As a self-professed “tech enthusiast” and an early adopter of telehealth, Michael Hasselberg is excited about the impact large-language models can have on healthcare.
But as a psychiatric nurse practitioner, he believes the real potential is in easing the burden on clinicians. By helping with administrative tasks such as authorizations and denials, as well as documentation, chatbots can free up providers to “do what we went into healthcare to do, which is to work with patients.”
AI-driven tools also have the potential to dramatically improve care in rural communities, which has long been a passion for Hasselberg. During an interview with Kate Gamble, Managing Editor and Director of Social Media at healthsystemCIO, he talked about critical role 5G networks play in the success of hospital-at-home models; how URMC is facing the daunting “supply and demand issue” with behavioral health; and why, even if the most sophisticated technology is available, it’s important to remember that “healthcare is local.”
LISTEN HERE USING THE PLAYER BELOW OR SUBSCRIBE THROUGH YOUR FAVORITE PODCASTING SERVICE.
- Hasselberg’s digital health journey started more than 10 years ago when he helped develop a statewide telemedicine program for psychiatry in New York. “I was really passionate about how to get care out to these communities.”
- While the transition to telemedicine wasn’t much of an adjustment for patients, it was a significant culture change for clinicians, “who were not trained to leverage digital modalities to interact with their patients,” he said. “We had to disrupt a lot of our operational workflows.”
- After using telehealth as a “band-aid” during the first two years of Covid, URMC had to “create better access points for patients in their homes and in their communities,” which meant “really cleaning up our data.”
- Although Hasselberg is encouraged by the proliferation of digital startups in the behavioral health space, it’s critical to remember that “healthcare is local, and we need to think about how local trusted health systems can develop these types of services.”
- “I’m even more convinced after playing with ChatGPT that it will not take the place of a clinical provider,” said Hasselberg. “It’s going to save clinical providers from leaving the profession. That’s where this technology is going to be most impactful — taking things off our plates that no provider wants to do.”
Q&A with Michael Hasselberg, PhD, Chief Digital Health Officer, University of Rochester Medical Center
Gamble: Hi Michael, thanks so much for speaking with me. The first thing I want to do is get some high-level information about University of Rochester Medical Center. Can you talk about where you’re located, kind of how you’re structured, things like that?
Hasselberg: University of Rochester Medical Center is located in Rochester, New York. It’s the upper part of western New York State, right below Lake Ontario. We serve a pretty large geographic region; we’re all the way out into the central New York area, almost to Syracuse, all the way down to the Pennsylvania border and creeping west out to the Ohio border.
Our patient population is quite diverse. Like any midsized city, we have disparities; if you go 20-25 miles outside of the city limits, you’re in rural farm country. We are an academic health system. We’re a little bit unique in that lens; we’re only one of a handful of academic health systems left in the country that is still fully integrated with its parent university. What that means is the health system’s budget actually rolls up to the university’s budget. Most academic health systems in the country have actually broken away from their parent university and have a separate budget but just have an affiliation relationship with the university.
URMC’s “true innovation incubator”
We’ve actually made a conscious decision at Rochester not to break away. From a digital innovation standpoint, we actually view that as a differentiator. We have a true innovation incubator for the health system that’s supported by the university. It’s not a research shop; it’s truly to innovate, to solve strategic gaps, to help our health system transform so we can better support the communities around us. Within this incubator, we have faculty from the engineering school, the computer science department, and data science institute combined with faculty from the business school. Even our school of music — we’ve got music faculty in the incubator under the same roof as faculty from the medical school, dental school, and nursing school; those faculty also serve as frontline clinicians in our clinical services.
As an innovation team, we build new technologies, everything from apps to chatbots to VR and AR applications. We build sensors for monitoring. We do a lot of machine learning; most of the machine learning that touches our clinical service lines comes through the incubator. We partner with start-up companies, typically series A or later. We put our arms around those companies and support them around integrating with our operational and clinical workflows.
Thinking differently about value
Our developer team does the direct integration into our EHR if needed. Oftentimes our business team, in partnership with our clinical leadership team, is defining new value propositions for these companies, where the business model for which they raised money on in the VC space is not the same business model that we would purchase for the health system.
We help them think differently about where their value is. That, in a nutshell, is who we are at the University of Rochester. We’ve got nursing homes. We’ve got a home care service line. We’ve got multiple community hospitals throughout our region. We’ve got a full ambulatory service with primary care and specialty services. We’re the fifth largest employer in the state of New York and probably the largest health system in the state outside of New York City.
Gamble: So, certainly a lot going on there. And you have the role of Chief Digital Health Officer. Can you talk about what that entails?
Hasselberg: Absolutely. First and foremost, I’m a clinician. I’m a trained psychiatric nurse practitioner who went on to get a PhD and then a business degree. My claim to fame early in my career, about 10 years ago, was developing a statewide telemedicine program for psychiatry. I got involved with telemedicine before it was a cool thing or even reimbursed, and that was driven by my own personal ‘why.’ I started my career in some really rural communities of New York State where I was the only psych provider for something like 6 counties.
