schedule and that’s what you need to set the schedule in the future.

What are the employee benefits of efficient scheduling?

It’s important to be able to quickly see your schedule on mobile or web. If the manager is putting out a paper schedule every other week that I take a picture of , what if it changes? Jenny calls out and now you don’t have an accurate view. That’s understanding your schedule, but it’s also important to be able to call out from your shift automatically so your manager doesn’t forget that you told her two weeks ago that you can’t come in.

People in all industries are looking for more flexible schedules. There’s remote work, or I want to be able to pick up a shift on my day off when someone calls out. Trying to manage a dynamic, flexible workforce is hard if you’re doing it on paper and spreadsheets.

Staff love the ability to see open shifts, pick up shift requests, and live in a more dynamic world. A lot of organizations are thinking about, should I pay a premium if I ask Joey to drive in from an hour away? If you pick up a shift inside of 24 hours, do I give you a little bit of a kicker? Staff are  excited about these things.

Can that help to reduce the cost of contracted workers, such as traveling nurses?

We often first think of managing the fixed workforce. You are hired to work Monday to Friday, 9:00 to 5:00, 40 hours.

How do I make sure that you’re providing that effort that you’re contracted for in the right place? Have I hired float pool or flex resources who I can tell where to report at a given time? Do I have per diem staff, either a little per diem group that is managed by the health system itself or engaged from nurse per diem companies to backfill shifts that I can’t fill from the first group? How do I get my best fit resource for the lowest expense and proper skill level alignment? 

Does AI have a potential role in your product?

It definitely does. We are constantly thinking about how to use AI behind the scenes, such as validating code or looking at user experience analytics. We use AI in a variety of ways today.

As we move forward, though, it’s important for our data science and analytics and recommendations to be well understood. Leaders and physicians and managers should be able to quickly understand why that recommendation was made, why this might be a better schedule, and how I should act on it.

At TeamBuilder, we are further clarifying what we do as an operational intelligence platform. We think of it as this intersection of  intelligence, which could include AI, and a practical reality that is well understood and explainable. The right answer can’t come from a black box, where nobody knows why the right answer today is 1.27 nurses.

I haven’t seen many CEOs and investors who have earned an MPH, which looks at how society can improve the health of the largest number of people rather than treating healthcare as a business. How do you see that intersection of healthcare and business?

I started in healthcare consulting out of business school after my MBA. I fell into it and grew to love it. Being able to drive business change inside of a clinical environment has been rewarding. When I was younger, I never foresaw myself getting into healthcare. I was doing turnarounds, distressed work, and strategy for health systems and growth. There becomes a time where you’re only looking at it through the business principles. I did not have as much exposure to the broader public health delivery ecosystem.

Going through the executive MPH program at Columbia rounded out that perspective. How is care delivered? Where does it need to be delivered? How is it done cost-effectively to provide value to community need?

There absolutely can be the intersection between running a business in an organization, but doing it in a way that benefits patients and providers optimally. The two are often at odds with each other, but don’t need to be. The backgrounds of the folks on our team let them live at this intersection between provider experience – which could be clinical or non-clinical – and business experience to be able to translate that. 

What factors will be important to the company over the next three or four years?

The care delivery environment continues to change. The mix of in-person, remote, inpatient, outpatient ambulatory surgery, and in-home care will need to be supported with flexible dynamism. The ways to support those settings are not well understood. A lot of our focus is to be nimble in helping organizations proactively recruit and retain talent and align it to drive care in different care settings.

HIStalk Interviews Jaideep Tandon, CEO, Infinx

Jaideep Tandon, MS is co-founder and CEO of Infinx.

Tell me about yourself and the company.

I come from a technology background in engineering, with a focus on doing back office work over the years in hardware and software. Our entry into healthcare happened by chance, where somebody we were talking to said, can you guys help us out on medical transcription? We said sure, why not? We’re enterprising. Let’s see what we can do.

That was the beginning of Infinx. We saw an opportunity around the overall burden that healthcare providers face when they are dispensing care. We started by providing back office help, but quickly realized that there’s an amazing opportunity here to reduce friction in overall RCM through the use of technology and trained resources.

That has been the journey that we’ve been on over the last 14 years since I co-founded this company. Today we provide a variety of RCM solutions to about 800 customers.

How has health system demand for RCM technology and assistance changed over the years?

Change is constant. It is inherent with the way that our health system is designed. It’s that principal-agent problem, where we’re all giving up some level of control to another party between patient, provider and payer. It is generally set up as a slightly adversarial system. As much as we talk about data interchange and things working smoothly, the incentives aren’t quite aligned for payers to share as much information with providers and vice versa. 

There’s always this friction that has been created inherently in the system. That leads to constant changes in payer guidelines, denials going up, and increased requirements for authorizations. We see no point in the future where suddenly everything will be solved. It’s an environment of managing and continuing to get more efficient as things change.

What progress has been made in making the prior authorization process less frictional?

The prior authorization burden is tremendous on providers as well as payers. Patients are the ones who suffer, because their access to care decreases or gets more complicated.

The end-to-end solution probably doesn’t ever get solved. Payers will always want some level of authorization, which they should in terms of making sure of medical necessity and that providers aren’t overprescribing certain things.  But even before you get to that aspect of it, a lot of information asymmetry exists as providers receive orders and submit prior authorizations. Missing information and incomplete orders are coming to providers, which can lead to denials on prior authorization. 

There’s a lot of low-hanging fruit that can be addressed through technology, as well as better business processes and having tighter controls in the front end of your RCM. That can stop revenue leaking, and more importantly, get patients the care they need when they need it.

How are these capabilities being integrated into the EHR?

EHRs are definitely making a lot of progress in being that single source of truth. Sadly, we still see that fax is still the lowest common denominator of communicating, which is absurd because I don’t think fax machines actually exist anywhere now. It’s just the fax protocol of the thing, “Oh, I’m receiving an e-fax.” 

