Iowa Health rechristened itself. In so doing, the health system took the name UnityPoint Health to signify its aim of putting patients at the point of care unification.
And in becoming an ACO and what CMIO John Frownfelter, MD, described as “an organized system to deliver population health management and coordinated care leveraging ACO contracts,” UnityPoint finds itself straddling two worlds: ACO and fee-for-service patients.
Government Health IT Editor Tom Sullivan spoke with Frownfelter about the ethical dilemma posed by having the ability to offer richer services to patients under ACO contract than those who are not, how UnityPoint is looking to analytics as the keys to resolving that dilemma, and its future plans for big data.
Q: A study published last week found that the ACO model used for Massachusetts Blue Cross Blue Shield patients actually lowered costs for non-ACO patients treated by the same providers. Are you seeing something similar?
A: Absolutely, and that’s the risk we incur. We’re investing in programs that will reduce cost, but we’re making the investment. So we spend, let’s say, an additional million dollars to set up a patient-centered medical home and the savings are reduced total cost, per member per month is the standard measure, if we use that service just for those patients under the ACO contract, we get the shared savings. If we apply that service to everybody, then for those in the fee-for-service world, the payers get the savings.
But how can we deny a quality service like that to patients just because they’re in the wrong sort of contract?
That’s been a moral struggle for us. We’re straddling two worlds with a lot of discomfort right now between giving those quality services to all patients we serve and yet staying financially solvent by ensuring that those services requiring significant investment bring a shared savings return to us. See that tension that we’re under?
Q: How do you handle balancing the ethical dilemma with finanical realities?
A: We’ve said for the most part, unless there are significant reasons to push patients and the insurance companies into more of a fee-for-value model, we’re going to provide these services to all patients. So we have a 1-800 nurse call center, that’s for all our patients. We have patient-centered medical home, that’s for everybody. But when we’re investing intently in something that we’re not yet sure of the return, we’re going to prove that in a fee-for-value shared savings model before we roll it out and bury ourselves financially. That’s just being prudent. The things that we know work and are cost-effective, we’ll give those to all our patients.
Q: So how do you envision reconciling this scenario you described as straddling two worlds?
A: Both our health system and the country are on a trajectory toward a fee-for-value model. We project that within the next couple years more than 50 percent of our patients will be in those types of contracts, in fact, fully going toward a capitated model for most of our patients. So we need to stay ahead of that, be prepared because our care delivery must be smart. We’ll incur some losses along the way but that cannot cause us to shrink back or to hesitate in investing to set these programs up, in proving their value, and taking care of our patients the right way. And improving their value comes down to big data analytics.
Q: Which brings us to another of today’s great trends: What is UnityPoint doing with big data?
A: We’re collecting it today. I don't want to overstate it and say we’re doing big data analytics as it’s cast in health IT literature today. But we are collecting information as a foundational starting point. And we have tools in place through Explorys, our vendor partner, to begin leveraging it but big data is complex. It’s not simply a matter of compiling the data and then asking intelligent questions against it to get great answers. I think we all know that.
Q: What’s on the big data horizon, say, for the next 5 years?
A: First, we’re solving some of the hurdles. We have internal and external hurdles. External hurdles are the increasingly strict requirements around not only data sharing or data access, behavioral medicine, protected health information particularly, the laws vary by state and so how we have to manage that data gets more and more complex. Internally, we have to reconcile things like attribution, so when we say this provider delivered that care, w have to know it was really that provider. Everyone dealing with analytics today has major questions around these core issues, like attribution, and another one is the definition of an encounter where care was delivered. Those can be addressed before a solution is in place, in fact, it must be done before the data is used effectively. So we're working on those today.
Our road map is to increase both the data inflow from various sources and the sophistication of the tools that we use to search and manage that data. That goes to things like Natural Language Processing; we don't have that in place yet and I’d say the majority of health systems today don’t but it’s trending upward. From a practical standpoint, we need those things realized as soon as possible, so we can ask the sophisticated questions that correlate data and give us answers to trends, patterns, or predictions in a way that we could not otherwise do so.
The simplest analogy is weather forecasting. Right now, all we can tell you is the weather that happened yesterday and maybe what the temperature and conditions are right now. We’d sure like to tell you what’s going to happen tomorrow and we’d also like to tell you how to prepare for it and — here’s where it breaks from the weather analogy — ultimately to modify it through interventions.
Related articles:
