Technology, analytics boost population health management at Geisinger [Q&A]
Population health management efforts increasingly rely on information technology and data analytics tools to achieve high-quality outcomes. Case in point, Danville, Pennsylvania-based Geisinger Health System, which uses a combination of patient- and provider-facing tools to ensure that consumers and doctors are on the same page, and that the latter can identify larger trends, both positive and negative.
In a recent interview with FierceHealthIT, Eric Newman, M.D., vice chairman for clinical innovations in Geisinger's division of medicine, and Chanin Wendling, director of Geisinger in Motion, discuss how each entity takes advantage of technology and innovation to close care gaps and improve results.
FierceHealthIT: What are the biggest technology and/or analytics challenges you face in improving population health management?
Newman: One is the point of, are we actually capturing the right things? I think we're getting smarter and better at that. The focus in the past has been on traditional medical elements, some of which are read in the electronic health record; there may be some questions as to how accurate that information is.
There's also expansion to new data streams with patient-reported outcomes, etc.; perhaps some additional domains of data to pull areas that we've not focused on with social determinants and other things that really help us to profile our patients in a politically acceptable way so we can understand truly about our patients and create more of a "many-sizes-fits-many" rather than a one-size-fits-all analytic model.
Challenges associated with that are cleaning up the data; choosing the right parameters to understand what other things we should be capturing; and how to successfully integrate that information on the front end. We've got these large, robust data sets, but realize that people have to use them; to that end, we've developed tools to provide information at the front end in much more interesting and meaningful ways than the typical EHR approach.
I think the final challenge is that Geisinger simply is one cog on a very large wheel. In the past, we've thought very centric on our own existing IT and our particular software that we use, but it's a much bigger world out there and all of us are challenged to integrate better with others and think more broadly about knitting the pieces of care together so that it's not just what happens in the hospital, it's the care that the patient receives before, during and after. We are not a closed system; we're really not Kaiser [Permanente], so we have to be able to effectively knit those pieces of patient care together in effective ways, and we can't do that unless information streams across multiple platforms in an interoperable fashion.
FHIT: Can you provide an example of how you're better implementing information on the front end for providers?
Newman: One that's sort of near and dear to my heart is a program that we developed internally that's now being commercialized. The original name of the program was PACER (Patient-Centric Electronic Redesign). It really helps to integrate information from four data streams: from the patient using a touchscreen questionnaire; from the electronic health record; from the nursing team; and from the provider physician team. It re-aggregates that information into new functionalities and views it in an environment that's much less constraining than the typical EHR view. Essentially, the information is viewable in a way that it's the right person, right information, right time. So the view that the nurse would see is completely different than the view that the provider would see. Those views can help pull information together in new and impactful ways that are either impossible or very difficult to demonstrate in the EHR.
That single one functionality--the ability to objectively see how our patients are doing--has been huge for us in terms of being able to better manage our populations and to better engage our patients, because the same view that the doc is seeing we can show to our patients in very meaningful ways.
And patients have some kind of a strange historical memory. They don't often remember correctly how they were actually doing at a certain time, so the ability to actually show that to them that, "Hey Jeannie, this drug is actually working; let's continue it or maybe even de-escalate it," or "Oh my goodness, Johnny, you're really not doing all that well, see here, this is how you were doing before and the treatments that we're giving you are not really improving your care, you're getting worse, time to change therapy;" that's huge. We use that as both a way to objectify and understand on a population level how our patients are doing, but also down to a micro level, down to a one-on-one.
The really cool part about all of this is we also can create patient-level scorecards about how patients are doing, we can attribute them to single physicians, and we can use that at point of service for task management. So we can understand where the gaps are in care and apply them at point of service or even between visits.
We have a specific example of that called AIM FARTHER (Attribution, Integration, Measurement, Finances, And Reporting of THERapies), where we co-managed all 2,400 rheumatoid arthritis patients across our healthcare system and we were able to show improvement in quality across a broad array of quality measures that are higher than almost any other institution has ever been able to report. We saved $1 million in biological costs over a single year in 2013, so it works.
It's an example of our ability to take information and use it in impactful ways. Getting the information is just the start; how you use it is really where the power is.