We’re seeing a lot of interesting things happening in document capture. As much as we’re saying that paper has been or will be eliminated, that’s the primary form of information exchange that we see when it’s a handoff between a referring physician to a specialist, and then from that specialist to a hospital or a health system. Obviously there are exceptions, but the industry standard involves a lot of disparate systems, so faxes end up becoming the way of life because they are low cost and you can get work done. Perhaps not the most efficient, but at least things keep moving along versus burdening IT teams to build broader integrations.

What RCM opportunities might AI provide beyond the earlier phases of offshoring and robotic process automation?

We look at technology solutions as first line of defense across any of the business processes that we are addressing for our customers, but we don’t leave it there. Our view is that our customers should demand outcomes, and that’s what we should deliver to them. 

For instance, in an authorization request, our customers and their patient customers don’t care how we get that authorization done. What they want is a clean authorization on file before date of service so that the patient can be seen in a timely manner and care can be dispensed as needed. Sometimes the ugly truth is that it will require somebody picking up the phone. It’s a stat requirement and you will need to talk to the payer and give all the clinical details about why the patient needs to be seen today, and we will support that.

But we see a lot of things that can be done from a technology perspective. That’s where early days we had machine learning and brought in RPA. Today we’re gradually bringing in AI agents to do more and more of those cognitive tasks that humans were doing. Reiterating the outcome-based approach, it doesn’t matter how we get it done, as long as we get it done with a quality output in a timely manner for our customers so that they can continue to focus on dispensing care.

Are health systems holding prospective vendors more accountable for outcomes that create measurable return on investment?

A lot of the technology spend these days in larger health systems is coming out of their innovation groups. Healthcare has been slow in technology adoption, but we are seeing more of a push to be on the cutting edge and not being left behind that is being driven by these innovation departments. But the folks who are actually driving the business processes, who have been living and breathing those inefficiencies, are pushing back about consuming yet another piece of technology. What is the value proposition that you are delivering to us? How will you ensure that we won’t increase our team size versus actually bringing efficiency? 

A lot of creative things are happening, but more often than not, our customers are defining an outcome and a success metric and saying that we are both going to work towards it. Nine out of 10 times, we’re going to get to those success metrics. Sometimes there are inherent workflow issues or business processes that can’t be changed, and perhaps the technology can’t deliver the value that it promised at the outset. It’s a joint effort between vendors and health systems to better define the problem, because once that’s defined, the guardrail is established, and technology can work really well within those framings.

Will payers use technology that is compatible with that of providers?

With Epic and other EHRs, we are seeing payers coming to the table to support various data interchange standards such as FHIR or previously HL7. There’s more and more of that happening in our ability to connect with benefit managers to get automated responses, be it on claim status checks or prior authorization requests. All of those things are definitely leading towards addressing some of the low-hanging fruit around what can be done through technology and EHR integrations.

But again, we feel that there are a lot of long-tail problems here in healthcare, RCM as well, that going back to my example on prior authorization, we just have to get it done. Let’s not wait for a technology to be 100% effective. If it is 80% effective, it’s a lot better than where some of the health systems are today.

As someone who has started, run, scaled, and sold businesses, how would you assess today’s environment?

Had you asked me that question maybe 10 months ago, my answer would have been very different. The general geopolitical environment worldwide is creeping into business decisions. Organizations are not taking a long view on things because they don’t know what the world will look like 12 months from now, and that makes it difficult for them to get tied into longer term contracts or buying into certain things that then they have to unwind. Commitment levels are getting tested.

But by the same token, innovation is at an all-time high. In the 15 years that I have been doing this, this is the first time that I have seen the investor community, healthcare leaders, the vendor community, and everyone aligned towards making this time different, with healthcare leading the charge versus being stodgy followers that never change. That is refreshing and exciting.

How do those factors affect the company’s strategy?

As we look at our various lines of business, and as we are looking to make investments across the organization, we ask ourselves the question — is AI going to disrupt us as we go down this path? Is AI going to be an opportunity for us as we go down this path?

More often than not, at least for now, we are seeing that the answer is the latter. We can definitely co-opt AI in many aspects of our business, which we continue to do. But it’s not the one silver bullet that will solve everything. Healthcare is an extremely fragmented industry and RCM is extremely fragmented across various specialties and geographies, so M&A is a key piece of our overall growth strategy. 

We feel that there is a lot of domain expertise that exists between various pockets around the country, whether it’s pathology billing, serving academic medical centers, or something complex like oncology billing. We keep looking at opportunities where we can partner with really smart people who have deep subject matter expertise in these specialties, then bring in our technology stack and our ability to globally scale to deliver value to our customers. AI continues to be a cornerstone of how we bring our solutions to market and how we service our customers, but domain knowledge will continue to exist and develop along with AI.

I have never been more excited about what we’re trying to do here at Infinx, along with the healthcare market in general. The ability to reduce friction between payers and providers, bring information to the forefront, and give agency to patients to better administer their own care is an industry opportunity. It obviously brings a lot of competition along the way, and a lot of noise as well, but we feel that we are well positioned. We are excited to be going forward and helping our provider partner customers across the board.

HIStalk Interviews Blake Walker, CEO, Inbox Health

Blake Walker is co-founder and CEO of Inbox Health.

Tell me about yourself and the company.

Inbox Health is a software platform that fully automates patient billing, payment collection, and patient support for medical practices and billing companies. We try to make bills clear, help them get to patients faster, and make them more convenient and affordable for patients to pay. Alongside that, and an important part of that process, is to make sure that patients get fast and empathetic support via phone, text, and real-time chat to get their questions about their bills answered while we’re doing that.

I started in the healthcare space right out of college. I worked on a patient financing startup and quickly learned how problematic the patient AR problem was becoming for medical practices, especially smaller medical practices. I then spent a couple of years working on another patient billing startup and then ultimately founded Inbox Health 11 years ago. I have been with the company ever since, growing it from zero to where it is today, with about 3,500 medical practices using the platform nationwide.