I was really passionate about how to get care out to these communities with so many social determinants of health barriers preventing them to get that access, and so, I got involved with the telehealth world. It was very successful there, and when Covid hit, just like every health system in the country, we had to turn on telehealth overnight. I was kind of tasked to work with some of our operational leadership to stand up telehealth for the whole health system in about a week.
About 6 months into COVID, we realized it wasn’t going away and we needed to digitalize pretty much everything we do in the health system, but we didn’t have a clear strategy on how to do that. We were getting stretched really thin because so many departments needed to leverage digital technologies, but we didn’t have the resources to do it.
And so, I was asked by our CEO of the health system to take on the role of Chief Digital Health Officer to essentially lay out the digital transformation strategy so we could prioritize all of the requests in a coherent way so we could go from care-as-usual to becoming a digital-first organization. It’s a new role for my health system and it was really COVID that kind of ignited this position to be established.
Managing Cultural Change
Gamble: I’m sure that it was advantageous that you had experience with telehealth before Covid. But of course, that time period was really challenging.
Hasselberg: Extremely challenging. It was a lot of culture change. The culture changed for our clinicians, who were not trained to leverage digital modalities to interact with their patients. It was also a lot of culture change on the operational side. We’re a health system that is primarily fee-for-service based in regard to reimbursement; that landscape doesn’t set up really nicely for digital health. We had to disrupt a lot of our operational workflows that work well when we see patients in person, but maybe do not align well for digital modalities.
For our patients, that culture change was maybe a little bit easier than it was within the walls of the health system, because all of us were forced into video conferencing in order to interact with each other. Our patients, in my opinion, were able to tolerate the change of receiving care through digital modalities better than those of us trying to deliver and operationalize that care within the health system.
Gamble: When we were about 6 months into Covid and it became clear that this form of care wasn’t going anywhere, how did your strategy change?
Hasselberg: The first few years of COVID were way beyond telehealth. Telehealth was a band-aid initially; we knew that we had to create better access points for patients in their homes and in their own communities to get connected with us. We focused heavily on the digital front door in our health system those first two years, and prioritized our patient portal.
“Cleaning up data”
Prior to Covid, our patient portal penetration in our primary care service lines was not very high. We had maybe 30 percent of patients utilizing the portal; we had to get that penetration up to like 90 percent pretty quickly. There was a lot of emphasis on making sure that folks were educated around the portal. We turned on as much functionality as possible.
We then had to start really cleaning our data up because we needed to make it easier for patients to schedule appointments with us. The old model of the call center really didn’t work before Covid, and it became a heck of a lot worse during Covid. We needed to stand up online scheduling and make those contact points easier and then we had to shift to the check-in process.
The old way of the patients coming in to the office and handing their insurance card and filling out forms around their demographics and doing questionnaires — we had to make sure all of those were in a digital format. After patients scheduled appointments on their own device, they could check in on the appointment. We could collect all of that data from the comfort of the patient’s home.
On-demand telemedicine platform
We actually had to pivot back to telemedicine after a year or so into it. By that, I mean that early in the pandemic, we were able to set up workflows for scheduling telemedicine appointments, but we needed an on-demand telemedicine service line where patients that have maybe an urgent care need can click a button and be connected to a trusted provider at the University of Rochester to deliver that care.
We had to stand up an on-demand telemedicine platform. Also, a lot of folks’ insurance was significantly impacted during Covid. A lot of people were unemployed and lost their insurance but still needed to receive care, and so we needed to create almost a retail option to allow patients that either didn’t have insurance or had really high deductibles to have a transparent price for on-demand telemedicine service.
Then, to finish out our front door rate regarding payments, we needed to create an Apple Pay experience where you can access your bill and click a button to pay it. That was the major emphasis those really first couple of years within Covid; to stand up easier access points using technology.
Analyzing telemedicine data to “understand the myths”
Gamble: It seems a lot of the focus at University of Rochester has been on the underserved population and leveraging telehealth to improve care.
Hasselberg: 100 percent, and we learned a lot. One of the things we were really proud of was that we did a really robust analysis specifically of our telehealth data during Covid. We honed in on our Medicaid population, which tends to be your more vulnerable patient population. We wanted to understand whether the myths around telemedicine reimbursement were true. A lot of payers made the argument that telemedicine would increase cost. There would be lower quality care and patients would end up in the ED more often or require more expensive imaging or lab tests or more follow-up visits. We wanted to understand if that was true.
What we had found was that our Medicaid patient population benefitted the most from telehealth. They had lower no-show and cancellation rates. They didn’t end up in the ED more often than other Medicaid patients who didn’t utilize telehealth. And actually, they required less expensive labs, imaging and follow up. We were really proud to publish that paper in the New England Journal of Medicine Catalyst back in October of 2022.
The digital divide
One of the caveats that we found specifically in our rural communities was that although our patients did really well with telehealth, the majority were engaging in audio-only visits during Covid, and not using the video component. Initially we thought it was a broadband issue; that there wasn’t internet access in these communities. However, that was not the case. New York State has done a really good job of investing in the broadband infrastructure.
