- Value-based care can improve outcomes and lower costs by getting patients the right care, at the right time, in the right place.
- Artificial intelligence (AI) is increasingly being used to automate critical business functions and support clinicians in making complex clinical decisions.
- AI tools created to reduce administrative burden could have a big impact on reducing clinician burnout and rekindle their love for the profession.
For more on this topic:
- Learn about what a partnership with PINC AI™ Clinical Decision Support can do for your healthcare organization.
- Read on for the full discussion, or listen to the full podcast below.
Recently, Ryan Nellis, Vice President and General Manager at PINC AI™ Clinical Decision Support, had the opportunity to sit down with Brian Zimmerman of the Becker’s Healthcare Podcast to discuss the role of AI in value-based care and patient engagement.
Brian: I want to begin with a question about value-based care. Ryan, how are payers and providers collaborating to navigate the shift to value-based care, and what are some challenges and areas of opportunities?
Ryan: Value-based care is about getting better healthcare for patients at lower costs. Some of the challenges historically were around misaligned incentives. For the last several decades we have been in a fee-for-service (FFS) environment where providers and healthcare administrators have been incentivized to do more tests, do more services, regardless of the costs. That environment is shifting to focus on patient outcomes, and managing the costs is difficult. However, there's been some new evolving payment models that are presenting opportunities for the community.
We're amidst a great shift in how Medicare, the largest payer in the United States, pays for healthcare. They've been shifting patient payments towards the Medicare Advantage (MA) programs, the bono programs and recently announced a new reach program around direct contracting, which is very similar to MA, and all these payment models are capitated payments. So, what they do differently is they pay providers up front, and they pay them well to treat the patients over the course of a year, and they make these payments based on the patients’ illness burden.
For example, if MA is only providing $10,000 to treat a patient, but the care actually costs $12,000, then the provider could actually be at risk for the extra cost. So, these new payment models focus on spend reduction and cost reduction, and that's the negative side of things. The positive side of things is these payments in advance, they really incent the provider to not just treat patients when they're in front of them and potentially acutely ill, they really force the system to think holistically about the drivers of cost. These payments up front can be used for mental health treatment, food insecurity, social determinants - some of the drivers of the underlying reasons that health costs have been driving up. Ultimately, I think this is really just good for patients. It's good for healthcare and it's good for the system. So, while I think there are still some challenges, things are looking bright with aligning incentive models.
Brian: As you pointed out, the mission behind value-based care is of course to lower cost, and then also it is focused on outcomes. That goal really points to a challenge that's been pervasive in U.S. healthcare for a long time, which is that healthcare in this country is very expensive, yet our health outcomes are behind other countries. Why do we spend so much on care and not see better results?
Ryan: I really think a lot of it is what I mentioned earlier around misaligned incentives. We were predominantly FFS for many decades. I was just listening to another podcast, and they were talking about cascades of care, and it's well known that there's waste in the system. It's been studied and it's been studied, and one of the things that has been going on is that unnecessary tests and unnecessary treatments lend themselves to more unnecessary tests and more unnecessary treatments. This can cause costs and patient anxiety to snowball.
A lot of time, the advantage of these tests doesn’t outweigh the potential for patient harm. So, there's been this Choosing Wisely movement, led by the American Board of Internal Medicine, that has created lists of guidelines focused on clinical tests we just shouldn't do, because again, the harm outweighs the benefit like unnecessary lab tests and unnecessary medications.
An easy example is the Choosing Wisely guidance that focuses on the use of benzodiazepines (heavy sedatives) in frail elderly patients. Studies have shown patients are taking these heavy sedatives and are experiencing more falls that lead to hip fractures and increased pain and cost for the patients, outweighing the benefit they receive from the sedative. We've been trying to help providers through our technology platform understand where these opportunities to reduce harm and waste are occurring in real time, in electronic health records (EHRs). What's nice - to kind of bring this full circle - is some of these payment models around capitation are now sort of forcing the system to pay better attention to reducing harm and waste.
Brian: Let's jump to AI and bring that into the conversation. Ryan, can you talk about how artificial intelligence is playing a role in delivering value in healthcare today?
Ryan: AI is working today, and it's being used. We've looked at several studies, and in 1950, the amount of information available to clinicians doubled every 50 years. So theoretically, if you needed to know everything about how to treat patients well, you would probably graduate medical school and be good for your career. The amount of information at our disposal has exponentially doubled. In 2021, it was estimated that the amount of medical information was doubling every 60 days. Clinicians just can't keep up; the human brain can't keep up.
One of the things we're doing is digitizing thousands and thousands of medical guidelines and paper, and helping clinicians use that in real time to better treat their patients. Another challenge is just the documentation around these patients. It's one thing to be able to have access to the medical knowledge, but you must know how to apply that smartly to patients. It's estimated that 60 percent of the documentation of patients in electronic health records are documented, on structured fields, blurbs and blobs of texts that are not easy for computers to get in and process and compare patients to these evidence-based guidelines. So, we've been spending a lot of time building out natural language processing to be able to go into these electronic records and understand with high confidence what's going on with patients. We then apply these clinical guidelines that have been digitized to help really nudge clinicians and be able to process this information in real time. Those are some of the things we're doing today, and there are other examples in market.
Brian: If you're having any conversation about the state of healthcare, workforce issues are certainly going to come up right now. I think when I first started reading and writing about healthcare some years ago, there was a lot of talk about the coming clinician shortage. Now, due to the pandemic and just the many challenges clinicians have faced in the last couple of years, there's a lot of folks leaving the field. It feels like that shortage that was on the horizon is really here. How much of an impact can AI have on really helping to bolster the healthcare workforce?
Ryan: AI can have a huge impact on the healthcare workforce. You know, clinicians got into the practice of medicine to help treat people, help treat patients. I don't think anybody intended when they went to medical school that there would be an intense amount of data work and paperwork as a part of the job. So, one of the things we're doing is going into those EHRs and being able to understand, with AI, where complex workflows exist. We're listening to the clinician reaction to these workflow events, we're reading their responses and we're able to help health systems take the burden of just getting through their day, their clinical process and their EHR better. I think many organizations are focused on using technology to reduce the burden associated with patient treatment.
Then there are other things like prior authorization which gets a lot of attention. If you think about it, the start of something that requires a prior auth is a clinician wanting to order something or treat the patient with something that's likely expensive. What they typically do is order something in the EHR. Then they or their staff must go to a separate portal and reenter everything that's already in the EHR into a second portal. That’s when other clinicians who work for insurance companies look at the information and transcribe it into the portal. There is tons of waste around that whole process. So, one of the things that we've been doing with AI and automation is actually being able to go in and capture those expensive orders in real time and look into the patient's chart around what's going on with the patient. Make sure that it meets medical necessity requirements in real time, and when possible, tell clinicians in an instant that the expensive thing that they want to do that's necessary for the patient is approved. The correlating benefit is his/her patients get access to quick and necessary treatment. So, at the end of the day, there is a lot of frustration, but I do have hope for the future that reducing burdensome workflows and addressing burdensome processes like prior authorization is going to help bring clinicians back to the joy of medicine and make it better for patients as well.
Brian: It’s certainly been a challenging time, but there is a lot to be excited about. Ryan, I really appreciate you taking the time to come speak with me today.
Ryan: Thank you. It's been a pleasure.
This podcast originally ran in Becker’s Hospital Review on March 31, 2022.
The insights you need to stay ahead in healthcare: Subscribe to Premier’s Power Rankings newsletter and get our experts’ original content delivered to your inbox once a month.