‘Big data’ and the Learning Healthcare System: Implications for leadership development
|Last week, the Health Organization Research Association held their annual meeting at Rush University Medical Center in Chicago, IL. Friday’s program was dedicated to an exploration of healthcare organizational excellence using publicly available datasets.The morning began with the National Center for Healthcare Leadership introducing their collaborative Organizational Excellence (‘OrgX’) initiative, whose goal is to reduce entry barriers for academics and practitioners to collaboratively access healthcare organization datasets and expertise. Next on the program|
was a presentation about the information-rich but very tricky to navigate Healthcare Cost Report Information System, provided by Jeff Helton. The next presentation, from Francis Fullam, provided a practitioner’s perspective on using data from the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey to drive process improvements. The morning sessions concluded with a highly engaging presentation from Tim Huerta about the data we have vs. the data we need for driving performance improvement, and how we might best collectively pursue the latter.Collectively, the speakers provided an intriguing perspective on the emerging competencies needed to support learning healthcare systems. An interesting summary of these competencies is provided in the Venn Diagram from data scientist Drew Conway.
Tim’s talk spoke most directly to the emerging competency needs referenced in Drew’s diagram. In his view, the ‘hacking skills’ most needed will involve expertise in data conduits and cleaning (a/k/a ‘data munging’), to broaden access to the raw material analytics will require. For ‘math and statistics knowledge,’ we will need professionals who understand the roles and trade-offs associated with different descriptive and inferential statistics. ‘Substantive expertise,’ in turn, will not only require contextual knowledge of the healthcare organizations the data represent, but also advanced competencies in ‘storytelling’ – translating statistics into practical wisdom and action, through compelling narrative and visualization.
What does all of this suggest for the learning healthcare system? Although there are skill development implications for each domain, as a package they imply the need for an interdisciplinary approach. For example, according to John Holland’s well-traveled career choice theory we would most likely find statisticians among the Investigative types, the data mungers among the Realistic/Conventional types, and storytellers among the Artistic/Social types. In short, data science in the learning healthcare system seems likely to become as much a team sport as the rest of healthcare leadership is becoming. If so, its professionals will ultimately require many of the same leadership competencies the rest of our healthcare leaders need.