We make judgments about the quality of physicians all the time – “She’s a star,” “He’s average,” or “I wouldn’t send my worst enemy to him.” But how do we make those judgments? The hallmarks of quality are complex. As primary care clinicians, we put weight in diagnostic prowess, knowledge base, good judgment, and communication skills. Patients also value communication skills, as well as compassion, accessibility, and responsiveness.

Current quality metrics tend to focus on positive health outcomes, and these measures are being used to judge (via payment adjustments) the quality of both individual physicians and health systems. But do they really tell us who provides the highest quality care?

In “Quality Measures: How to Get Them Right,” Ronald Adler, MD (no relation), and colleagues say that most of the quality measures used in pay-for-performance programs have not been shown to improve quality. They suggest that all quality measures should be assessed to make sure they are truly evidence-based and relevant and oriented toward positive outcomes. They conclude that payers should pull back, reduce the number of quality measures to just those proven to provide outcomes that matter to patients, and reevaluate the others.

I would take it a step further. Outcome-oriented quality measures generally shouldn’t be applied to individuals at all but only to health systems. Current measures are influenced by many variables besides the physician’s skill, such as the patient’s economic situation, psychology, and beliefs about medications and the health care system.

The measures that better reflect a family physician’s technical and communication skills tend to be process measures. For instance, I don’t mind being judged on what percent of my patients got a flu shot, but I’d be reluctant to be judged on what percent of my patients got the flu.

So where should we go from here? I agree that we should work to improve quality measures and make them more evidence-based. But outcome measures should be used for judgment and quality improvement only in large health systems. In primary care, if we are going to reward individual physicians for quality, we should focus more on process measures (e.g., did the physician do the right thing, communicate well, and show compassion?). Meaningful process measures and well-designed patient experience surveys could be used to that end.