Editors note: This transcript was edited for clarity.
Caroline Pearson: Across all of the reports where we see the greatest potential for digital health solutions is in areas where we have existing care models or interventions that are very effective, but that too few people receive. So let me give you an example. If we look at musculoskeletal disorders, we know physical therapy is incredibly effective in terms of improving functional outcomes and reducing pain to a greater extent and more quickly than rest and other clinical interventions. People don't do it, they don't do PT. It can be hard to access. It can be annoying to get there. It can be expensive if it's not adequately covered by your insurance. Where virtual solutions can take what is already a known clinical intervention that works PT and make it more convenient, more accessible, and more affordable you really start to see the scale of those interventions.
Matthew Weinstock: That was Caroline Pearson, executive director of the Peterson Health Technology Institute. The institute conducts independent evaluations of digital health technologies. The goal is to see if technology solutions drive both clinical improvement and lower costs. The institute has assessed technologies aimed at depression, hypertension, diabetes, and musculoskeletal conditions. In this podcast, Oliver Wyman's Ran Strul and Pearson talk about the value of having an independent and thorough assessment of technology, especially as executives face an increasingly crowded field of solutions. Pearson also shares insights into how they determine which solutions should be adopted, which may need more testing and which ones aren't viable. The Oliver Wyman Health Podcast is brought to you by the global management consulting firm, Oliver Wyman. For more insights on the business of transforming healthcare, visit our online publication, health.oliverwyman.com. And now let's pick things up with Pearson talking about why the Peterson Health Technology Institute was created.
Caroline Pearson: The Peterson Health Technology Institute was launched in 2023. We are a subsidiary of the Peterson Center on Healthcare, all of which is part of the Peter G. Peterson Foundation. The foundation is really a philanthropy that has been long focused on reducing the federal deficit and very quickly, Pete Peterson figured out that healthcare was the primary driver of our national debt. So he launched the Center on Healthcare with a mission to improve affordability and outcomes of healthcare in the United States. We've been working hard for about a decade trying to achieve that goal, but we understand that technology has a really significant role to play in terms of all of our ability to close workforce gaps, improve healthcare outcomes, and hopefully do that at a cost and scale that we're going to be able to afford for generations. So PHTI is really part of that. We perform independent assessments of digital health tools. We really look at categories of solutions and try to answer two questions, what works and what's worth it as employers health plans and health systems are making purchasing decisions.
Ran Strul: What domains are you focusing on? Are there areas where you are staying away from the time being?
Caroline Pearson: Well, we have so far completed four assessments of health technologies. We've looked at digital solutions for type two diabetes we've looked at some of the virtual PT solutions to treat musculoskeletal disorders interventions for hypertension and mental health care. All of those have a few big things in common. One, they are very large areas of healthcare spending. So we're trying to start off by tackling big chronic conditions that drive both a lot of disease burden in the country and a lot of healthcare costs. We are also looking for areas that are at the right level of maturity of the solutions that are being offered. We don't want to evaluate any area too soon and really quash innovation or have unrealistic demands for what kinds of evidence those solutions can have produced early in their life cycle. We do want to pick categories that are at the steep part of their adoption curve that are beginning to take off in the market, and we can really help give clear guidance to healthcare purchasers about the impact that those solutions are having, both on clinical outcomes and on economic impact. Those are really the two things that we are primarily looking at as we're going across any of our areas of work.
Ran Strul: And what have you been finding? Are these solutions actually working?
Caroline Pearson: Across all of the reports where we see the greatest potential for digital health solutions is in areas where we have existing care models or interventions that are very effective, but that too few people receive. So let me give you an example. If we look at musculoskeletal disorders, we know physical therapy is incredibly effective in terms of improving functional outcomes and reducing pain to a greater extent and more quickly than rest and other clinical interventions. People don't do it. They don't do PT. It can be hard to access. It can be annoying to get there. It can be expensive if it's not adequately covered by your insurance. Where virtual solutions can take what is already a known clinical intervention that works, PT and make it more convenient, more accessible, and more affordable, you really start to see the scale of those interventions. We saw similar things on some of the hypertension interventions that have great medicines. If we get people on the right blood pressure medicines, we can really improve their outcomes, but too few people get treated for hypertension even when it's diagnosed. Conversely, what doesn't work so well is behavior change. Across all of our reports solutions that are focused on behavior change have really struggled, and that is because behavior change is hard. If someone had a chronic condition, they've been managing that condition the same way for many years, giving them a digital solution or an app, no matter how engaging and motivational it may be, it's very hard to change all of those existing habits. That's a place where we have consistently found that the digital tools have not performed as well as many people would hope.
