"It’s a myth that delivering care digitally will result in higher costs"

Q. Michael, can you talk about the URMC and what makes it unique?

Michael: We’re actually even more unique than most academic medical centers left in the country these days in the sense that our health system is still truly fully integrated into our university.

What that means is the budget on the health system side rolls up to the budget of the university. Most academic health systems, today, are no longer fully integrated with their parent university in that they have broken off from the parent university. On the health system side, we tend to make money but on the academic side, it’s much harder to do so. The health system then, ends up subsidizing and sending a lot of their margins over to the college to help support those missions.

A lot of academic medical centers said, “Hey! If we broke away from our parent university, it’s going to be easier for us to obtain our 1-2% annual margins per year that we’re trying to achieve.” At the University of Rochester, we have made the conscious decision that we are not breaking away from our parent university and we actually leverage that as a differentiator for us.

When we think about digital health and digital transformation, I have access to some of the most brilliant engineers, computer scientists, data scientists, business faculty in the country. I have access to even the faculty from our music school. I can apply that expertise and capacity to solving some of the most difficult problems in our health system. I can leverage that expertise to build, create, and deploy new technology solutions into our ecosystem. It’s a unique place and I love it.

Q. Does it also influence your priorities regarding the kind of digital health initiatives you should be in, in addition to serving? Also, tell us about the populations you serve.

Michael: We serve a very diverse patient population. To give you some context around the URMC and Health System, we’re the largest health system outside of New York City. In terms of geography, we have a large geography in the state of New York from central New York all the way out to the Ohio border and all the way down to the Pennsylvania border. This entire region has patients that we serve.

In terms of the kind of diversity of these patients, we have everything from the inner city of Rochester, which looks like the inner city of most moderate sized cities across the country. But if you go 20-25 miles outside of the city, you could be in some of the more rural areas in the States or in the country.

A good portion of our patient population is safety nets and Medicaid. We have a lot of underserved and vulnerable patients that seek care out at our academic medical center. For those reasons and in trying to engage and reach those patient populations, we’ve had to think outside the box and other technology solutions to, not only meet the needs of patients in the inner cities but also meet those for who, there may not be a specific specialist for four counties around them. How do we get care out to them?

With regard to our technology priorities and the influence that the college has on that, actually, there’s not a whole lot of influence from the college in the normal sense. We have a very clear digital transformation strategy that’s set out. When we have gaps in our technology stack, we say, “Hey! We need to solve this problem.” If we don’t have a solution in our technology stack, we may lean on the college. That expertise—if we can’t find a solution or an external vendor that we think is best of breed to fill it in—is what we will leverage and say, “Hey! Can you help us develop the solution in-house?”

It’s not that we don’t develop technologies for the purpose of spinning out companies. We don’t have a true investment arm, so, we’re different from another one of your honorees like Providence Health, which has a $300 million venture arm where they actually incubate a lot of companies in-house. They invest in them and spin them out. We don’t do that.

When we build our technologies, they are truly being built to serve our patient population and community. We build to open source our code and give our technologies away to other health systems in the country. We have a lot of examples of doing that and of other health systems and industry coming in, taking our code, and applying it to their systems.

So, the college, I would say, kind of augments the strategy but doesn’t satay or drive it. They help us fill the gaps.

Q. What are you hearing from these populations in terms of what they want from URMC? Can you talk about the solutions you’ve developed that are technology enabled from a digital health standpoint? What kind of benefits have you delivered?

Michael: Access to health care is something that we hear across the board that these populations are seeking. So, it doesn’t matter if you’re in the inner city or if you’re in rural America. Folks want to have access.

I think one of the myths that jumps out often in the digital health space is this digital divide — that some of these populations don’t have access to the technology needed to receive care or there’s not sufficient broadband in these communities — so they cannot engage.

