VOICES OF REFORM: Geisinger's "Experiment of Scalability"

Dr. Glenn Steele, the president and CEO of Geisinger Health System, recently co-authored an article in Health Affairs and appeared at a Health Affairs forum here in DC to talk about the innovations at Geisinger, and the ways that Geisinger's unique characteristics might limit their adaptability elsewhere. Geisinger is both a health plan and a health network spanning 41 counties in Pennsylvania, and it has an unusual mix of Geisinger and non-Geisinger physicians, of open-and closed-staff facilities. It is blazing trails on many fronts, in chronic and acute care, from medical homes to its ProvenCare initiative for certain acute inpatient conditions and procedures.

We spoke with Dr. Steele recently and turned his question around. We weren't so much interested in the limits of Geisinger's example; we wanted to know what would work elsewhere—which of those trails could be followed or adapted by others. To our relief, Dr. Steele told us that other hospitals and policy experts have been flocking to Pennsylvania recently trying to answer that precise question. And the answer is that quite a bit of it would work elsewhere.

"An experiment of scalability and generalizing is underway," Dr. Steele said. "There are a number of hospital-based provider organizations that are intrigued with some aspects of our ProvenCare."

"We are just out to prove that some of our innovations help human beings," he added. "People come to us and learn."

Geisinger's experiments are coming at a time when the models for relationships between doctors and hospitals are changing. A number of hospitals are expanding their employed doctor groups, as one model, or developing more integrated delivery systems, as another model. Geisinger, in a sense a hybrid, has lessons applicable to both, in both chronic and acute care.

One target for Dr. Steele has been the unjustified variation in care. Not between one region of the country and another, or even between different hospitals. He was distressed by the wide variation within a hospital from one doctor to another. The ProvenCare model seeks a sensible level of scientifically-sound consistency on certain common hospital-based procedures. "We ratcheted down unjustified variation," Dr. Steele said, creating what are basically care templates for procedures like coronary bypasses. "We come to a consensus on best practices," he said. The ProvenCare model blends aspects of best practice, ongoing quality improvement, and innovative payment models that reward quality, and incentivize good risk management -- or better yet, sensible risk reduction.

The ProvenCare consensus also has a bit of built-in fluidity. For instance, the coronary artery bypass graft care model at Geisinger has 41 best practices "baked in." These aren't all defined by randomized clinical trials, because those studies are not always available, and best practices aren't always all agreed on. But at Geisinger the experts did come up with a consensus. And the model can adapt, even encourage, changes in consensus as more clinical evidence is amassed. Since the template has those 41 practices "baked in," they can remove an outdated one, and easily slot in the new one.

Another example he gave was managing type-2 diabetes. There's a debate about how low A1c levels should go, particularly in elderly diabetics. "We are not necessarily committed to a rigid set of best practices. We are engineering a system where, when there's new data, new consensus, we can slot something out, slot something else in, and monitor the effects," he said.

Health IT has been integral to Geisinger since it introduced electronic medical records back in 1995. Both Geisinger and nonGeisinger providers can access the records. Some patients can also access it to see their lab results, communicate with providers, make appointments, request prescription refills, and get educational materials about their health. But IT alone brings neither savings nor quality improvement. IT has to be integral to the delivery and coordination of care.

"A lot of folks have been tinkering with, or have purchased, IT as a predicate. A lot of them are intrigued by how we've used the IT not to cut down on pieces of paper that are generated, not to get rid of full-time equivalent (staff) doing transcriptions etc, but how the information technology is a necessary precondition to a lot of real time data that's used in our re-engineering of care," Dr. Steele said. That clinical data is necessary both for the ProvenCare programs as well as helping medical home care teams stay on top of managing chronic diseases and keep patients from unnecessary hospitalizations. The medical home includes a "Personal Health Navigator" that incorporates 24/7 access to primary and specialty care, nurse care coordinators, home-monitoring and predictive analytics to identify risk trends. (Outcome data is preliminary and limited, but so far encouraging. The Health Affairs article said that that early data from the first year at two pilot sites show 20 percent reduction in hospital admissions and seven percent reduction in total medical costs). Health IT is important throughout the Geisinger system, not just the new medical home pilots.

Dr. Steele isn't all that focused on reform from Washington in the next administration. If he has a wish list for the next president, he isn't sharing it. "What can Washington do? Who the heck knows?" He's more focused on showing dramatic benefits in patient outcomes, and helping other health systems figure out how to follow Geisinger's path. A lot of the care Geisinger provides requires what Dr. Steele calls "blocking and tackling," i.e. planning and managing and doing things that make sense, not just by ordering more expensive tests and technology. "If other systems, other markets continue to come to us to try to replicate it.. if some insurers would get involved, that would be a breakthrough."

"A lot of health policy people know a lot more economics than I ever will. Yet they seem to crave real life experiments like ours. They are really interested in this. To translate a little of what we are doing into new demonstration projects, new waivers." Not that he wouldn't like some funding streams dedicated to promoting more innovation. "We'd like it if Medicare didn't have to do one-size fits all, if it could allow a cohort of innovative systems to do some cool stuff."