Ran Strul: I'm curious in how you guys were thinking about behavioral health, virtual behavioral health, where would that sit in the spectrum between we have an intervention that works and we just need to deploy it at scale versus the intervention is not quite defined yet and no technology will help until we actually figure out the clinical model?
Caroline Pearson: Well, you are right that virtual mental health care is a little bit different. We looked at specifically solutions that were treating mild to moderate anxiety and depression symptoms. Actually, as it turns out, in mental health care, lots of different interventions work. This is a case where it is really all about what is the treatment or intervention that you can get a patient to engage with. Whatever the modality or form of that engagement is going to be is what is likely to be most effective for them. When we look at the clinical literature, there's obviously great evidence that talk therapy or psychotherapy is clinically effective. We know that medications for depression and anxiety are also clinically effective, but there are a wide range of other activities, whether those be CBT based activities and homework and self- guided tools, or that's things like exercise, like engaging in community. In this case, there are lots of options, and when we look at how digital health companies have implemented these options, we really see quite promising results, which is that digital self- guided activities can be almost as effective as talk therapy for people who are not otherwise receiving treatment. We see similar promise among things like chatbots, like asynchronous therapy, like text- based therapy. That really creates a huge amount of opportunity for virtual mental health companies to say, " How can we expand access to care? How can we create a range of treatment modalities that are going to appeal to different types of people in different contexts of their lives?"
Ran Strul: Can you share a little bit about how you do the work? You come back with pretty conclusive statements about the efficacy and the cost outcomes of these solutions. So, one would ask, what type of research are you doing in order to make these determinations?
Caroline Pearson: Well, when we set out to do this work, people said, " You're never going to be able to do this. There's not enough evidence." What has been surprising to some people is exactly how much evidence really does exist. We are looking only at secondary evidence on the clinical basis. We're not doing our own data analysis. We're not receiving data directly from health plans or employers. But there is a lot of published literature in any one of these areas. So we are generally looking between 2000 and 5,000 pieces of evidence that we review in our systematic literature reviews. We then work very closely with clinical advisors who we recruit specifically for each of the reports to help us understand what exactly the most important primary clinical outcomes should be. We understand that for patients, there's a wide variety of things that matter, and we want to look at all of those outcomes. But when you look at people with type two diabetes, the ultimate clinical goal is to reduce their hemoglobin A1C. When you look at people with hypertension, we want to see their blood pressure come down into blood pressure control, and we often measure that through looking at changes in their systolic blood pressure. It is helpful in fact to say to the market, these are the clinical outcomes that matter most, this is what we're going to prioritize, we do a deep dive in looking at that evidence, and then we look at a range of other outcomes including user experience, engagement, measures of health equity, and other important features of these solutions. We also then create a budget impact model. Our budget impact model is really looking at things from the view of the health insurance system as a whole. So whether that's going to be commercial health insurance coverage, Medicare, or Medicaid we're building sort of one to three- year budget models looking at total cost of care. A solution that makes someone better tends to also reduce their other healthcare utilization. But then we need to offset that by the added cost that any given purchaser is going to be paying for the solution. And we stitch those two pieces of research together into what we hope are clear and directive findings to the market. We are certainly not neutral. We are empiric and evidence- based, but then we try to take all of that work and communicate very clearly with strong recommendations about which solutions should be adopted, which solutions need more testing, and which solutions simply aren't worth the effort, time and money to which we're devoting to them.
Ran Strul: I'll give some insight or example from one of the reports that I've read for our listeners for the diabetes management solution. One of the statements for one of the vendors was, and I quote "Small but not clinically meaningful reduction in H1C and a net increase in spending. The conclusion is that current evidence does not support broader adoption." Now, just to paint the picture, almost all of the names that you see there, these are vendors, many of them capital-funded, private capital-funded, or otherwise with clear growth expectations and so on. You are making pretty clear and strong statements about that. What pushback are you getting, and how do you manage that?