What we have found in Rochester, for our market and the patients we serve is, that it’s a total myth especially, in some of the more rural areas of the state. What our previous Governor of New York state did was, they invested really heavily in getting broadband access across the state so, there isn’t a problem of Internet not being out in some of these more underserved communities.

The other myth is, a lot of our patients don’t have one of these devices – the smartphone. Pretty much everybody has one of these and you can do a lot with engaging patients on that smartphone. However, as we started deploying things like telemedicine very broadly during the pandemic, what we found was especially in some of these more rural areas and with our safety net patient population, while they engaged quite a bit through the telephonic interactions with our care providers, it wasn’t very significant on the video side. We did a deeper dive in that and found that although there’s Internet access out in these rural communities, the only Internet that’s available to them is through their data plans on their phone. When you’re pushing out a video conferencing feed to somebody’s data plan on their phone, it eats up that data plan quite significantly. So, we’ve thought of and engaged them via more text-based and mobile applications and we’re thinking outside the box around how we can identify other partners and where we can meet these patients in their communities to actually deliver video-based care.

A selfish plug here — just last month we had a publication in the New York-New England Journal of Medicine, Catalyst, which actually talked about our experience delivering telemedicine to the safety net Medicaid population in these rural areas. As they engaged in care, we found out that these populations engaged more via digital modalities than actually in person. On comparing them, we found they engaged more in the digital modalities than even some of our commercial payer patients did. Not only did they engage, they required less in-person care after that engagement in their video consult. They weren’t ending up in the EDs more often than our patients coming in-person. They also required less expensive imaging and lab work than those that were coming in-person.

All these myths then, that delivering care digitally is going to result in higher costs because providers are going to lay hands on them, so, they’re going to need to order more tasks, more imaging to get the data to make those confident care decisions, is not something we saw. The idea that, because the provider’s not going to lay hands on these patients, they’re going to require more in-person follow ups because they’re not going to get their care needs met is not what we saw at all. Again, the patient population that did the best to decrease cancelations, no shows and more follow up was the safety net Patient population engaging in telephonic and video digital modalities who received care.

Q. Does this hold true for all types of care — episodic, preventative, or chronic disease management — or is it more pronounced for one type of care?

Michael: Where I think we had the most success was in primary care because our primary care sees whatever comes through the door.

Another area that we continue to have success in is behavioral health. We’re also having a lot of continued engagements and considerable success in the urgent care and emergency department settings with these modalities.

In terms of the types of digital modalities we have success in some of our subspecialty areas actually may not be telemedicine. And part of that is, digital health in some ways really disrupts their current workflows. Those workflows and more procedural based subspecialty disciplines are set up to be successful with that patient showing up in the office and being seen in-person. If you apply too much digital transformation to those subspecialty areas, it disrupts what’s working for them now.

Being in a health system that’s primarily reimbursed or still in fee for service, we have very little value-based reimbursement contracts. We really don’t want to disrupt a whole lot of our high-cost procedural based subspecialists and what they’re doing. So, in some of those areas, digital engagement has perhaps not been as strong as it’s been in primary care, behavioral health, geriatrics, and urgent care and some of our more non procedural based specialty kind of discipline areas.

Q. How have the caregivers and physicians responded to these digital modalities even if it is for primary care or urgent care or something more specific? What have they had to change or adapt to in terms of their own training, reorientation? Can you talk about their expectations and how you met those?

Michael: I suspect a lot of your listeners — the other health systems — are going to have experienced a lot of what I’m going to say. When we started our digital transformation strategy in our health system, the first two years of the strategy were primarily focused on access and on how we could essentially create a digital front door where our physical front door was located. Our physical front door is primary care. That is where we narrowly focused the beginning of our transformation strategy.

When we started in primary care, we had a significant amount of resistance around, “Right now, my caseload is falling. I don’t have room to take on any more patients. What do you mean you want me to use more technology? This electronic health record that you have for me is the bane of my existence. I am documenting all day long and answering messages from my patients and looking at labs all day. You can’t add another technology on top of this. I can’t do it.” That was a lot of what we heard in the resistance.

