I disagree that quality is a zero-sum game. Werner and Chang do a good job demonstrating this.

They also show that time spent with the doctor was one of the important factors in quality care. And they suggest that the traditional quality metrics might be encouraging that they spend this time by requiring multiple follow up visits.

In which case, I say, good.

But I also say, if the time doctors spend with their patients is so empirically valuable, why don’t we just say so and set up our system that way?

What I find most troubling is that these measures are being presented as consumer tools for finding good quality. They are not. To the extent they suggest to a consumer that practice A is better than practice B on an objective scale, they are deeply misleading.

But more broadly, I am less interested in how people find the “best doctor,” as I am in the question of getting the right diagnosis and treatment.

Measures of quality ought to be focused on these issues – how well doctors puzzle out these problems – than on whether they do the specific tasks these groups measure.

The key thing you must understand is that nobody suggests that the currently-available publicly reported quality measures represent a complete capture of health care quality. They are quite limited in scope. This is not news.

Quality measurement is already evolving in the direction you suggest. This will require considerable resources, but I believe these would be well-spent. I would love for patients to have a sense of all aspects of quality, including acute care, chronic care, diagnosis, patient experience, and costs.

The real debate is over whether imperfect and incomplete quality measure reporting is better than no reporting at all. In other words, is there some kind of detrimental effect–on patient health–of public reporting on the currently available measures? For this to be the case, you’d have to make the case that reporting on available measures distracts from unmeasured aspects of quality, or that higher performance on the available measures is inherently negatively correlated with other aspects of quality. For example, you might argue that aggressive treatment of blood pressure will cause physicians to ignore other health problems…or to treat too aggressively. Or you might argue that quality is some kind of zero-sum game: if you’re good at one thing, you’re bad at something else.

These are empirical questions, not conceptual ones. Therefore I would encourage you to review the existing literature on them. Though this literature is young (you’re really only looking at a few years of peer-reviewed articles), what you find might surprise you. In particular, check out Werner and Chang in JGIM 2008.

Finally, I would encourage you to think more broadly about the appropriate locus of measuring quality. It is not always the doctor (those who stay fixed on this unit of analysis systematically undervalue other participants in care). As health care is delivered by a system, quality measures are often intended to system-level performance. So again, nobody (at least nobody intelligent, which excepts the Massachusetts GIC) is arguing that the existing quality measures should be reported at the individual-doctor level…or that performance on these measures will help a patient select the “best doctor” for him or her.