Caroline Pearson: Well, we have been very fortunate to have a very high degree of engagement with the companies that are in our reports as well as companies that haven't yet been assessed. We didn't know when we set out what the reaction would be. We thought companies might hide from us. But we have been very fortunate that actually in all of our reports, the vast majority of companies that are included have chosen to work with us. They engage directly, they submit lots of evidence. They also submit commercial information, some proprietary pricing information. They generally believe that working with us is better than hiding from us, and we have greatly appreciated that engagement. Certainly no company wants me to indicate in our report that we don't think that their solution warrants broader adoption. The market is at a place where right now, companies do not have clarity about what is going to help them win. Are we focused on engagement? Are we focused on having a multi- chronic condition platform so that it's easy for people to adopt? There's lots of competing value propositions in the market. Vendors really appreciate understanding that there is a clear set of rules against which some of their customers are going to begin to assess their products. That gives them a sense of, okay, if clinical outcomes are going to matter, then we can begin to invest more of our resources in generating that evidence and selling against those metrics. And so I think we've really begun to see that conversation change over the long term. Hopefully that creates more predictability about how you can successfully grow and scale a company. And it also helps companies innovate. If a solution isn't working as well as we all as a society hope it is, then we need to push for faster changes and updates to those products to try to do better by all of the patients that we are trying to help.
Ran Strul: And aside from the institute team, essentially that performed the assessments, who else is involved in this and how are maybe purchasers are involved in these assessments?
Caroline Pearson: Well, we work very closely with our purchaser advisory council, which includes over 40 employers, health plans and health systems that are responsible for deciding on which digital health solutions they're going to adopt and pay for on behalf of their members or their employees. They are incredibly helpful because they are really the end customer for all of these solutions. They can tell us both what data and information they are looking for, that then we can try to fill that gap. They can also tell us what areas of the market they are going to be focused on so they can help us point future reports at areas that are going to be upcoming decisions for any of those employers and health plans. Their input is incredibly important. They also tell us other things they need. They will frequently say, " We're so glad to have clarity around how these solutions are performing. Now we need to be able to contract better, so then we can set up teams to help them collectively figure out how to do better performance- based contracting." The Purchaser advisory council is very important to our work, but once we get into an assessment that is really a very tight team, both for folks that are working at PHTI and our close partners who include health economists, health economics, and outcomes researchers and other partners.
Ran Strul: How do you know that you made an impact at the end of the day?
Caroline Pearson: Well, we are early in our work, and I certainly hope that we'll be able to measure our impact more quantitatively over time. For now, we are very excited to see so much engagement from the market. Every day we get an email from a new purchaser saying, " How can I get access to your reports? Will you help me with this new area? Can I join the purchaser advisory council?" And it just speaks to the demand from the market. We also get a lot of interest from investors. Investors drive a lot of focus about where we're going to continue to innovate and how their portfolio companies choose to spend their money. And increasingly, investors are saying, " Help us think about how you do these assessments so we can incorporate that into our diligence process." And I think that's another indicator of how people are really changing the way they think about making these decisions. And then lastly, that engagement from the companies. I go to conferences and I am constantly having companies walk up to me and say, " Will you please assess us? We believe that we have great clinical outcomes. We believe that we're delivering economic value and we want help telling that story to the market." And again, that's a great sign for us that there is really a increased emphasis on clinical outcomes, economic impact, and generating strong credible evidence to support those findings.
Ran Strul: I think you are touching on a point I've heard from your team in the past of matching supply and demand of these solutions, and I think especially during the pandemic years and when capital was cheaper and the market was flooded, and I think even to this day, it seems like there's more supply and the solutions are more supply driven based on what maybe a founder has an idea or what capital markets are pushing versus hearing from the purchaser and what they need. And sometimes there's a match and sometimes there isn't, but you feel like you are helping to shape and match what is coming to the market with what is actually needed in the market.
