Skip to content Skip to navigation

Using Clinical Data Repositories to Manage Population Health

February 9, 2012
| Reprints
Anceta’s CMIO John Cuddeback shares ways to use predictive models for case management

For Healthcare Informatics’ annual Top Tech Trends feature, I recently spoke with American Medical Group Association’s John Cuddeback, M.D., Ph.D., chief medical informatics officer of Anceta, AMGA’s collaborative data warehouse, to see what was in store for future development of accountable care organizations (ACOs) and what care coordination tools will be necessary to lay the groundwork. Dr. Cuddeback said that some AMGA members have been utilizing Anceta to help manage complex patients. This is similar to what other organizations like the American Academy of Family Physicians’ Center for Health IT have been doing for their members. The AAFP and Emdeon launched its clinical data repository pilot project last September to explore ways in which administrative and clinical data can be leveraged to benefit AAFP members. I think this trend of professional organizations creating their own data repositories will only continue in this new era of accountable care to help providers better care for their patients and focus on population health. Here are some excerpts of my conversation with Dr. Cuddeback:  

How many organizations are using Anceta right now?

We have 20 participants now, and we’re right in the growth phase now. We’re working with Humedica [Boston, Mass.], and they do the actual normalization and mapping of the data to be able to do meaningful comparisons and they deliver an interactive query tool that is designed for people to use who aren’t programmers. One of the big issues is with performance measures, is the question of measures for accountability versus measures for improvement. Measures for accountability, like HEDIS measures, have all of the complex patients sort of negotiated out of the denominator. But it’s the patients that go beyond the guidelines, patients where conflicting guidelines may apply, are where the greatest opportunity exists. So one of the issues is being able to look at those exceptional people within your population and that’s really what our analyses are about is to understand the entire scope of the population, and seeing where the opportunities are.

As we look across the groups we see very little variation in the outcomes they are achieving, what we see is the variation in the processes they use and the cost of achieving those outcomes. Some groups are using more expensive drugs than others; some people are using more visits than others.

In what ways are AMGA members using Anceta?

Some medical groups are managing to achieve improvements in blood pressure control in diabetes patients with fewer visits than others. It’s identifying those [improvements], and we tee those up for discussion in the collaborative. We found the opportunity to do patient outreach, and as a matter of fact, our partner Humedica has just introduced predictive analytics for patients with congestive heart failure. So now we have a few medical groups looking at their congestive heart failure patients, using our data as part of the case management activities that they’re doing.

One of the things that has evolved as a discussion among AMGA members is the difference between case management vs. care management because the two sound very similar. But they’re really quite different. Care management, or care coordination, is the idea of designing the mainstream process of care, so the idea of planned visits and stuff like that. That’s what you put care coordinators in place and you design processes to go smoothly for the patient.

The case management is the individualized attention for the really complex patient. It’s the recognition for those people, the standard processes are probably not sufficient. You’re probably going to have issues to deal with that require a bit more creativity. These patients are at the intersection of different guidelines and care imperatives, and there are often some trade-offs that need to be made.  It’s the distinction of making the system as efficient as possible for the bulk of patients, and then providing individualized attention for the people who are going to be outliers. The trick is figuring out who is who.