Dr. Rob Lamberts also comments on my paper on pay-for-performance (P4P) in Medicare. Lamberts (like Holt) seems to have blogged that paper having only read the press release. Though the paper probably would answer most of the questions they raise, I'll respond to two of Lamberts' comments.

1. Lamberts argues that a P4P experiment in Britain's National Health Service (NHS) refutes my claim that "provider-focused P4P incentives can encourage inappropriate care or reduce access to care for patients with multiple illnesses or low incomes."

Not quite. A P4P scheme can avoid those effects, but not without causing other problems. For example, the financial incentives could involve only additional payments to physicians and no payment reductions for "low-quality" care. That's what the NHS did; physicians' gross incomes increased by an average of $40,000.

A rewards-only approach reduces the incentive for physicians to avoid very sick or very poor patients, who make it difficult for the physician to meet the performance goals. However, that approach makes the P4P effort more costly. Lamberts himself suggests that Medicare's P4P efforts should be budget-neutral, which would make it more likely that physicians would give outlier patients inappropriate care, avoid those patients, or otherwise game the system.

Another way the NHS experiment avoided inappropriate care or a reduction in access for outliers was by allowing physicians the discretion to disregard as many of their patients as they wished when calculating their compliance score. But the availability of such "exclusion reporting" also gave physicians an opportunity to game the system. Rather than provide the desired type of care to their patients, physicians could use exclusion reporting to increase their incomes without changing their behavior. The authors of the study cited by Lamberts note: "More research is needed to determine whether these practices are excluding patients for sound clinical reasons or in order to increase income."

2. Lamberts writes that the Brits "were able to achieve astonishing improvements to their quality numbers and improve physician incomes at the same time."

Of course, these two ends are not in conflict. It's easy to get people to do what you want when you dangle $40,000 in front of them.

But we can't even be sure that the NHS P4P experiment made any improvements in quality — much less astonishing improvments in quality. Although median reported achievement was an impressive-sounding 83.4 percent, according to the authors of that study:

There is no baseline with which to compare performance in the first year of the U.K. program, although the quality of care was already improving before its introduction.

If we don't know what compliance rates were before the NHS introduced financial incentives for compliance, and quality was improving anyway for other reasons, how do we know whether or how much their quality numbers improved, or how much of that change was due to P4P? 

If we don't even know that, we certainly don't know whether the effort was worth the $3.2 billion the NHS spent in 2004.