Unwarranted Variation: A Preventable National Crisis

If you knew visiting a bottom decile hospital more than doubled your chance of dying compared to a top decile hospital¹, would that change where you sought care? If your mother’s risk of getting a central venous catheter blood stream infection was nearly 20 times higher at a bottom decile hospital¹, would you try to steer her to another facility? Would you be surprised to learn that your uncle in Idaho Falls is 3 times more likely to undergo knee replacement surgery than your friend in the Bronx², or that your cousin in Louisiana is more than twice as likely to be on a high-risk medication than your co-worker in Minnesota²?  

These large differences in care, which directly impact us and our loved ones, are a reality of health care in the United States today. Differences in medical practices that cannot be explained by the severity of illness or patient care preferences are what we call unwarranted variation³. This unwarranted variation is routinely observed between states, zip codes, and even between providers in the same facility. We believe over half of this unwarranted variation is avoidable, and current methods to address variation have made little progress. Common attempts to minimize unwarranted variation include purchasing expensive business intelligence technology, creating dashboards, forming committees, designing physician alignment structures, and changing incentive models. The missing piece in most of these approaches is true clinical collaboration with physicians. Without physicians actively engaged in the process, it is difficult or impossible to make lasting change.

The good news is that reducing unwarranted variation is good for hospitals and medical groups in both a fee-for-service and a value-based payment world. Doing this well reduces hospital cost structures in a fee-for-service environment and provides flexibility within a value-based payment environment to deliver returns in bundle payment arrangements, price competitively in narrow networks, and/or attract price sensitive, high-deductible consumers. Improving unwarranted variation will be a critical success factor for the nearly 1,300 entities, including 832 hospitals, who applied for the new Bundled Payments for Care Improvement Advanced model.

As an example of the opportunity, our friends at Ancore Health have a new and insightful way of breaking down opportunities to reduce unwarranted variation within medical groups and hospitals across the country. Ancore Health used Medicare claims data to generate variation scores specific to disease states, where a higher variation score indicates greater opportunity for care standardization. When examining variation in hospitals across five high volume DRGs, the Pacific and Middle Atlantic regions show the highest variation in costs, while the East and West South-Central show the lowest. DRG 871 (sepsis) shows the highest variation in cost, followed closely by DRG 64 (stroke) and DRG 189 (pulmonary edema). Surprisingly, perhaps, DRG 470 (joint replacement) shows some of the lowest variation among these DRGs.


It is not surprising that sepsis topped Ancore Health’s variation opportunity list. Sepsis is one of the most common (750,000 annual hospitalizations), costly ($24 billion in hospital costs), and deadly (215,000 deaths annually) conditions in the US. Every hospital’s sepsis performance is becoming increasingly transparent and visible to the public. In July, CMS began reporting individual hospital adherence rates for the 3-hour sepsis bundle, which has been shown to reduce sepsis deaths. Performance data for hospitals in the US range from 100% to 1%, and each hospital’s performance can be found on the Hospital Compare website. The average overall rate across the country was just 49% through the first three quarters of 2017, suggesting significant nationwide opportunity for improvement.

Predictably, reducing unwarranted variation associated with sepsis is a major focal point for health system leaders. However, working with physicians to reduce variation can be hard work because no two patients are alike. Differences in patient presentation, co-morbidities, and disease course make it difficult to drill down into the root causes of this variation and provide meaningful feedback to physicians. Engaging providers with patient simulations offers an opportunity to remove confounding patient variation and focus instead on the differences in clinical decision making that drive unwarranted variation across groups of providers.


QURE and Ancore Health’s unique point of view, based on experience, shows that when all providers care for the same (virtual) patients, get feedback on their individual performance, and come together to discuss their care decisions, clinical practice improves.


Physicians, who influence 80% of health care spending⁴, want to know how their care compares to their peers and how to improve care for their patients.  Unwarranted clinical variation is common and pervasive, but it does not have to be. Engaging providers with patient simulations and feedback represents an enormous opportunity for health systems to improve quality and financial performance, regardless of which side of the value-based care coin they operate on.

To see more examples of how this data-driven, collaborative approach has reduced variation across different disease areas, visit us qurehealthcare.com and ancorehealth.com or reach out to us for more details and discussion.


John Peabody, MD, PhD
President, QURE Healthcare


Eric Passon
CEO, Ancore Health


¹Rosenberg BL, Kellar JA, Labno A, et al. Quantifying Geographic Variation in Health Care Outcomes in the United States before and after Risk-Adjustment. Plos One. 2016;11(12). doi:10.1371/journal.pone.0166762

²The data was obtained from The Dartmouth Atlas, which is funded by the Robert Wood Johnson Foundation and the Dartmouth Clinical and Translational Science Institute, under award number UL1TR001086 from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH).

³Wennberg JE. Unwarranted variations in healthcare delivery: implications for academic medical centres. BMJ. 2002;325(7370):961-964. doi:10.1136/bmj.325.7370.961

⁴Crosson FJ. Change the microenvironment. Delivery system reform essential to control costs. Mod Healthc. 2009;39(17):20–21.



Othman Ouenes