Caroline Pearson: That is really our goal. Our goal is to make sure that companies are being developed and built that meet purchaser needs and that serve the health and well- being of patients, and that those companies that are successful can both raise capital from investors and win contracts and revenue from their customers. And that is really the market shift that we're trying to achieve. I think to your point, post- COVID, we saw so many health plans, employers and health systems that needed to rapidly adopt technology, and that jumped the field of healthcare forward many decades. It wasn't necessarily the most well thought through and planned approach, and so we've had a lot of point solutions get adopted. There was a lot of capital in the market at that time, and now is just the chance to take a step back and be a little bit more prudent and thoughtful in thinking about what do we want out of our health technology? What are the biggest problems we're trying to solve? How do we help achieve those goals?
Ran Strul: So we would be remiss having conversation on health technology these days without mentioning the magic words of artificial intelligence. You can probably, even since 2023, you can probably see a growth I would expect in the share of solutions that are AI based. It comes with its own set of issues as everyone knows in terms of how explainable and how clear is the correlation to an input and output. How do you guys handle that at PHTI when AI- based solution is being evaluated?
Caroline Pearson: Well, one of the things about PHTI is that we sometimes ask delightfully simple questions. When we say, " Does it work and is it worth it?", what we mean is we're focused on the end outcome and not all of the mechanics of how a solution gets there. And so that means that whether you are using AI or any other back- end technology, we are most focused on the outcomes that you're producing for your users, for your patients. And so we really don't get deep into the business of evaluating the models, checking for bias, assessing the training data on which they are built. Those are all very important. Many people are looking at it. That's not our focus. We are really looking at how is the AI producing a better experience or clinical benefit for the patient? And on the administrative side, how is any AI based solution integrating into existing healthcare workflows and changing the way they function? Let me tell you an example. We put together what we called an AI task force, which had 12 large health systems and most of the AI vendors that are selling ambient scribe technology right now. We talk to them about what is going to be a very rapid adoption of ambient documentation tools to help reduce note writing burden. As these solutions are rolling out. You hear a lot of discussion about the importance that they're going to play in terms of driving efficiency and productivity for doctors who are overburdened with paperwork. But when you talk to the systems about their experience and what their decision- making is being based on, they tell you the doctors love it, the doctor experience is really good. The patients are happy because the doctors are turning around and making eye contact. We're seeing great outcomes. And we say, " Wait, wait, wait. Let's take a pause and let's think about what are we trying to achieve and how are we going to measure it." If your goal is to reduce provider burnout, that's one set of measures. If your goal is to increase productivity or top line revenue, that's a different set of measures. Why don't you think about setting up your systems to make sure you can track that data against whatever it is that your goals are. But that's the bottom line. What are you trying to achieve and how will you measure it? Question that we tend to ask when we're looking at anything, whether it be AI or a simple text message.
Ran Strul: What assessment has been completed? Let's just get quick inventory for those who listen, and importantly, what's on the horizon? What can we expect?
Caroline Pearson: Well, we think about our work in two big buckets. One are the technology assessments that we produce. So far there have been four. Those include diabetes, musculoskeletal, hypertension, and anxiety and depression. The next assessment that will come out will be focused on virtual care for opioid use disorder, and then we'll announce future topics here pretty soon. The other set of work that we do is more thematic, and this is really because we are out talking to folks in the industry. People say, " Help us with big key issues that are emerging." That's where the AI taskforce work came out of. Our next big project on that is really focused on performance- based contracting. So consistently, our reports have made recommendations to say, one of the ways that we're going to get the best outcomes from digital health is to make sure that we are paying for performance, not just paying for engagement or paying for users. Everybody agrees that that's where they want to get to. And negotiating those contracts is really hard. We've brought together a big working group to try to create some standard definitions, some best practices, and some contracting templates, and all of that should be out either late this year or early next.
Ran Strul: And all of that available on your website, phti.org.
Caroline Pearson: Exactly.
Ran Strul: Caroline Pearson from PHTI, thank you so much for joining us.
Caroline Pearson: Thank you so much for having me. It was delightful, Ran.
Matthew Weinstock: Thank you for listening to the Oliver Wyman Health Podcast. This podcast is brought to you by the global management consulting firm, Oliver Wyman. For more insights on the business of transforming healthcare, visit our online publication health.oliverwyman.com.
