Pay-for-Performance Plan Shows Limited Benefit in Quality of Care
I tweeted the NEJM link yesterday and I have read the entire article. This news piece on the article is accurate.
During the pre-introduction period from 1998 to 2003, quality of care improved at a steady rate for all three conditions.
Following the short-term gains in the year after introduction, the rate of improvement leveled off somewhat for asthma and diabetes, such that the rate was similar to that during the pre-introduction period.
At all time points, the mean quality-of-care scores were higher for those aspects of care linked to an incentive — for instance, prescription of short-acting bronchodilators or HbA1C testing.
The quality of those aspects of care not linked to an incentive declined for patients with asthma or heart disease, whereas those governed by an incentive continued to increase.
The researchers also looked at patients’ perceptions about certain aspects of care, including access to a doctor within 48 hours (which was associated with an incentive under the pay-for-performance plan), continuity of care, and ability to communicate with the physician.Introduction of pay for performance did not have any effect on access to care or on communication, but there was an immediate reduction in continuity of care (P<0.001), and it persisted at the lower level.
The researchers said this could be the result of practices focusing on the incentive associated with getting patients in to see any doctor within 48 hours.
These results should not surprise anyone. They highlight a potential danger of P4P programs. When you shine a light on a corner of practice, you logically slightly darken the remainder of practice, because we each only can bring a limited amount of energy to each encounter.
Everyone plays to the test. Thus, we should not establish tests without understanding what the unintended consequences will be.
P4P does influence performance measures in a minor way. I continue to oppose P4P as a concept. This article convinces me that my position has merit.
{ 3 comments… read them below or add one }
The largest risk of P4P is the reality that high socioeconomic patients will perform better than low. Any rudimentary study of health will uncover this reality. Poor people die at a rate 3 times that of rich. Thus if the measure of my medical practice is going to be performance , choose to take care of well off patients.
So if you measure outcomes, rich people will pay much better than poor. If you pay for process type performance, you may do a bit better.
Will P4P efforts actually change long term outcomes? The jury is out, simply because we have not had enough time to do studies to measure this. One reason that P4P may not make much difference is that in general health itself is only weakly related to medical services anyway. Most experts on this field will tell you that your health and longevity are influenced by medical services only to a maximum of 25% of total effects. Your social and economic circumstances create the 75% effects.
The most interesting effect of universal health care that I see is that it will in reality raise the real income and relative status of the currently uninsured. This by a direct effect will imporve thier health,independent of health care services. This effect is well known, but will be unpredictabe in its size economically. But it definitely will push people up the socioeconomic ladder, improving their health to some degree.
You should take a closer look at Figure 2 in the NEJM article you reference. This figure, and the accompanying text, deal specifically with the issue of how P4P affected quality of non-incentivized care. It’s not surprising that for maybe 3 of the 4 conditions (I say maybe due to the multiple comparisons issues in their statistical analyses), financial incentives led to larger quality gains than for the non-incentivized conditions. BUT ALL MEASURES–INCENTIVIZED OR NOT–EXPERIENCED IMPRESSIVE OVERALL QUALITY GAINS WITH THE INTRODUCTION OF P4P.
So the “but-for” question of whether P4P resulted in quality degradation for any non-incentivized condition is as follows: if anything, P4P raised quality for all measures, not just those incentivized. Whether the trends for asthma or heart disease will persist is purely speculative…and even if they did, in the case of asthma it looks like it would take many years for quality to fall back down to the pre-P4P level for asthma (in the non-incentivized category).
Sooo…the news article you quote is indeed misleading through selective reporting of minor findings rather than the larger issue. Interesting that you didn’t pick this up.
Apparel|Ed Hardy Men Clothing|Ed Hardy Men T-Shirts|Ed Hardy Women T-Shirts|Footwear|Jerseys|Sunglasses|Tiffany Jewelry http://www.80trade.com
Nike Air Max|Jordan Shoes|Puma Shoes|Prada Shoes|Supra Shoes http://www.vmytrade.com