David K. Vawdrey, PhD, who recently joined Danville, Pa.-based Geisinger Health System as its chief data informatics officer, weighs in on his new role, biggest challenges in healthcare data analytics and why culture is key for innovation in healthcare.
Question: What are your key responsibilities as chief data informatics officer?
Dr. David Vawdrey: I'm responsible for implementing transformational technologies and leveraging Geisinger's advanced data and informatics infrastructure to create value for patients, clinicians, researchers and members across our enterprise that includes 13 hospital campuses, over 250 clinics, a health plan that includes nearly 600,000 members, and the Geisinger Commonwealth School of Medicine. My role is housed within Geisinger's Steele Institute for Health Innovation, which was established to elevate and share our efforts to transform healthcare delivery by improving patient experience, quality, efficiency and outcomes.
Q: What excites you most in your new role?
DV: To me, it's the confluence of three things that make Geisinger such an amazing place for informatics: first, a remarkably dedicated and talented team with support at the most senior level of the organization; second, a culture of innovation spanning more than a century; and third, a comprehensive and integrated data infrastructure containing clinical, financial, genomic and other information.
It's not by accident that Geisinger has been repeatedly recognized by policymakers and other experts as a blueprint for healthcare transformation. But the thing that excites me most about working at Geisinger is standing shoulder-to-shoulder with 30,000-plus people who are deeply committed to improving the health and well-being in the communities we serve. As we say at Geisinger, everything we do is about caring for our patients, our members, our students, our Geisinger family and our communities. And this is a model for the country and for the world.
Q: What are some key initiatives you plan to roll out as chief data informatics officer?
DV: We are focused on making life better for patients and for care providers. Patients and family members have a terribly difficult time navigating the healthcare system. There is so much anxiety, frustration and dissatisfaction, and these sentiments are felt not only by patients, but also by our frontline clinical care providers. In many ways the EHR and other health IT have made these problems worse, not better. It's not always the fault of the technology or the people creating it, but we need to collectively evolve from 'use' to 'usefulness.'
So, we are working on key initiatives to improve the way clinicians use technology to do their jobs more effectively — getting their heads out of the computer by providing a clinical digital assistant that uses context-aware voice recognition and harmonizes EHR, genomic, claims and social determinants data to provide evidence-based, patient-tailored care recommendations. We are working to improve care across the continuum — especially in patients' homes, by way of remote monitoring and technology-supported care coordination. Our measure of success is not whether we deploy advanced technologies, but whether we add value and reduce waste — in the Lean sense — for our patients, our clinical care providers and our community.
Q: What are some of the biggest challenges to leveraging data analytics in healthcare?
DV: We have some serious challenges in healthcare, and in spite of all the hype, data analytics hasn't yet solved much of anything. The truth is, the data analytics industry spends a lot of time peddling 'solutions looking for problems.' Instead of listening to sales pitches, we need to do the harder work of listening to frontline clinicians, patients, researchers, payers, etc. We need to work together with them to identify the most pressing problems that need solving, and then deeply understand those problems.
A somewhat facetious example is a model that predicts mortality with nearly 100 percent positive-predictive value. It sounds great, until you find out that the model's most predictive input feature is the presence/absence of a 'death note,' and thus, it's of no clinical use at all. This is an exaggerated example, but we have a lot of similar things going on right now in our field. So, the big challenges I see are: one, to figure out how to measure the right things; two, to figure out how to measure things in the right way; and three to make information actionable — to provide the right information, in the right format, to the right people, at the right time, so they can make better decisions.
Q: What is your advice for healthcare organizations looking to enhance their data informatics strategy?
DV: A data informatics strategy shouldn't be based on technology buzzwords. There is no silver-bullet solution: transformation requires patience and persistence. Most of all, it requires a culture of trust, teamwork and transparency. It's almost always a people problem, not a technology problem. The hard problems in healthcare require a willingness on the part of doctors, nurses, administrators and consumers to climb the path of continuous improvement together. Work to align incentives that make it easy for people to do the right thing. Measure progress and hold leaders and teams accountable.