We found that these were high poverty communities and folks couldn’t afford the internet in their home. The only internet access they had was relying on the data plans on their cell phones and we all know that video is very data intensive. Folks weren’t willing to use their data plans to do a video interaction. They interacted with our providers just using audio only, and we know that when the public health emergency ends, Medicare has already said they’re not looking to continue to reimburse audio only telehealth, and commercial insurers will probably follow that.
Now, we’re trying to think outside of the box around how can we engage this population in video telehealth; how can we create new access points to telehealth in these patients’ communities on their daily journeys so they can receive care in a place that’s already has broadband enabled. We’re also looking at other opportunities like SMS texting applications and mobile applications that don’t require significant data plan usage.
We’re trying to now think outside the box so that we can continue the great results that we achieved during Covid, but in a new reality of pulling back some of those flexibilities that happened during the pandemic.
Behavioral health supply & demand
Gamble: When you talked about the need for behavioral health providers, it’s staggering. And I know that there’s a lot of trepidation around what’s going to happen with reimbursement, but this has made a huge difference in the mental health field.
Hasselberg: It really has. Being a behavioral health provider and especially starting in rural communities, I can say firsthand we have a supply and demand issue. The demand for behavioral health services is so high — it has only become higher since Covid started, and we just don’t have the supply to meet that demand. The reality is we’re not going to double our supply anytime soon. It’s just not going to happen. We’ve seen a proliferation of digital health start-ups in the behavioral health space. We’re seeing companies like Lyra Health and Modern Health come in to try to fill these gaps, but I would argue that healthcare is local. We really need to think about how local trusted health systems can develop these types of services for their community and not be so reliant on these national providers who may not know the nuances of the local community or be fully connected to the other resources available within those communities.
We’ve really put an emphasis around innovation on the behavioral health side. And moving beyond telehealth, starting to think about how can we leverage technologies that may not be so reliant on the need for a clinician on the other side. For example, how can we use chatbots or mobile apps to provide some of these services, knowing that there’s are always going to be the need for a higher level of care for some patients, and open up the care capacity for those patients to be triaged. I’m super excited about what is happening in the world with large language models in the AI space and I think that technology could be really transformative specifically for behavioral health.
ChatGPT & clinical burnout
Gamble: It’s definitely something we’re starting to hear more about. Tools like ChatGPT have so much potential there to help ease some of the burden on providers.
Hasselberg: As a tech enthusiast, I’m in awe of these large language models. I’m having so much fun playing with them and trying to think of new creative ways to use them. I am 100 percent convinced that AI will not replace the clinician. Where I’m excited about this technology is that it’s going to save clinical providers from leaving the profession. I believe that where this technology is going to be most impactful — taking things off our plates that no provider wants to do.
Things like insurance denials and prior authorizations and so many of these administrative tasks, even documentation — this technology could potentially solve those issues for us. That way, we can do what we really went into healthcare to do was to work with the patients and talk to the patients, and maybe also keep ourselves well because we can actually finish our clinical days or our work shifts on time. We won’t have to worry about going home to then catch up with all of our documentation. We can spend that time with our loved ones and get our own respite.
That’s where I’m most excited about the future of AI in healthcare — not that it’s going to replace that therapeutic relationship between the provider and clinician, but that it’s going to support the clinician to make that relationship even stronger.
Gamble: I can tell that you’re really passionate about this. I think one of the biggest barriers is fear — more specifically, fear of change. That’s certainly something that has plagued healthcare.
Hasselberg: Yes, and we need to change. Healthcare is broken in the United States. We absolutely need to disrupt healthcare. I’m super excited about the new market entrants in healthcare, the Amazons of the world, the Wal-Marts, the Microsofts, and the Googles. I’m excited that retail is investing heavily in healthcare because that’s what is going to push the traditional incumbents like myself forward. It’s going to push us to be disruptive and to innovate. I love that other verticals and other industries are investing heavily in the healthcare space. It’s needed in the United States.
5G’s enormous potential
Gamble: Right. So, the last thing I want to talk about was 5G and the potential there within healthcare.
Hasselberg: I’m intrigued. 5G is a technology that I think could be totally transformative to healthcare and be really disruptive. The idea of low latency, high frequency data transmission, and edge computing could really push healthcare forward. Where I’m really excited about the potential of 5G is outside the walls of the hospital and I know most of the 5G work has been done within hospitals, but where this is going to be really beneficial is in those rural communities that I’m passionate about.
The ability to leverage all of these IoT devices that are in patients’ homes and be able to get that data back to the hospital and to the clinicians in almost real time to make assessments — that will change the world in regard to how we conceive the walls of the hospital. It will really set that idea of hospital at home up to be successful. I think there’s a lot of opportunity that 5G is going to bring into healthcare, but it’s still really early. We’re still trying to figure out what is that killer use case.
Again, it’s exciting to see companies like Verizon, AT&T and others that are moving into healthcare and leveraging their tech stacks to see if they can help fix this healthcare problem we have in the US.
Gamble: It’s very exciting, I agree. Well, I really appreciate your perspective and I think you’ve definitely given our audience a lot to think about. Thank you so much for your time.
Hasselberg: Thanks for having me, Kate. This was enjoyable.