How does the patient’s payment experience influence their satisfaction with the provider?

It’s a huge factor. It’s so intertwined today, the way the clinical experience then carries over into the billing experience. The patient can leave that visit feeling good about the clinical care, but then have such a bad billing experience that their entire perception of that provider is dragged down. 

If there are mistakes in the way their insurance was billed, the provider may end up getting into an argument with them on the phone after the fact. Or even worse, it’s not even the provider, it’s the provider’s billing service provider. The patient may feel that the clinician provided a worse clinical experience because of how the billing went.

Having a great billing experience is critical to making sure that the entire clinical visit is perceived as positive. So much more frequently than ever, that billing experience is a negative one. Patients owe more money. High-deductible health plans are common now, and patients are surprised by the bills they get and are frustrated by them.

As providers, we owe it to the patients to do everything we possibly can to make sure that the billing is done accurately and clearly and that the experience that they have is convenient. Often it’s paper checks only and poor patient support to answer their questions. That will obviously leave the patient with a bad outcome.

How do practices prevent those awkward financial moments that can start or finish a visit, especially when the practice may not know what services the patient will need ahead of time?

Educating the patient about their coverage at the time of the visit, and how that visit is likely to be billed to the patient, is an important starting point. You don’t necessarily need to have it down to the dollar, just that the patient understands that a bill is likely. If they come in to a nutritionist’s office, their child is sick, and you see that they have a high-deductible plan, give them a sense of expectations, such as that it usually takes about two weeks for us to send bills out. Or after your insurance is adjudicated and we know that you’re on a high-deductible plan, I want to make sure to flag that for you. That’s realistic for most practices with just their standard processes for eligibility checks prior to visit and understanding a little bit about the patient’s insurance. The patient can fill in the rest.

Then, whatever you can do to get the bill to the patient as quickly as possible. You don’t want situations where it’s months and then the patient’s getting a bill 90 or 120 days after they came in to see you. I understand that there’s often trouble getting it through the adjudication process with insurance, but getting that timeline as fast as possible so that the patient is in that same frame of mind as when they came in to see the clinician in the first place. Having that be an easy, convenient digital-native experience as much as possible helps to streamline that whole experience. 

How does the method and timing of presenting the bill to the patient affect getting paid?

Most patients want to have both the digital presentment and also the tactile patient statement in the mailbox to know that it’s legitimate. Sending a text message or an email captures their attention, but it doesn’t feel real to them. Once they get the statement in the mail, then it feels more real and they are more likely to pay from the email or text that follows up after the statement. It’s a combination of demographics and who’s more likely to pay from email or text message than a paper statement. But for most patients, email is the most likely way to drive payment.

No method on its own works particularly well. Everyone pays attention to text messages, but are hesitant to click on them and pay because texts are often used for scams. It has to be a holistic approach, where at least in our case, we’re using artificial intelligence to identify what will probably work best for most patients. It’s all dynamic. It has to be an omni-channel, holistic approach to trying to reach the patient in the way that will work best for them and meeting them where they are at any particular time through a process over several weeks and sometimes months.

How common is it for the patient to need or want to contact someone at the practice once they’ve received a bill?

About a quarter of the bills that go out create a question. That’s obviously a huge factor in terms of how you’re running a practice now. If you’re sending these bills out without anticipating and being prepared to answer those questions, it’s going to pull down your collection rates. It’s going to negatively impact the consumer experience with those bills. Most of us aren’t well prepared to do it, but it’s a high proportion that are coming back to the practice with questions. 

A big factor is the amount of money owed. The average family has a $4,000 deductible and it’s not uncommon to end up with a $500 or $600 bill from a standard medical visit. That’s a huge number for most families. You shouldn’t expect that someone who gets a $500 or $600 bill will just blindly pay it from all of this wording that’s on the bill, and why it was billed that way. It’s unlikely that someone will just pay without asking a question.

What is the outcome of those billing calls to the practice? Are patients looking for reassurance that the number is correct or perhaps wanting more details that could have been provided on the bill initially?

I would say that about half could have been addressed upfront. Better setting expectations, providing better information on the bill itself, is often a root cause of the questions that come through. But a significant number of them are related to how the billing was done. It’s often somehow related to coordination of benefits, meaning what insurance was billed and in what order was it billed, particularly for patients who have multiple insurances like Medicare or Medicare Advantage plans or multiple commercial plans. A lot of those cases have legitimate issues that feed back to it, and ultimately, that could be prevented to some degree upstream just by collecting better information at the front desk.

I’m just isolating the patient billing itself. It’s a little difficult to control what problems land on the lap of your patient support team because something wasn’t done well up front.

How often does the patient get frustrated by trying to coordinate the practice’s billing, the insurance payment, and their own financial responsibility?

It’s frustrating for everyone. The provider obviously wants the procedures to be covered to whatever extent they possibly could be. The patient is stuck in this loop where they’re asking the provider questions, the providers are deferring to the payer, and the payer defers to the provider. All sides don’t have a full picture. 

The patient is the one who’s left holding the bag with a bill that someone is demanding to be paid and the frustration of two parties that aren’t seeing eye to eye. It’s common for the provider’s answer to be “ask your payer” and for the payer’s answer to be “ask your provider.” The patient may finally give up and pay the bill or ignore it and see what happens. Patients are seeking that alternative more and more.

What are some best practices for reducing how long it takes to receive payment for patient responsibility?

Optimizing the number of touch points and the channels that you are able to reach a patient on in that first 15 days is critical. That’s the first thing.

Second is meeting patients, from an affordability perspective, where they are. Understanding where a patient’s threshold is for when they might need payment plan options and making those payment plans available to a patient readily. You don’t want the patient waiting 45 or 60 days, getting three bills from you, and then picking calling you and saying, “I know you keep asking, but I don’t have $1,500. I just don’t.” Then you tell them that you can take $50 a month and that’s fine. You need to be proactive about how you engage the patient, which channels you engage them on, and then offering the payment plans when it’s applicable to that particular patient.

