Terri Steinberg, MD, CHIO and VP Population Health Informatics, Christiana Care Health System
For Terri Steinberg, population health isn’t just another project; it’s a passion. She has spent the last five years working to build a program that leverages analytics – and the use of embedded care managers – to manage the care of 200,000 patients. But the journey hasn’t always been easy. Christiana Care had to first create an infrastructure that would enable more sophisticated use of data, then demonstrate that it could be utilized to achieve improved outcomes. And then, of course, there’s the tweaking that needs to be done to help the program continue to thrive.
In this interview, Steinberg, who serves as Chief Health Information Officer and VP of Population Health Informatics, talks about what is has taken for Christiana Care to start the transition from fee-for-service to value, including securing a grant from CMS and having leadership that bought in to the vision. She also discusses the “evolution” required to align incentives, the pivotal role that care managers play in improving care and decreasing readmissions, and what it’s been like to build Carelink CareNow from the ground up.
Chapter 3
- Aligning incentives – “It’s been an evolution.”
- Focus on risk contracts
- “We have really changed the mix of services that members seek.”
- Active care management – “you don’t have to go looking for things, because the data come to you.”
- Embedded care managers working in partnership with providers
- Right-sized care
- Misconceptions about managed care
- “If you have needs, we’re all over you.”
Bold Statements
We requested risk relationships from our payers because we believe the services that most directly impact care — the outcomes, quality, satisfaction and everything about care — are those for which you can’t drop a bill.
We’re heavy on social services and we use our clinical services in the right way. We make use of telemedicine. We make use of virtual medicine. We have nurse visits.
The notion of care manager to member ratio is sort of an anachronism in this technology, because you actively care manage everybody in your population. You don’t have to go looking for things, because the data come to you.
Care managing is not rationing care, and it’s not depriving people of what they need. It’s giving everyone the right sized care, using technology tools to help determine that.
It was a transition from a single person taking the reins of doing the work to create an organizational infrastructure of informatics professionals that divides up the work in a way that makes sense.
Gamble: One of the key issues that comes up when we talk about population health is the need to align payment performance incentives. How has your organization worked to address this?
Steinberg: It’s been an evolution. If you look at how well providers understand many of this issues with payment, it’s still bourgeoning. It turns out that doing the right thing will get you where you need to be. Christiana Care wasn’t forced to embrace value-based payment structures; we requested risk relationships from our payers because we believe the services that most directly impact care — the outcomes, quality, satisfaction and everything about care — are those for which you can’t drop a bill.
One example is social work services. It’s pretty obvious that if somebody doesn’t have a roof over their heads, they’re not going to do as well as somebody who sleeps in a warm bed. Primary care physicians who don’t have a social worker or a nutritionist cannot do much to get to the real root of illnesses across a population. Even clinics in underserved areas, such as Federally Qualified Health Centers, are just not incented to have the right complement of services.
That was one of the most important things we did at Carelink. We said, we’re going to aim for risk contracts where dropping a bill is not what it’s about. You’re capitated. You can spend your money however you want; we’re going to spend it where we think it’s important, and that means we’re heavy on social services and we use our clinical services in the right way. We make use of telemedicine. We make use of virtual medicine. We have nurse visits. We have really changed the mix of services that members seek. A visit might be a social work visit, it might be a pharmacist visit, it might be a nurse visit — not every visit has to be a doctor visit. And people are increasingly getting what they need with better outcomes. We think that we are moving the needle on the health of the population.
Gamble: Right. And of course the numbers have to come into play.
Steinberg: Of course. That’s where it gets tricky, because the things that we’re being asked to measure by federal and private payers don’t necessarily indicate the health of the population. They may indicate the sufficiency of your screening. They may indicate many important things, but I don’t know of a measure yet that says whether your incidence or penetrance of homelessness has gone down. We look at those things and we pay a lot of attention to the social factors as well as making that part of our risk model.
In terms of the economics, we have sophisticated reporting that tracks what our PMPM (per-member, per-month) should be, what our PMPM is, and where we have room for improvement. We divide it by line of business. If it’s something like a Medicare Shared Savings population, we look at it by the reasons you qualified for Medicare, whether you’re disabled, you’re dual-eligible, you’re end-stage renal disease, or you just aged into Medicare. Obviously the PMPM is different, and the health status is different.
We endeavor to catch things early. If you look at it, the PMPM target for end-stage renal disease is something like 7-times the amount compared to the PMPM target for aging into Medicare, which is less than 10,000. Obviously we want to prevent end-stage renal disease, which can be done by making sure you’re managing people with diabetes. It doesn’t take a lot of years with diabetes to develop end-stage renal disease. You want to make sure people have come in for screening blood pressures. You don’t want somebody to have 5 or 10 years of hypertension that went undetected, because when they do, they present with end-stage renal disease.
