By Andrew Trees, Agathos Co-Founder & CEO
Photo by BMJ Blog
As U.S. healthcare payments become increasingly value-based, quality improvement initiatives will enable hospitals to improve patient outcomes, realize operational efficiencies, and better manage the financial risk associated with emerging payment models.
Quality improvement (QI) is the systematic, formal approach to analyzing and improving practice performance. There are already numerous proven best practices for managing QI toward given metrics. However, objective definition and measurement of healthcare quality—particularly as it pertains to individual providers—is a very challenging task. While quality metrics theoretically give physicians useful benchmarks, they too often foster distrust or apathy when not carefully developed, implemented, and measured.
Hospitalists (hospital-based internists) are increasingly expected to pursue QI but often lack the data, tools, and analytics support to do so responsibly and sustainably.
Challenges with measuring physician quality are well-documented. As a start, data integrity, technology limitations, and statistical challenges (e.g., small sample size) present formidable barriers to entry. But even assuming one’s data is clean and well-powered, at least five clinical barriers stand in the way of objective quality measurement for physicians:
1. Attribution: when many agents (e.g., physicians, allied care providers, ambient factors) influence a patient’s outcome, how can we attribute ownership of that outcome to an individual physician?
2. Complexity: when some patients are sicker or otherwise more medically complex than others, how can we normalize metrics and reward physicians for tackling more difficult cases?
3. Regionality: when patient demographics, access, and flow dynamics vary across regions and individual medical centers, how can we normalize metrics responsibly?
4. Seasonality: when select seasons (e.g., winter) present more illness, or when select times of the day (e.g., night), week (e.g., weekends), or year (e.g., holidays) present unique census or staffing dynamics, how can we objectively benchmark physicians?
5. Tradeoffs: when physicians must determine whether to help a patient feel better quickly, prevent a future illness, initiate a difficult conversation about palliative care, manage hospital costs, or simply get to the next patient on time, how we can balance these competing goals within quality metrics?