We do predicted payment plan offers, where we’re looking at various data points about a patient, their bill, and their past history with the practice and then determining which ones to offer payment plans to and what kind to offer.

But if you can do those two things well, that will get you the best possible result. Some of this comes back to the more that you do at the front desk to educate the patient and collect cards on the file, the more you can accelerate that back end as well. But if you can’t influence that or change that for whatever reason, then obviously on the back side, that approach makes the most sense.

How are you using AI now and how will you use it in the next year or two?

AI has always played a role in how we manage the outgoing patient billing process. The biggest changes in how we’re using AI, and how AI will be used in the patient experience moving forward as it relates to patient billing, is on patient support. We are investing heavily in making the patient support experience better by training large language models to answer the patient questions that come back, feeding it data from the patient record to be able to help it answer patient questions, and letting it actually take action, such as the patient didn’t get a paper bill and wants one, so AI sends it. Or creating a payment plan.

Over the next two to three years, you will see a transformational change in how patient phone calls are answered and how patient chats are answered relative to where we were a couple of years ago, or a year ago. Or even right now, where most of that is either going to the practice staff in the office or it’s being outsourced to the Philippines or India to lower-cost resources. The quality of AI for patient support is rapidly improving and will play a cool role in improving the patient experience in many ways, but in particular, around patient billing.

What factors will drive the company’s strategy over the next three or four years?

Investing heavily in the role of artificial intelligence in the patient experience is a main focus for us over the next few years. And in general, partnering as closely as we can with the best-in-class EHRs and practice management systems to make the experience as seamless as possible for patients where their providers are using different EHR platforms is really important to us. Those are the areas we’re investing heavily in. We believe there’s a lot of opportunity to improve the front desk experience. That’s another area where we’re focused on trying to build technology to improve how the front desk experience connects back to the patient billing experience post-visit.

HIStalk Interviews Don Woodlock, Head of Global Healthcare Solutions, InterSystems

Don Woodlock is head of global healthcare solutions at InterSystems.

Tell me about yourself and the company.

I’ve been in our industry for my whole career – 14 years at IDX, 15 years at GE Healthcare, and eight years here at InterSystems, where I run our healthcare solutions. These are the applications that we sell to the healthcare industry, which includes an EMR that we sell outside the US and an interoperability data product line called HealthShare that we sell around the world. 

Otherwise, InterSystems is also a data platform company. Most famously, Epic is built on our technology, but about 1,000 other systems have been built on our technology as well. We’ve been in the industry a long time, specializing in data and interoperability, with special skills in healthcare.

How central is healthcare to the overall mission of the company, which has $1 billion in annual revenue and 2,000 employees?

We are primarily in healthcare, where we have a deep focus and experience base. It has been our longest business. Our technology is applicable to a few other industries, so we have built up financial services and supply chain product lines and teams, but our heart remains in healthcare.

We have been foundational, so everybody doesn’t necessarily know our name. We are behind other vendors as the interoperability or data platform for many customers. We’re not always out there with the front-end face of healthcare software, but we’re certainly in there on the heavy duty lifting, performance, and scalability of the healthcare data side.

What are the latest developments in the company’s technology?

We are not unusual in that we’ve been focused a lot on how Gen AI can make a big difference to our products and to our customers’ workflow. Each of our products has had exciting innovations, with new features and modules that are enabled by Gen AI. We are certainly in that AI era, a good AI era, and we’re a couple of years into it.  We have a lot more years of innovation and hopefully making healthcare a lot better with this technology.

What parts of healthcare do you expect to see profoundly affected by AI, especially in quality, cost, and access?

At least the first few years, we’re focused on the user experience and having the use of technology like ours, like EMR, be a whole lot more fun, delightful, natural, and more human to use. We moved to graphical user interfaces 20 or 25 years ago. We thought that was a big shift, but it’s still hard to use. It’s still a lot of pointing and clicking, dropdowns, and tabs. It’s not a very natural experience.

With GenAI and approaches like ambient or natural chat user experiences, we will be able to create software that’s a whole lot easier to use, to get information out of, and have it pay attention to our instructions and do useful things for us. Historically, software has been kind of dumb. It follows the instructions of the user, stores the information that I type in, and then shows it back to me a few days later when I ask for it again. AI can allow us to build a lot smarter software that will be more helpful to us as users and hopefully will help transform the industry.

We first focus on the user experience. Down the road, we’ll start to move into other areas around clinical decision-making, workflow optimization, and a better patient experience. There’s a lot of places we could go with this technology. 

You just announced IntelliCare, a next-generation, AI-centered EHR that is available only outside the US. What were the lessons learned in developing it?

We took a bet that worked out, which is that AI should be natively built in the EHR versus just a partnership with somebody else. That’s really working out. It’s enabling us to do closer integration of the technology into the workflows of the user instead of having it be an arm’s length relationship. That’s been good.

It takes more R&D to do that. We’ve had seven teams working on this across our EHR development teams. With enterprise EHRs versus best-of-breed departmental systems, enterprise has won out as the right strategy. I don’t think that AI is that different. You want to embed it into that enterprise feel versus having it be a best-of-breed type system. We made that decision early on that we would do this natively. That cost us more, but I think it’s going to pay off, and it is already paying off with some of our early adopter sites.

Other lessons learned with AI is that it’s important to work closely with our customers. There’s a lot of trust issues with AI. There’s a lot of education issues in terms of how these systems work, how we test and validate them, and how we get comfortable with the way that our data is handled by a cloud AI provider.

There’s a lot of new ground on the InterSystems side, but also on the customer side in terms of governance, legal, safety, and a comfort level with AI overall. We’ve had to spend more time than I would have guessed on the customer side, educating them and getting them comfortable with what we’re doing. Maybe part of my education push on AI was observing how much the market needed to learn about AI in order to adopt it well. We’ve just encountered a lot of that with our early sites.

