Speaker 1: Welcome to HIMSS TV's deep dive on population health. The term population health is used often in healthcare, but for such a widely discussed concept, priorities and approaches to pop health strategy, and even definitions of the term itself, can vary widely among provider organizations. Is the goal the successful management of chronic conditions? Is it about improving health outcomes, or about effective tracking and metrics for at-risk patient groups? Is it about outreach to underserved patients and addressing the fundamental social determinants of health and wellness? And how does population health square with a more traditional definition of public health, especially during a crisis like the COVID-19 pandemic? Studies have shown that perhaps just 10% or 20% of health outcomes have to do with medical care, with the other 80% attributable to factors such as personal behavior, physical environment, and genetics. Increasingly, healthcare organizations have come to understand that success with value-based care, improved access and experience with better outcomes at lower costs, depend on looking beyond more episodic care and toward a more holistic view of patient health.

Speaker 2: Population health is a game changer for us in healthcare. Why? Because we are very good at when you need to be seen, you come in and you see your physician, you see your nurse practitioner, you see your care provider. You come in for an issue, they address your issue, you go home. We're very good at the care we do when you come in to see us. And what population health said was, that's not good enough. You're accountable for their care even when they don't come in to see you. That mother who's 45 hasn't been in for her mammogram, we need to engage her and get her involved and get her mammogram done. We're going to hold you accountable to that.

Speaker 1: Effective population health management is essential to any success with value-based reimbursement, and advanced predictive analytics are key to realizing that goal. Increasingly, artificial intelligence, able to mine longitudinal data and automate many information management tasks, is redefining approaches to an array of population health challenges.

Speaker 3: Health systems are interested in defining populations in different ways, so they want to either look at all the people that they manage care for or all of the people who have a certain type of disease. But then we're also working with governmental public health agencies to look at the community level, as you mentioned, or both at the state health department level or the CDC. We also work at Regan Street with ministries of health in other countries to look at population health in that context. We try to automate as much as possible, because we think that automation will reduce burden on providers, so the people who are collecting the data, as well as reducing burden on those who are sort of receiving the data and trying to analyze it. So we're trying to build sort of dashboards that will auto-populate with data that's collected from routine care delivery, as well as pulling in data from other large data sets, so like the American Community Survey or the Behavioral Risk Factor Surveillance System data, so that we can automate the delivery of that data into dashboards, which will reduce burden on those who are trying to sort of look at population trends instead of working with individual data sets and trying to munge and manage those individually, right? Which is a lot of work, and that's how we did things 20, 30 years ago.

Speaker 1: AI and machine learning have helped put healthcare at a pivotal moment, where providers can now capitalize on the IT infrastructure investments and data digitization of the past decade plus and gain new insights about how they should deliver care for their patients more effectively and more efficiently.

Speaker 4: I'm incredibly excited about where I think we stand right now on the precipice of being able to make really fundamental changes in population health. If you think about how technology has evolved in healthcare over the last several decades, many of the problems that we've sought to solve have not been able to be challenged until now. Machine learning models think much in the same way that humans learn, and in order to solve for big population health challenges, that's the kind of power that we need. One of the biggest challenges that you face is trying to distinguish sick from not sick, toxic from not toxic, right? It's a classic sorting problem. As a medical student, you see these patients one by one, and you begin to figure out the tiny differences in signs and symptoms that help predict which course they're going to take. My challenge as a doctor is I can only see one patient at a time, so I'm always a step behind disease. That's where AI changes the game.

Speaker 1: At its most basic level, of course, population health is not about technology, but about people. More and more healthcare organizations now understand that successful outcomes start with addressing social determinants of health, such as food security, home environment, and access to transportation.

Speaker 5: Three and a half million Americans were missing appointments every year due to a lack of transportation, and that missed appointments would have significant downstream costs on the entire system. So it was estimated that around $150 billion were wasted every year due to those missed appointments. So if you take a dialysis patient, for example, if they have to go to treatment every Monday, Wednesday, and Friday, and they missed that appointment on Friday, their mortality rate goes up by 30% over the next six months, and it's estimated to cost around $200,000 to $300,000 in a catastrophic ER admission. And so a single transportation event could ultimately have significant downstream cost savings there.

Speaker 1: Now more than ever, it's time to rethink old models of episodic care and embrace a more proactive approach to help ensure health and wellness when patients are living their lives away from the doctor's office.

Speaker 6: Those of us in health care, especially doctors, like to sit in our clinics or in our hospitals and wait for patients to come to us. And so some of it is kind of flipping that on its head and going outside of our clinic or hospital walls and delivering care in communities or out to the people. Most patient populations, particularly my patient population, don't just need health care. They need physical health care, mental health care, other social services, housing, food, transportation, and sort of taking responsibility or thinking about how we can take shared responsibility for all of those pieces and connect the dots to get people the care, holistic care and services they need.

Speaker 1: Going forward, some of the most important players in helping these new care models evolve and succeed will be the patients themselves. And equipping them with the right tools, devices, and apps to help them monitor their conditions at home and relay that data back to clinicians will be essential to helping them play their part. Technologies to help provider organizations aggregate, analyze, and act upon that patient-generated data are just as important.

Speaker 7: A couple of really unique things that we do. Number one, we focus really closely on patient-reported outcomes. So we continuously survey from before a patient arrives all the way through their discharge process and then when they go home. We then follow up many iterative times after that. And what's interesting about that is you oftentimes catch patients in moments when they need you and they maybe haven't been able to vocalize it. I'll give you an example. Our psychiatric services often catch a patient when they're depressed, after they've had a procedure or whatnot, and they don't know who to talk to, who to reach out to, and we're able to have specialists intervene and say, hey, we got you, we're here to help. It's also amazing that for patients that come in and maybe need to be seen by our musculoskeletal group that they often have psychological issues around that as well. Maybe they think the worst has happened to them. They call it catastrophic-type planning. And so we're able to intervene with multiple types of services, again, because we constantly keep the patient at the core of the whole process and are constantly surveying and looking at data, looking for trends, and looking for analysis.

Speaker 1: More broadly, population health depends on addressing basic building blocks of safety, security, and behavior. It means refocusing attention away from health care and towards lasting approaches to better health.

Speaker 8: Too often we focus on health care and termite health. And so we've been stepping back from that and really thinking about what does that environment more broadly look at and how do we engage. And that requires rethinking a lot of the models that we have. It requires sharing data across entities that haven't in the past necessarily shared data. It requires really putting the person at the center rather than your programs and departments.

Speaker 1: Health care was already navigating big changes to how it manages the health of patient populations even before the coronavirus crisis shifted attention to some fundamental challenges of global public health. Now, with the pandemic having put the spotlight on valuable new technologies such as telemedicine, remote patient monitoring, AI chatbots, advanced analytics, and data visualization tools, patients and providers worldwide are looking toward a future where these and other approaches can work together to help achieve better and more equitable health outcomes. Thank you for watching HIMSS TV's deep dive on population health. This is Health Care IT News Executive Editor Mike Milliard signing off. ♪