So we monitor the finances, we monitor the clinical values, and we monitor the outcomes. Every time somebody has got a hemoglobin A1c, it comes into the care managers. If it’s above a certain level, they’re tasked to take a look at those numbers. And so the notion of care manager to member ratio is sort of an anachronism in this technology, because you actively care manage everybody in your population. You don’t have to go looking for things, because the data come to you.
Gamble: Right. That speaks to how having that having an embedded care management model can make a difference.
Steinberg: Absolutely. Another thing is that, in some cases, care managers are empowered to execute a protocol. Let’s say a patient had ischemic heart disease. If he had bypass surgery, he’d go home from the hospital and then 10 days later, he’d end up coming back. Why is that happening? All the fluid he was given during surgery was coming back into his system, and he was going into heart failure. So the nurses work with the cardiothoracic surgeons and cardiologists, and they all agree that within a certain framework they’re going to monitor a patient’s weight, and if they see that the weight is above a certain amount, they’re going to send a visiting nurse to the house for an examination, and they’re going to execute a protocol.
That is a highly effective component of how we manage this particular population. These were people who otherwise would have shown up in the ED, and very likely would’ve been admitted because they had suffered a heart attack 10 days before. So that made a big impact, having the ability to change how we manage a predictable complication of a procedure. And it was highly effective.
Gamble: And this is different from the care managers that are part of the payer model?
Steinberg: Yes. In our embedded care management model, we work closely with providers, and so it creates a very, very different relationship than the care management model for payers. On the payer side, care managers are tasked with calling people after surgery to say, ‘how’s everything going? What do you need?’ That’s not what Carelink CareNow does. Carelink CareNow has an opportunity to take it further. For example, ‘You’re having surgery in three weeks, and I see that you smoke. Let’s get you on a smoking cessation protocol right away.’ Or maybe, ‘Let’s put your surgery off for another six weeks. If you can stop smoking for a month before surgery, we can improved your outcomes.’ That’s a whole different animal, because the care managers in Carelink are working in an embedded model with the providers in partnership. They extend the reach of the providers. And when they call, they’re not saying, ‘Hello, this is Suzy from Carelink CareNow.’ They’re saying, ‘This is Suzy, the care manager from Dr. Smith’s practice,’ and it’s a whole different relationship. The doctors are highly satisfied and the patients feel taken care of.
Our care managers follow patients, whether they’re going to a hospital, nursing home, or any care facility. It’s not always the same care manager, but it’s the same care management record. It’s the same group of people. They talk to one another and they report to the same senior vice president, and so there is continuity.
The net result is that not only has patient satisfaction been high, but readmissions from nursing homes have decreased dramatically, because we help manage the members in the nursing home. We provide another set of hands. We provide information. We’ve shortened length of stay across the board in nursing homes because there’s a safety network for the patient to go home earlier instead of the other alternatives. Very often one of our members will show up in the ED and the decision will be made not to admit because Carelink is notified as soon as the patient registers, so very often we will call the ED and say, ‘One of our members has just registered, here’s some history. If you feel comfortable sending the patient home, we can have a visit later in the afternoon or tomorrow morning.’ It’s a safety net. As someone who’s been on the provider side for many years, I can attest that sometimes you admit a patient just because you’re worried or because you want to observe or something isn’t quite right, and you feel differently if you know that there’s an observation team available to you. The embedded care model really works quite well.
Gamble: When you spoke at the New Jersey HIMSS conference this past fall, you talked about that ability to right-size care, and it seems like that’s such a big component of population health.
Steinberg: That is very true. And I want to make this important point that I wish patients would embrace. Managing care, at least in my observation, results in more care. If you have needs, we are all over you. We are so in your face you can’t even believe it. You get much more care than you ever could’ve have imagined. The flipside of that is some people actually need less, and they get what they need as well. Care managing is not rationing care, and it’s not depriving people of what they need. It’s giving everyone the right sized care, using technology tools to help determine that. We don’t hear, ‘you’re not taking care of us.’ We never hear that from people. We hear, ‘thank you so much for being all over us.’
Gamble: It’s such an interesting shift and I’m sure it’s been interesting from your perspective to watch this unfold. The last thing I wanted to talk about was your role as Chief Health Information Officer, which you were named to in 2015. Was that a reflection on the organization’s shift to focus more on population health and integrated care?
Steinberg: That was part of it. It was in recognition of the fact that population health technology was now part of our suite of applications. It was also in recognition of the fact that we did need a CMIO, assistant CMIOs, and a CNIO, and that there needed be a right-sized umbrella. It was essentially creating an oversight and management responsibility of the informatics professional in a way that hadn’t occurred before. It was a transition from a single person taking the reins of doing the work to create an organizational infrastructure of informatics professionals that divides up the work in a way that makes sense, because it’s too much work for one person. So it really was a growth of the scope of the informatics organization.
Gamble: Very interesting. I want to thank you so much for taking the time to speak with us about the work your organization is doing, and what you believe it takes to build a successful population health model.
Steinberg: Thank you, Kate.
Share Your Thoughts
You must be logged in to post a comment.