More evidence that micromeasures provide poor insight into medical quality

Up to yesterday it was common knowledge that more mammography was a good thing.  This well accepted notion lead reasonable and thoughtful people to assume that we can judge the quality of care in a doctors' office by (in part) looking at the percent of women who have had a mammogram.  Low % = bad, high % = good.

Trying to get doctors to toe the quality line lead rational people to establish 'pay for performance' standards attaching dollars to more mammograms.

As of yesterday, we're in a bit of a bind.  One of the very thoughtful groups that make recommendations on screening is the US Preventive Services Task Force.  Yesterday they changed their recommendation saying that women between 40-50 should not receive mammograms and 50-70 year old women should only get them every two years (as opposed to the earlier 'every year' recommendation). 

Why the change?  Turns out that doing more mammograms results in more harm that good to the women according to recent studies.

This is hardly the first time we've seen such changes.  It used to be common knowledge that radical mastectomy - an extremely disfiguring surgical procedure - was the best way to manage breast cancer.  That is until studies demonstrated no additional value to women when compared to much less disfiguring approaches.

Now we have to unravel a complex tapestry of quality measures.  Of course quality measures will shift over time, but there's a significant chance of shift and perverse incentives when our quality focus is on the details of care we deliver.  Details are very important in care - but they provide a poor proxy for understanding the quality of care delivery. 

The US is too deep in the weeds on quality measurement.  We're chasing mammography rates, prostate cancer screening (yet another screening test with dubious credentials) and a host of other details.  Those metrics are important, but pale in comparison to how well an office eliminates barriers to access, how well clinicians communicate with and support their patients in becoming effective self-managers.

Big issues that make a huge difference in the quality of care: access, continuity, comprehensiveness of services & care coordination get short shrift or lip service while we chase the shifting sands of medical details.  These cardinal features of high quality care can be measured, can be used to improve office practices, and have a proven and substantial impact on overall quality of care, the experience of care and when patients receive care in offices performing well on these criteria, the total cost of their care is substantially less.

Enough.  The measurement community needs to break out of its rut & create a new agenda focused on what's meaningful, not what they've deemed measurable.

Comments

[L Gordon Moore]

Evan, great comments and Iove your blog post on the topic.

Gordon

[Evan Falchuk]

Excellent post, and great observations.

Too often, quality measurement treats medicine as an assembly line process. It's not.

Quality medical care requires judgment, reflection, thought. These things are systematically undervalued.

The failure to place those things at the center of any consideration of quality is at the core of so many of our problems in health care.

Some related points here: http://bit.ly/ox73Z

Cheers,

Evan Falchuk