How does traditional software development, maintenance, and support change when you add an AI component?

The good news is that all of these large language model vendors basically use the same APIs and the same way to call them. There’s not a lot of technical investment that you’re making in one road that’s not useful for another road if people continue to leapfrog each other and things change. 

What is really tricky is the testing and validation process, because when you are dealing with generative AI and you ask the same question multiple times, you’re not going to get the same answer back. There’s a non-deterministic aspect to the way large language models work, even on the inbound side. If somebody’s asking a question about a patient chart or whether they have been seen for this condition before, there might be multiple ways that that clinician might ask essentially the same question. There’s non-deterministic aspects on the user side and then certainly non-deterministic aspects on the answer side. 

We had to invent our automated testing process and our validation process from scratch. That is much different than our traditional process, where we want them to fill out these four dropdowns and get the answer “32” in the end. For this non-deterministic process, we’ve had to build up a completely different automated testing infrastructure and validation infrastructure. We have a lot more human validation with real physicians and nurses in the process. Testing, measuring accuracy, and then maintaining that accuracy as the model providers come up with new versions is a whole different design and architecture that we needed to build around this.

How are you using AI tools personally?

We provide our employees with OpenAI licenses with an enterprise agreement, where they can use it for company confidential stuff. We’re enabling our employees, myself included, to use and take advantage of the technology.

For me, I use it most for coding side projects. I do a lot of AI side projects just to keep current with the technology. These large language models are excellent at writing code, answering technology questions, debugging, and stuff like that. It’s remarkable how well these technologies work as maybe junior programmers or code developers along with you. 

One way to view these AI technologies, at least for the next couple of years, is for empowering every human employee here with a co-marketer, co-developer, or a co-implementation person who can help them be better at their job, be more productive, debug problems faster, and that kind of thing. 

The industry could use basic AI education to navigate the opportunities and risks with AI effectively. I’ve always enjoyed teaching, so I am doing five or 10-minute videos called “Code to Care” to explain AI concepts. I always have enough content because buzzwords are being thrown out that people don’t understand or that vendors overuse. I am enjoying putting together that AI education. It’s important. HIMSS, HLTH, and ViVE have a lot of sessions where educators don’t get into enough depth, or maybe they don’t know enough depth, to help you understand some of the newer topics and approaches.

I don’t know if it’s to the company’s benefit or not, but I certainly enjoy doing it. I enjoy hearing from people across the industry who have known me over the years who like my video content. It’s important that we navigate this AI wave effectively.

What has been the impact of moving to the cloud?

I’m finding that our customers are struggling with anything on-prem these days. Maintaining a data center and keeping hardware and storage current, updated, and patched for security vulnerabilities is a growing challenge. More and more of our existing customers are asking us to host their platforms or offer the same functionality as some kind of service or equivalent. 

For our net new business, we almost do everything as a service. People within health systems and payers don’t want to be doing this anymore. It just doesn’t make sense economically. It’s the predominant model that we find to make software and technology available to customers. We do the heavy lifting, such as maintaining the staff, buying hardware if we’re doing it ourselves, or procuring it from one of our cloud partners. The industry is just kind of done with on-prem software and relying on their software vendors to manage it as a hosted or software-as-a-service platform.

Is interoperability a solved problem?

[Laughs]. No, no, no, no. I definitely think that the ball has moved, which is great. When I started in interoperability, the use case was a provider seeing a patient, let’s say in the ED, and wanting to know what happened with this patient outside of my health system. That is getting solved. National networks like CommonWell or vendor networks like Epic’s Care Everywhere have done a fabulous job with that use case, and the ball has moved.

But we’re trying to do new things. We are working hard on the payer-provider interaction, like electronic prior authorization, clinical data exchange, payer data exchange, and patient and member access to their information. Those are new exciting use cases that we’re working on as an industry.

The industry still struggles. We are in the middle of this with our technology and services with mapping data in one format and making it consumable and useful in another format. So it’s definitely not a solved problem. We are enjoying a great growth of FHIR as an approach and a set of standards, and that is helping with all of these new use cases. 

Things are getting better. We’re moving on to slightly more advanced problems from an interoperability point of view, but it’s certainly not a solved problem at all.

What near-term trends will influence the industry and the company?

InterSystems has been around for 47 years. We have a slide that we talk about, which is the advent of micro-computing, PCs, the Internet, cloud computing, and now Gen AI.  Each of these is maybe a 10-year-long transformation that has allowed us to do great new things. All of those significantly advance the impact that computers and software have had in healthcare. Gen AI is going to be either no different, or even better, than some of those prior transformations. That’s a terrific trend. 

I also think that cooperation among payers, providers, public health, Medicare plans, and others within a community is getting stronger. It will make it easier as a patient and as a caregiver for your family to navigate the healthcare system. I hope that technology, interoperability and cooperation across communities will continue to improve. I certainly see it improving with customers that we work with.

HIStalk Interviews Anthony Lucatuorto, CEO, Sphere

Anthony Lucatuorto, MBA is CEO of Sphere.

Tell me about yourself and the company.

I’ve enjoyed a 25-year plus career in the fintech space, with leadership experience in embedded payments, digital engagement, and high-growth partnerships. I began my career working at Mastercard and American Express in the early 1990s and held executive roles at First Data, TransFirst, and TSYS before arriving here at Sphere, Powered by TrustCommerce in 2018.

TrustCommerce is a financial technology company that provides secure, integrated healthcare payment solutions to some of the largest health systems in the US. We’ve been doing this for over 25 years. More recently, we are proud to have launched our next-generation card present payment solution, called Cloud Payments, that advances our EHR integrations and continues to help make the patient payment experience more seamless, flexible, and secure.

How are providers supporting newer payment methods and technologies?

Collectively, the industry is on the right track. We need to implement it with what I always preach to my team, which is a quicker speed of play. The act of paying for healthcare is just different than in retail. We need to continue to ensure price transparency, educate patients on their responsibility prior to treatment, and provide patient billing plans. We need to capture cards on file, along with an account updater tool that helps keep our tokenized card information current.

