The trusted source for
healthcare information and
Study casts new doubt on effectiveness of P4P
Study in JAMA compares two groups of hospitals
Does a pay-for-performance (P4P) approach to quality improvement really work? A new study in the June 6 issue of JAMA raises some serious questions as to whether it provides a superior incentive for hospitals to improve their performance.1
For the purposes of the study, the authors chose to compare two groups of hospitals: one that participated in the highly touted Demonstration Project by the Centers for Medicare & Medicaid Services (CMS) — the largest pay-for-performance pilot project to date in the United States; and one group that did not.
"Prior studies had come out, and using Premier [CMS' partner in the project] data had shown that care had improved since the beginning of the pilot," notes Eric D. Peterson, MD, MPH, director of cardiovascular research at the Duke Clinical Research Institute in Durham, NC, and one of the article's authors. "It had been contented that the data showed the program alone was highly successful, but obviously other things had been going on at the same time. So it was not clear how much of the improvement was a direct result of the pilot project."
Peterson says the JAMA study adopted "a more classic control design" to see whether hospitals that had enrolled in a national voluntary quality initiative — Crusade 400 — but did not participate in the CMS project improved at a similar rate (All of the hospitals studied were part of Crusade 400; only a small percentage also participated in the CMS project.) The Crusade 400 collaborative collected clinical information on patients admitted with acute coronary syndrome (ACS) and participants received quarterly feedback, plus a series of other QI tools.
The study included an analysis of data for 105,383 patients with acute non-ST-segment elevation myocardial infarction. Patients were treated between July 2003 and June 2006 at 54 hospitals in the CMS program and at 446 control hospitals. The main outcome measures were the differences in the use of ACC/AHA Class I guideline-recommended therapies and in-hospital mortality between pay for performance and control hospitals.
The researchers found that composite measure scores for CMS processes showed significant improvement at both pay-for-performance and control hospitals. There was no significant difference in the rate of improvement in the composite score between the two hospital groups. Two of the six CMS measures, aspirin prescription at discharge and smoking cessation counseling, had slightly higher rates of improvement at pay-for-performance hospitals than control hospitals. For composite measures of heart attack treatments not subject to incentives, rates of improvement were not significantly different. There was a slight reduction in the mortality rates over time at both pay-for-performance and control hospitals, although the difference in the rate of the reductions between the groups was not statistically significant.
Posing key questions
Peterson said that he and his fellow researchers sought the answers to three key questions:
"All of the hospitals improved over time," notes Peterson. "We could not find an incremental benefit of the P4P relative to those who were not in it, so this was kind of a negative finding."
The "good news," he continues, is that the researchers did not find any harm arising from the P4P project. "Those hospitals did just as well in the adoption of newer therapies that were not part of the CMS 'package,'" Peterson reports. "We also found out overall that as hospitals got better with their care, the outcomes got better — and at a similar rate [among the two groups of hospitals]."
Peterson's overall conclusion: "We did not find that P4P had a large incremental impact on care or outcomes."
This is not good news for the proponents of P4P, to be sure. "I think, at least in its current iteration, P4P may not be the ultimate panacea for quality improvement," Peterson asserts. "Hospitals that are committed to doing quality improvement generally improve with or without this financial incentive."
All hospitals, Peterson contends, are truly interested in QI and in receiving feedback on their progress. "My general recommendation would be for the government to create incentives for all hospitals to be engaged in quality improvement and get feedback on their progress, but whether a financial incentive should be tied to it is unclear," he says. "At least in its current iteration in terms of the amount or the ways of paying for hospitals and health systems to improve, it does not seem to be changing hospital or physician behaviors."
A larger financial incentive, he concedes, "could have a larger effect. Or, you could use funds simply to pay for participation in the program."
[For additional information, contact:
Eric D. Peterson, MD, MPH, Director of Cardiovascular Research, Duke Clinical Research Institute, Durham, NC. Phone: (919) 668-8830.]