We listened to and understood that. We needed to help relieve some of their pain points. We realized the need for a true digital patient portal into our health system. We are an Epic shop and MyChart is the patient portal for Epic. So, we started there in our MyChart penetration. Our digital transformation was not high in primary care — in fact, it was below 30% — and we knew that if we were to engage our patients through digital mechanisms, we had to get that MyChart and that patient portal penetration up. However, the resistance from the primary care site meant they were not championing the patient portal in MyChart because they didn’t want more messages coming in. They equated that patient portal to being their in-baskets, which was overwhelming them.

To get early wins and buy in from our providers, we had to help them out and do a deeper dive into what was clogging up their in-baskets. We found some low hanging fruit here and made system level decisions of getting all that out. We were able to really quickly reduce the in-basket burden on our clinicians by 15% and all this by clicking a button in our system. Getting that win had never happened for these primary care providers in the years that I’ve been in the institution, since we’ve gone live with Epic, and so, that was huge for us.

That gave them more confidence so they said, “Hey! Let’s give this a try.” They started engaging in the digital transformation strategy and started championing MyChart. Now, our patient portal penetration in primary care has gone from less than 30% to up about 90% in a two-year period in the primary care setting. There was resistance at the beginning, but we had to get those early wins.

Along that transformation in primary care, we celebrated those early wins with our providers. We showed the benefits of, “Hey! We’re going to save you more time and free you up to do the things that you really want to do. You can see patients and not be documenting or doing the rest of the stuff.” That’s how we were successful.

We find the same kind of experience in our specialty service lines. We’ve expanded our transformation and one of the things I’m very grateful for is having a great partner in crime. Dr. Gregg Nicandri, our Chief Medical Information Officer and I are attached at the hip. He leads the clinical informatics teams.

We help with the translation, enable getting by at the provider level, and really, leaning on the clinical informatics team. Leaning in on Rosemary Ventura, our Chief Nursing Informatics Officer on the nursing side has also been really helpful to move forward this digital transformation with our providers.

Q. With regard to the technology landscape, how do you make technology choices? What are the tradeoffs you make when you consider something that is native to your EHR platform versus something that may be a standalone tool which is best in class but also has its own set of tradeoffs?

Michael: Folk that have heard me speak in other forums know that I’m, in some ways, a little bullish in my response because we’re an Epic shop. We take an Epic-first mentality. What that means is, if Epic has the functionality and it’s good enough — it does not have to be the best or peripheral but if it has patient experience or patient access functionality that’s a little bit outside of Epic’s bread and butter, then, we’ll go with the Epic solution every single time, even if there is a better solution out there. Part of the reason we’ve just made so much of an investment as a health system into Epic is because we have to maximize that investment is as best as we can.

That being said, if Epic doesn’t have the functionality or it’s on their roadmap but there’s no real clear indication of when it’s actually going to go live, which happens a lot, then, that’s when we make a call about whether this is a high-enough priority. We can’t wait until Epic gets there on the roadmap. We need to find a solution.

The way we evaluate external vendors is not the typical way a vendor may think they would get evaluated. I don’t really care if you’re the best-in-class vendor out there. My first priority is less about your success with regard to your UI, UX and results there. It’s truly about the level of your integration into Epic. If you don’t have a nicely integrated package within Epic already, you’re probably not going to make it on our list of even a vendor to consider. That level of integration is priority number one for us. If we then find a solution that integrates well with Epic’s hyperspace and with the patient MyChart portal in a way that it the patients continue to have that omni channel experience, then, we can onboard that into our health care ecosystem to fill that gap.

One of the things that’s really unique about Rochester and what probably excites me the most is we actually have a true digital innovation incubator. It’s not a research shop. It has faculty from all of our schools — the medical, dental and nursing schools under the same roof — and it uses design thinking methodologies to build solutions in-house to fill those gaps. We build them fully integrated into Epic. That’s the thought process at the URMC as we think about our technology stack and how we take on new solutions.