Over 50% of private employees in the US who participate in medical care plans are enrolled in high-deductible plans, according to the Bureau of Labor Statistics. These deductibles keep getting higher year after year. Embracing technology like digital wallets capabilities and recurring payment tools are great examples of ways that providers could help collect more of the growing patient responsibility.

How does healthcare compare to other industries in that regard?

If you or I have a retail-like experience, we take for granted that it comes with speed, convenience, transparency, and trust. The healthcare industry is catching up to this. The industry is a little bit behind, but is now providing more advanced omni-channel payment options, more payment methodologies such as Venmo and PayPal, and more digital wallet products like Apple Pay and Google Pay. These are great examples of ways to get closer to what the patient experience needs to be, which is what they are experiencing on that retail side.

Are providers generally accepting patient payments from Venmo and digital wallets? Is use of those methods skewed to a particular demographic?

It’s funny that you say that. I’m Gen X and I don’t know how many baby boomers are using these, but my generation certainly has embraced it. Millennials and Gen Zs are certainly embracing Apple Pay, Google Pay, and all of the digital wallet products very well. They are after speed, convenience, transparency, and trust.

From the provider side, you have to do that in all of the omni-channel payment options. Whether it’s at the point of sale, in front of your doctor’s office at a kiosk about to check in, or you go online, all of these omni-channel payment methodologies need to accept these forms of payments. More and more, they are.

How are virtual credit card numbers and tokenization being used?

More and more of our clients are capturing tokens. In fact, in our experience, over 70% of healthcare organizations offer patients the ability to store a payment method, and that’s super important. Providing an account updater tool is important, so as a card expires or gets lost, you’re constantly updating the information. Keeping that tokenized card on file helps the provider collect payments today for the future, post-treatment billing down the road, and recurring billing options.

There’s a wealth of opportunities for them. It’s a growing tool and product and providers are certainly starting to embrace it.

For that virtual card question, we’re discussing this a little bit more than we used to. It’s more on the insurance side at this point, where insurance companies send a virtual card to providers. It allows the provider to collect quicker and maybe with more enhanced data for their reconciliation. However, it comes with a cost, because now you’re introducing card brand fees and acquiring fees, which the providers wouldn’t have had with just an insurance payment. Providers absolutely have to weigh the cost benefit of these virtual cards.

What do pre-arrival financial activities look like?

We don’t see payments as a standalone event that always happens at a certain time. The payment needs to be woven in throughout the patient’s journey, complemented by all the tools that are available to help the patient know what they owe, why they owe it, and who they owe it to. Then, to set them up for the best chance of being able to pay their bills. Patients don’t know or care that the appointment reminder system might be a different company than the scheduling system or the patient billing system.

Providers that are being successful in this area are the ones that step back and think about the entire patient experience from beginning to end, giving the patients the right information at the right time to take the right action. That’s really the key.

How does EHR integration work?

We focus on helping our providers collect payments. We are super proud of the fact that we’ve been integrated into Epic, as a great example, for more than 15 years. We’ve done so in all of their native workflows. 

From a provider standpoint, we are embedded in all of the workflows of the EHR, Epic as a great example. They see that as a great experience and greater opportunity to collect payments. It becomes more streamlined workflows for the provider’s patients. It allows centralized reporting for analytics across locations. On the patient side, they have greater information, which is greater cost transparency, and simpler flexible payment options. It’s all within the native workflows, which helps make reconciliation seamless.

How are providers implementing propensity to pay and payment plans?

Most of all of those tools exist in the EHRs, so from my vantage point, I’m making sure that my solutions are embedded into all of those collection points. When they get that that pre-estimate, if they want to make a payment, I’m providing the tool and the access for that provider to collect that. If they want to wait until after service, I’m providing the tool in that omni-channel environment to make that payment. I’m making sure that all payment methodologies are captured, whether it’s Apple Pay, Google Pay, Venmo, PayPal, or ACH.

How will AI impact your business?

It’s growing exponentially. I expect it to play a larger part, exponentially, quite honestly. We’ll see it in the service side. We’ll see it in our development side. It is exciting and we’re absolutely diving into it and analyzing everything we can.

What will be important to the company over the next two or three years?

We are going to continue to advance our products so that they remain on the cutting edge of being seamless and secure. It always starts with security. We’re going to make sure that we know where the puck is going as it relates to whatever is the next form of card payment. What’s the next Venmo or Apple Pay that’s coming around the corner that the next generation of payers want to use? We’ll make sure that we invest in that technology.

We see healthcare providers heading in the right direction. We’re happy to be a part of it. To summarize the ways that they could continue to help build a path to better collections of patient payments, continue to think of the journey from beginning to end of the whole patient experience. Provide those cost estimates upfront, support the flexible payment methods, provide those omni-channel payment options, ensure that the secure payment storage for future treatments and recurring billing, and continue to communicate early and often. That’s the best thing we can do.

HIStalk Interviews G. Cameron Deemer, CEO, DrFirst

G. Cameron “Cam” Deemer is CEO of DrFirst of Rockville, MD.

Tell me about yourself and the company.

I’ve been with the company for 21 years, as CEO for the last two and a half years. I started my career in the ministry, where I worked many years with churches in Arizona and did some time in Papua New Guinea working with Bible translation. I got into healthcare IT totally accidentally during the recession in 1992, when I came back to the US. I grew up here in the PBM industry. I worked for the old PCS Health Systems in Scottsdale and then building what eventually became Surescripts while I was at NDC Health. I have been with DrFirst ever since.

DrFirst is celebrating its 25th anniversary this year. It has been a really interesting ride. The company was originally founded to eliminate what I would call technology friction for the doctor. I was the first product manager for this little 30-person company when I came here. The CEO and founder, Jim Chen, sat down with me and said, “Look, Cam, you need to understand one thing. Here’s what I need you to do. I need you to make sure that DrFirst is number three in the e-prescribing space. That’s your target.”

