Walking the fine line between the burden of too much quality measurement and not enough

April 11th, 2018 / By L. Gordon Moore, MD

Does the “more is better” approach to quality measurement create a burden on healthcare delivery? Does it distract attention from actual care and consume resources that might be used to help improve outcomes? A recent commentary in the New England Journal of Medicine (NEJM) Perspective, authored by Michael Chernew, PhD and Mary Beth Landrum, PhD, describes a way to walk the fine line between the burden of too much measurement and not enough.

The “more is better” approach recognizes that tracking a set of disease parameters often fails to generalize to overall performance. Does great diabetes care result in great care for people with congestive heart failure and/or cancer, for example? We’ve seen that focused disease management might not translate to overall improvement. The “more is better” approach is based on the premise that we must add more disease measures to the mix to better reflect overall quality.

The problem with “more is better,” say the authors, is that U.S. healthcare providers are spending $15.4 billion a year on quality measurement and the pushback on “more is better” is significant. Chernew and Landrum suggest an alternative approach where a core set of measures could act as an “initial pass.” This core set could be designed as a minimally burdensome set of metrics serving as a screening test.  Those who fail the screening would be required to submit a set of supplemental metrics. This supplemental set would provide more granular information that would assess quality performance with greater certainty though with greater burden.

The authors’ description of the inadequacy of the core measure set is based in part on the limitations of disease focused, small denominator measures and the lack of risk-adjustment. These problems could be addressed by using risk-adjusted outcome metrics like potentially preventable hospital admissions, emergency department visits, readmissions, and the like.  These metrics are better because:

  • Coming from administrative data they have very low/no reporting burden
  • Are risk adjusted
  • Are more reflective of a system of care delivery as opposed to a measure that might not represent systematic improvement beyond the a specific disease process

The bottom line: Chernew and Landrum provide a good framework for quality assurance measurement, one that can be improved by using risk-adjusted population-based core measures.

L. Gordon Moore, MD, is Senior Medical Director, Clinical Strategy and Value-based Care for 3M Health Information Systems. 


How do we develop quality measures that help improve outcomes without creating more work for clinicians? Listen to our podcast episode.