“All initial results are positive.” Legions of medical students have had this aphorism beaten into their brains as they (we) learned to evaluate the scientific literature. A British Medical Journal article in 1978 introduced the term “therapeutic illusion” to describe unwarranted enthusiasm for positive intervention by both physicians and patients—ignoring coincident causative factors.
Measurement of physician performance is a perfect example of these phenomena. Every practicing physician in the United States is measured. His or her resource use (aka “cost”), processes of care (did the patient get a glycosylated hemoglobin this year?), coordination of care (did you send a letter to the primary care physician?), and outcomes of care (return to the OR rate) are collected, sampled, and analyzed. They are increasingly linked to access to patients and to payment and are used as proxies for determining “value.”
Bending of the health care cost curve, changes in patient satisfaction, and avoidance of surgical errors have all been attributed to 1 or more “quality measurement” metrics. How many times have you heard the maxim “If you can’t measure it, you can’t manage it”? It’s attributed to W. Edwards Deming, PhD, a leader in the field of quality improvement. However, as Robert A. Berenson, MD, recently pointed out in the JAMA Forum,1 Dr. Deming actually said, “It is wrong to suppose that if you can’t measure it, you can’t manage it—a costly myth.” Further, according to Dr. Berenson, Dr. Deming cautioned that the most important data needed to manage often are unknown and unknowable.
Those of us who deal almost constantly with health care policy and regulations have long decried an even more perverse factor in some health care metrics: They are not just useless and time consuming, but they lead to harmful conclusions and actions. How can you compare the costs and outcomes for 2 equally capable physicians with different practices types without adjusting for case severity? The tertiary care expert who manages the most complex patients could have her or his access to patients and payments restricted as a result of faulty metrics and analytics.
This is not to say that metrics have no place. They are critically important, but great care must be taken to ensure that they are well constructed, appropriately adjusted, accurately tallied, standardized, and thoughtfully reported. Most important, they must be clinically meaningful! And, finally, the entire process should be designed primarily not to be punitive, controlling, burdensome, and even threatening. It should be designed and implemented to recognize the genuine desire of physicians to practice nothing but the best medicine and to assist in our lifelong struggle to maintain scientific currency and to improve quality.
In a recent Health Affairs article, Casalino et al. concluded that reporting quality measures cost the physician practices in only 4 specialties $15.4 billion—mostly in time devoted to the task.2 Over 90% of that burden was attributed to entering information. This is not just a financial burden; a 2014 Merritt Hawkins survey found that 5% of physician time is spent dealing with external quality measures—or an annual “loss” of 40,000 full-time equivalent physicians at a time when many specialties acknowledge a manpower shortage.3
One of our Academy’s primary health policy and advocacy objectives will continue to be battling against poorly conceived, badly designed, and erroneously analyzed physician performance measures. At the same time, we will work to promote those few metrics that are clinically meaningful to patients and physicians, are not burdensome, can be accurately measured and risk adjusted, and will help each of us render better care in a value-sensitive system.
1 http://newsatjama.jama.com/2016/01/13/jama-forum-if-you-cant-measure-performance-can-you-improve-it. Accessed April 7, 2016.
2 Casalino LP et al. Health Aff. 2016;35(3):401-406.
3 A Survey of America’s Physicians: Practice Plans and Perspectives. 2014. www.merritthawkins.com/compensation-surveys.aspx. Accessed April 7, 2016.