Q. We are going through a very challenging year in 2022. What do you see ahead for health systems? What are you planning from a digital health priorities/investment standpoint going into 2023?

Michael: Our big investment in priority is actually data. We’re collecting a lot of new data from new technologies that we’ve to had before in our databases within the health system. So, getting our data organized and in good shape is top priority.

As a large academic medical center or health system, we also have a lot of data silos and no source truth of data. It’s important then to build our enterprise data warehouse and break down those silos, bring in all of this new data from these technologies that we’ve rolled out over the last couple of years and make sense of it. That’s actually going to set us up uniquely in two different areas.

One, it’s going to help my health system make more strategic decisions around taking on risk from maybe a payer standpoint in the future. It will also get us set up nicely for moving into more value-based arrangements. That’s priority one.

Priority two is our workforce struggles and shortages. Data will allow us to understand where to start making investments in the workforce. When data is all cleaned and aggregated, we can start taking advantage of some of these Machine Learning products that are popping into the markets. A lot of that machine learning and artificial intelligence technologies that are coming out can potentially significantly impact the workforce shortages, help us start automating things, and supplementing where we have gaps in our workforce.

I have a lot of AI vendors that approach me and want to partner with Rochester. However, my response to them is, we’re not ready yet. The reason is, you may have the best algorithm or model built, but my data isn’t there, yet. If I was to roll out your model now, and I put my data in, then, the results emerging will probably not be the results I was hoping for.

We want to put ourselves in a good place to not just take advantage of machine learning and artificial intelligence in the future to help our workforce but also to help us make better strategic decisions around transformation, in general. That may be on the digital side or even on the payment side.

Q. Are you looking more at data from the point of view of a consumer data strategy that helps you improve your engagement and outreach? Or is it more from the standpoint of improving health care outcomes? Or, both?

Michael: It’s absolutely both. One of the things that excites us is on the outcome standpoint and actually merging that data. We’re very proud of the fact that Rochester’s the home of the bio psychosocial model of medicine and it’s all about focusing on care from these holistic and broad domains.

We made a strategic decision about seven years ago to profile our patients using patient-reported outcomes within those broad domains. It didn’t matter if you came to my health system with a toenail injury but we were going to ask you about your emotional distress, physical functioning, pain interference, and social functioning every single time. We collected this data on iPads and integrated it right into Epic. My health system has, as far as we’re aware, the largest patient reported outcome data set in the entire country. All this is systematically collected in these broad domains. Now we’ve got these outcomes based off the patient’s own perceptions and in their own voice about how they’re doing in health care. We can combine that with some of the more quantitative data from the EHR about lab or mortality outcomes and these newer consumer engagement data that we didn’t otherwise collect using technology. That’s going to be the secret sauce.

When I think about disruptors, in general, in other verticals, Amazon comes to mind. They totally disrupted retail, but Amazon didn’t do it because they were setting up an e-commerce website. It was the data behind that. Amazon knows you as a consumer better than you know yourself. That’s where we want to get to in health care. I want to be able to help predict what you’re going to need as a patient before you know you even need it and get you to the right level of care at the right time. We think that combination of patient reported outcomes collected in these broad domains combined with our EHR data combined with this new consumer data that we’re getting from technology combined with claims data and others will help us at Rochester develop our own Amazon recommendation algorithm that they’ve patented. We’re going to do it for health care. That’s where the future is going and that’s why we’re now heavily invested in getting our data to a point where we can start leveraging our Data Science Institute and the college and some of these AI vendors to help us get there.

We hope you enjoyed this podcast. Subscribe to our podcast series at  www.thebigunlock.com and write to us at  info@thebigunlock.com

Disclaimer: This Q&A has been derived from the podcast transcript and has been edited for readability and clarity.