I was like, “Why would you want to be number three?” He said, “Any industry always ends up with three players. I don’t care if we’re number one or two, I just want to be number three.” With that challenge, I started my career at DrFirst. 

We have developed a huge footprint over time. We are handling scripts for about 550,000 doctors a quarter. We provide the e-prescribing back end for 275 EHRs. We are actively engaging 6 million patients a week for various programs, most of them adherence related. We have always been an innovative company that is trying to solve what we think are the real, gritty problems in healthcare.

You said in a recent article that the pharmacist’s role in patient care was limited by the original design of the electronic prescribing network. Do you see that changing, particularly given the struggles of the drugstore chains?

When I joined PCS in 1992, the company was already experimenting with e-prescribing. They were doing some of the early seminal work of getting some doctors to use the system. I believe at the time that they were using email to send prescriptions to Walgreens, and then the pharmacist would pick it up from the email. In those days, there were a lot of early attempts to do e-prescribing in different ways and PCS wasn’t the only company involved. That was eventually all destroyed by a patent troll who went through the industry suing everybody. Then that industry had to reinvent itself. 

While that was all going on, PCS selected me to be their representative on the script task force at NCPDP when it was first put together. I was one of the three original co-chairs working on SCRIPT 1.0 and continued to be loosely involved with e-prescribing at PCS. I was thrilled to death when we announced the creation of RxHub in collaboration with ESI, Medco and PCS at the time. We built that system based on technology that already existed, the pharmacy claims switching technology. 

It made all the sense in the world. The PBMs had already transformed pharmacy, and healthcare IT around pharmacy, by fully digitizing the claims process. It worked great. We went from paper claims to terminals and then PC or mainframe-based transmission of claims through a central switch. It was great, so you can imagine that when they turned their attention to e-prescribing and said let’s digitize that, they had every belief they could do it. They had what they believed was the right technology, and they built it. 

When Surescripts saw what was happening, they decided they had to do the same thing and not let the PBMs cut them off from the doctors. So they also implemented e-prescribing based on claims switch technology, which I had been helping develop while I was at NDC Health. When I came to DrFirst, RxHub was already there. Surescripts was already there. DrFirst was working to get the doctors to adopt e-prescribing. None of us had any time to really think about it.

This just seemed like the right solution. It did what it was supposed to do. It let 600 participants on the EHR side talk to 600 participants on the pharmacy side and solved that many-to-many problem. It converted the prescription from writing into a digital format so it could be picked up on the other end and imported into systems. It accomplished its goals. 

We are 20 years down the road now, and it is just now that the cracks are starting to show, as people are saying, “This doesn’t quite work right. This isn’t quite what we need.” It has become a real cap on innovation in the industry, the way the e-prescribing network works. Fundamentally, the problem is that it’s a technology that was designed for pharmacy claims.

A pharmacy claim is a financial instrument. It’s a request for reimbursement from the pharmacy, an accounts receivable, essentially. It goes to a PBM, who runs it against a contract in the adjudication process and promises to pay. They send back an accounts payable transaction to the pharmacy so they can reconcile that to the accounts receivable that they sent forward. So it’s basically two participants, a pharmacy and a PBM. It’s based on a contract, and it’s a mathematical process. There’s nothing more to it than how it matches against the contract.

Prescriptions are totally different. A prescription is a clinical order, not a financial instrument. It is initially ordered by a highly trained clinician who has evaluated a patient, considered their current problem, their ongoing problems, their other treatments, and the other medications they’re taking and then making a decision based on what they know about that patient. Not just clinically, but also all the other factors, such as what they can afford and what they are willing to do.

Then they send that clinical order to another highly trained professional at the pharmacy, who would like to be able to evaluate that, add their thinking to it, and interact with the provider who wrote the script in case something needs to change. Eventually that’s filled by the pharmacist and the patient needs to pick it up, so the patient is another participant here who maybe doesn’t know whether their prior auth has been approved, whether it’s ready at the pharmacy, and how badly they really need it.

There’s a lot that the patient also has to think about through this process. There’s also pharma, who really, really, really wants the drug to be filled at the pharmacy once it’s written. And there’s of course the PBM, who’s interested in getting as clean a script that ideally matches what they’re trying to do with the patient as well. You have five key participants in this process, all trying to work around a transaction that is flowing through something that was designed for claims, and it just doesn’t work. 

The biggest failure point is that the script arrives at the pharmacy with no context. The pharmacist can’t really do their job, the job they’ve been trained for, because all they get is the digital version of the script. We’ve been taking a look at that for many years. It was the recent FTC settlement with Surescripts that opened the opportunity for other networks to enter the market, so we have introduced our version, which includes the ability to carry the extra data needed with the transaction to establish rule sets by which we can manage workflow issues.

For instance, some scripts get to a pharmacy that the pharmacist is never going to be able to fill, so they have to call the doctor to get clarification. We can handle that on the front end, just based on a rule that says if you get a script like this, ask the doctor to correct it in these ways before it actually goes pharmacy. It saves the doctor a phone call, saves the pharmacy a phone call, and the patient has access to therapy more quickly.

The solution is freeing up the clinical order to be a clinical order and to have everything it needs to be processed without a lot of friction at the pharmacy.

Early e-prescribing was done on a standalone PC or terminal. How has it progressed to integrate back into the EHR?

You raise a good point that initially e-prescribing was standalone. Now it’s fully integrated into the EHR. It’s just part of the standard EHR workflow. 

What’s been done over the last several years is bringing more of the information the doctor needs to make a decision into that system. One is real-time benefit check, where you’re doing a pre-adjudication of the script to give the doctor an idea of what the impact will be on the patient when they show up at the pharmacy and having to pay for that prescription. Also giving the doctor the same information on alternative drugs that would also be applicable under the therapeutic class so the doc can make a more informed decision based on plan design. That has helped people avoid prior authorizations, so that the doctor can see one drug that requires prior auth, so I’ll go with the one that doesn’t. Along those lines, information about the patient’s plan has been useful.

