Technology is only part of the solution at Geisinger
CHICAGO—Geisinger Health System is working on data analytics, innovation and incremental behavior change to improve patient outcomes, said Alistair Erskine, MD, the health system's chief clinical informatics officer, speaking at the 2014 Healthcare Leadership Forum. “That’s what we’re convinced will make the changes in the end.”
Geisinger has invested heavily in technology with a thriving telemedicine practice; an active, financially stable health information exchange; a move away from pagers to smartphones; and heavy use of its patient portal. “We are very much engaged with at least 35 percent of our patient cachement area through our patient portal.”
This progress has its cost, however. “All these decisions to digitize come with unintended consequences that need to be addressed.” Erskine said technology is just part of the solution to improve healthcare.
Geisinger is half health plan and half clinical enterprise, he said. Despite that, half of its patients don’t have a Geisinger health plan. That results “in a petri dish of different ways of having risk-based and fee-for-service patients so we can try different models.”
The organization processes its data through both a traditional data warehouse as well as access to real-time information from its database and some cloud-based natural language processing. That allows for very quick processing of very large amounts of data.
That information is used to do some predictive analytics, he said. In fact, Geisinger is betting that genomic and genetic testing will let them start patients from birth with a full set of genomic sequencing and throughout patient lives, develop illness correlations.
Geisinger also has adopted OpenNotes—a system in which patient receive all notes clinicians write about them. The patient gets the note as it’s written, not a summarized version. “It’s a way of making sure that the two are connected and that what I’m writing is something I know my patients will read so I will write a note more intelligible to the patient. It has been wildly successful. Patients love it and the providers who were initially terrified are finding a lot of benefit from it as well.”
Erskine also suggested that healthcare reconsiders the way medicine is taught, which is still primarily through textbooks and lectures. Why not expose medical students earlier to the EMR so they become more familiar?
Instead of interoperability, Erskine said “interappability” might be the way forward. The market now offers several monolithic EHR systems but he is concerned what could happen if the healthcare industry goes the way of several other industries. Books and records have gone digital, travel has gone online and big box stores have gone to Amazon. While he conceded that some EHR vendors are working toward interoperability but “if we continue down that path, we may get in trouble.” Those systems don’t really allow for much innovation at the edges. “They may be part of the solution but they are not the solution by themselves.”
One way to adapt, he said, is with apps. Several vendors now have app stores and adding these layers “is just a different way the industry is negotiating these costs.”
Echoing several other forum speakers, Erskine said “all this intelligence with data and technology doesn’t manifest itself with the clinical workflow to get good outcomes.” It’s really important to reengineer and redesign the actual workflow, he said. “Just working faster doesn’t necessarily help the patients.” Geisinger’s rigorous design practice strives to remove embedded paper workflows.
Outcomes is a recurrent theme, he said. He cited the example of patients with diabetes. There are nine measures that everyone agrees are important to this patient population. The workflow needs to go outside office encounters because that time is stressed and rushed and the information shared won’t always be retained. The work can be distributed across the continuum by prompting clinicians with reminders to help make it a more reliable process, Erskine said.
Geisinger looked at measures for 27,000 patients with diabetes. At first, just 2.4 percent of them were getting all the treatment they were supposed to. “That was a sobering number.” They got that up to 14 percent—quadruple from initiation. They have seen the same pattern for other conditions which, he noted, is sustainable over time.
“It’s not the tool; it’s how we implement. We can pay people more but it won’t change everything; it won’t change behavior.”
All the available technology can make a difference to patients when providers engage patients and transform their culture, he added. Organizations also should be sure to be patient-focused and use data mining to support their care.
The improvement in metrics for that population of diabetes patients prevented 305 heart attacks, 140 strokes and 166 cases of diabetes retinopathy, Erskine said. “We’re on a journey to take technology, data analytics and the concept of innovating along the edges to better processes to change outcomes and make our patients better.”