How Do You Define and Find who Needs Care?
After delivering an opening keynote at the Care Innovations Summit, Dr. Atul Gawande returned later to moderate a panel on delivering care to the chronically ill. In a conversation with Gawande, CEO of WellPoint/CareMore Alan Hoops told him that CareMore, a health plan and heal care provider, is able to deliver more tailored care than a senior’s regular primary care provider. Hoops said that primary care doctors are relieved to know that these patients are being cared for in an intensive system.
This prompted Gawande to respond, “You need to be selective about who you put that intensive care with. We have intensive care units in hospitals, and part of making them really valuable is ensuring that it is for the people who need it the most.”
Then he asked of the panel: “How do you find and define who’s appropriate?”
CEO of WellPoint/CareMore Alan Hoops:
“For CareMore we define it however we possibly can. I indicated that one of the critical success factors is predictive modeling. And frankly, we throw everything but the kitchen sink at predictive modeling. The best predictive modeling event is the acute care episode for the senior.
Statistically 50% of all seniors who are admitted die or are readmitted within a year. About 20 to 25% of seniors are dead within a year. So this is a big-time predictor for the seniors.”
“But we use everything we use some statistical modeling, we use a program … we come up with home made things. We, for a while, looked at any female 90 years old, living alone. That was a predictive model. So we use any way we possibly can to give us as much runway as possible to determine when that patient is going to hit the point of inflection on that slope and go from being a relatively healthy person to being a very needy and complex person.”
Professor in Gerontology at University of Pennsylvania School of Nursing Mary Naylor:
“Ours is an evidence-based model. Risk screen has been grounded in multiple NIH clinical trials where we’ve tracked for over a year, the experience of people and their outcomes and been able to predict, not predict, but to identify those individuals who are at a point of risk that if we don’t intervene will be back into our costly health care system.
So it is based on both randomized control trials as well as testing the transitional care models against other evidenced-base approaches.”
“It’s a risk screen that has a number of factors so we are focused on Medicare beneficiaries. We are focused on dual eligible populations. We are focused on individuals with multiple complex conditions, individuals who in the past 30 days have some kind of acute service use, individuals who in the past six months have multiple acute service use, individuals who screen as cognitively impaired (which we often don’t screen consistently in our health care system), individuals who screen as depressed. So those are among the risk factors.”
CEO and Medical Director of Health Quality Partners Kenneth Coburn:
“This in the area of hot research right now. And on the slide I showed several risk groups and that’s because when we started the model with Medicare coordinated care demonstration, we had low, medium and high risk groups based on classification using a questionnaire with permission from the Sutter Health Organization, which we found very helpful for teeing up the needs of patients for the model with nurse. Mathematical policy research, the CMS rapid evaluation team have looked at this data lots of different ways and sliced and diced it, and the group that we’re targeting now, based on their analysis, are folks that have either heart failure, coronary artery disease, diabetes, COPD and one or more hospitalization in the preceding year. So I think the estimate, if I have this figure right was somewhere between 14 to 18% of all Medicare beneficiaries would fall in that group.”