What’s coming now is more information about the prior authorization question sets that the doctor needs to answer ahead of time. The ability to grab that information from the EHR and send it along with the request for the PA so that you don’t have to have back-and-forth between the doctor and the PBM to get the PA approved. A lot of what’s happening now is pulling information into the doctor’s office that would avoid them having to have phone calls or electronic back-and-forth with pharmacies or PBMs.

How are you looking at AI?

We definitely don’t believe that just slapping a chatbot on top of our existing stuff counts as AI. We’re actually trying to take a much deeper approach to it and go at it from three levels.

Probably most important and foundationally, we’re trying to train every single person in DrFirst to be comfortable with the concepts of AI. Comfortable with interacting with it to help them develop individual visions around how AI can be used to automate processes at the company, as well as be incorporated into our products to improve workflow for other people. So we’re starting with our people first. We move it then into the feature set of our products to make workflow better for the folks who use what we do. In other words, it becomes a feature.

We are just now starting to work on an actual product that is completely AI based. For us, the most important applications of AI are practical. We’ve been using it, actually for quite a while, for interoperability solutions that are e-prescribing. We’ll continue to expand focusing on AI’s ability to help people get work done quicker with more information and fewer redos and stuff like that.

What kinds of medication-related engagement do patients want or need?

Some interesting things are happening right now. Consumer engagement is really hot in healthcare. It has been interesting to watch how that is expressing itself out in the wild. This is one of the areas that we’re intensely interested in, but I probably should have said this earlier. We like to think of ourselves as productively contrarian. It doesn’t matter to us so much what other models are being built right now. We are more interested in what’s the right way to handle the situation. 

If you think about what happens with patients today, there’s a lot of activity around patient choice of pharmacy. What if the script is written for a patient, and then for some reason, the patient wants to go to a different pharmacy? I’m going to be passing this pharmacy on the way home. I’d rather pick it up there, or I found that I can get a consumer card that would be cheaper at this pharmacy than that pharmacy. There’s a lot of talk of use cases like that that aren’t really all that interesting because they’re probably fairly rare based on how people tend to develop habits and how they pick their pharmacy in the first place. I usually think of that as trying to come up with a solution where there’s no problem.

But there are other more significant things. Patients who have important drugs that they need to receive, but the script has been written to the wrong pharmacy. For instance, a specialty drug that has a limited distribution network is sent to their regular retail pharmacy which may be reluctant to give it up, because if they can special order the drug, they can probably make a decent profit on it. But the patient’s going to have to wait much longer than they would if it just were going to the right pharmacy in the first place. Being able to alert the patient that the drug has been sent to a pharmacy that can’t fill it for them immediately is an issue.

Another example would be that the patient shows up at the pharmacy and the pharmacy says, “We don’t have that in stock. Give us until Wednesday and we’ll have it.” The patient may not want to wait till Wednesday and they need to have it filled at a different pharmacy if they can find one that does have it in stock. The ability to switch that script without having to rely on the pharmacy being willing to give it up, or a doctor being willing to rewrite the script to a different pharmacy, that’s all friction from the patient side.

You see a few different solutions in the industry. For instance, having the doctor write the script to a company that will then show the patient on an app that they’ve downloaded that a prescription has been written for them, then giving them a choice of pharmacies so the patient can pick a pharmacy. Another model might be to persuade the doctor to write the script to a non-dispensing pharmacy, which would then determine the best place for the patient to fill it and then reach out to the patient in different ways to give the patient the option of which pharmacy to use.

These solutions are pretty hot right now. There’s a lot of talk about those. But they suffer from the fact that they require everybody to be out of workflow. The doctor has to not use the default pharmacy, they have to write to a pharmacy that’s not actually going to fill the script, but it’s going to get the patient to fill the script. The patient has to download something and go through an extra step , where otherwise they would just show up at the pharmacy. The physician is out of the workflow and the patient is out of workflow. Typically the folks that are doing these kind of models struggle with volume, and no wonder since everybody’s being required to do something unusual.

Another dirty little secret is why, in the early days of e-prescribing, NCPDP picked this model instead of a more European model, where the script would go to the cloud, the patient would just show up at whatever pharmacy they wanted, and the pharmacy would pull it down from the cloud. That was actually considered in the early days of the SCRIPT standard, and it was promoted at the time by a representative from the University of Alabama.

I remember the meeting where this happened. Everybody else in the meeting disagreed with that approach because they were trying to solve for the adherence problem, that patients are given scripts and are then trusted to deliver that piece of paper to the pharmacy. That creates one more barrier for the patient actually getting the script filled. The pharmacist isn’t doing their job, the doctor’s not getting the results they want, and pharma certainly isn’t getting revenue from the drugs being filled.  So instead, they decided to have it sent to the pharmacy. That will set up an expectation in the patient. They need to go pick it up. It can create a little work for the pharmacist because if the patient doesn’t show up, they have to return it to stock, which is a pain in the neck. 

Nonetheless, it has worked really well. It did in fact improve adherence dramatically. Patients are much more used to picking up their drugs now than they used to be. When we go to a model where the patient becomes the delivery mechanism again, you’re just stepping back into the past to a time when compliance was happening at a lower rate. First-fill adherence was lower than it is today and patients weren’t getting on therapy. We believe the right way to do this is to keep it all in workflow. Let the doctor write the script and let the patient interact with the physician directly whenever the script can’t be filled for some reason. Don’t force the patient into making a selection if they don’t want to make a selection, because if they don’t make one, nothing happens. Make sure the script still gets to the pharmacy.

What will be most important to the company over the next two or three years?

Number one is that we are reinventing the e-prescribing platform. We’re going to give the industry what it deserves. Doing that is very important to us.

We are working to eliminate all the points of friction in the specialty drug workflow. That will become increasingly important with new developments in medicine.

Those are the two challenges we’re taking on right now.