P4P: An Evidence Based Approach

By Bernd Wollschlaeger,MD,FAAFP

" The fact that nearly three-quarters of the general internists we surveyed support financial incentives for quality could be surprising, given their concerns about unintended consequences, possible fear of the unknown, and literature on physician professionalism suggesting that physicians prefer not to subject their performance to external oversight. Respondents might simply have seen such incentives as a way to gain more income; this belief might have outweighed
any resistance they had to external oversight."

"The majority supported financial incentives for quality, although they conditioned their support on measures’ being accurate."

Dear Friends and Colleagues:
In the heated discussion about the so-called Pay-For Performance (P4P) Programs we often forget to consider the views of those we claim to represent: the practicing physicians.
I am concerned that ideological tainted viewpoints distort rational approaches towards this issue, thereby failing to consider the pros and cons of a topic that may affect us all.
In a series of reviews I will therefore present on my blog a variety of different articles discussing evidence-based P4P applications:

In the first article ( see attached abstract) the authors present the views of General Internists and their attitude towards public reporting of quality scores.

DESIGN:
To learn more about physicians’ views, the authors conducted a national survey of general internists. Their views, although not necessarily representative of all physicians, are important both because of the large numbers of general internists and because most physician P4P programs focus on primary care.
METHODS:
The sampling frame included 1,668 randomly selected general internists listed in the AMA Physician Masterfile as working in one of the twelve broadly nationally representative metropolitan areas included in the Community Tracking Study (CTS) of the Center for Studying Health System Change (HSC).A cover letter and a seven-page self-administered questionnaire were mailed to the sample in April 2005, with follow-up surveys to nonrespondents mailed in four subsequent waves. The survey was developed after review of the relevant literature, input from experts in P4P and public reporting, and pilot testing with academic and community physicians. The Institutional Review Board (IRB) at the University of Chicago approved the study protocol.
Of the 1,668 physicians selected, 500 were ineligible: 188 surveys were undeliverable; 14 were returned uncompleted with a note stating that the physician was no longer in practice; 46 respondents were not general internists; and 252 additional physicians were not at the address to which the survey had been mailed, according to the phone calls made to nonrespondents’ offices. Of the 1,168 remaining subjects, we received 556 completed surveys, for an adjusted response rate of 48 percent.
RESULTS:
Attitudes toward P4P and public reporting. Almost three-fourths of responding physicians agreed that "if the measures are accurate, physicians should be given financial incentives for quality"
However, only 4 percent strongly agreed and 26 percent somewhat agreed that measures of quality are generally accurate at present. Most respondents had little confidence that this will change: 38 percent believed that health plans will try hard to make the measures accurate, and 35 percent believed that government will do so
There was much less support for public reporting than for financial incentives for quality. Only 45 percent of respondents supported public reporting of medical group performance , and only 32 percent supported reporting of individual physicians’ performance. One in three physicians strongly supported financial incentives, but only one in twelve strongly supported public reporting for medical groups, and one in nineteen strongly supported it for individuals

Attitudes toward quality measures and possible unintended consequences. Only 30 percent of physicians agreed that measures of quality are generally accurate. Eighty-eight percent believed that measures are not adequately adjusted for patients’ medical conditions; 85 percent believed that they are not adequately adjusted for patients’ socioeconomic status; and 82 percent stated that quality measures could lead physicians to avoid high-risk patients. Many physicians added written comments stating that poorly compliant patients would also be avoided: For example, "If my pay depended on A1c values, I have 10–15 patients whom I would have to fire. The poor, unmotivated, obese, and noncompliant would all have to find new physicians."
The survey results suggest that physicians are also concerned that "measuring quality will divert physicians’ attention from important types of care for which quality is not measured." Sixty-one percent strongly or somewhat agreed with this statement; many added written comments stating that many quality measures cover physician activities that they do not consider very important compared with many other things they do.

Are some physicians more likely to support P4P or public reporting? The authors performed bivariate and multivariate analyses to learn whether physician support for P4P or public reporting is associated with whether the physician has financial incentives for quality, sees relatively poor patients (measured by percentage of Medicaid patients), and has an income that is primarily based on patient volume.They also tested they associations with physicians’ age, sex, and board certification.
The main finding on bivariate analyses was that physicians who now have financial incentives for quality were more likely to support such incentives than those who do not. These physicians were also more likely to support public reports on medical groups. One-fourth of respondents (130 physicians) reported having financial incentives for quality;for most, the incentive amounted to 5 percent or less of their income (data not shown). Multivariate analysis produced results consistent with the bivariate analyses.

DISCUSSION:
" Support for P4P. Our results suggest that there is a sizable potential reservoir of physician support for P4P: Nearly three in four internists stated that physicians should be given financial incentives for quality if the measures are accurate. However, responses also suggest that there are barriers to overcome if P4P programs are to gain physicians’ support: Fewer than one-third of internists stated that quality measures are accurate at present, and only slightly more than one-third believed that health plans and the government will try hard to make measures accurate.

Support for public reporting. Internists were far less supportive of public reporting than of financial incentives for quality: One-third supported public reporting of individual physicians’ quality scores, and only 45 percent supported public reporting at the medical group level.

Concerns about financial incentives for quality. Although the majority of respondents supported financial incentives for quality in principle, they appear to have been very concerned about possible unintended and undesirable consequences. Large majorities of respondents stated that quality measures are not adequately adjusted for patients’ medical conditions or socioeconomic status; that measuring quality may lead physicians to avoid high-risk patients; and that measuring quality will divert physicians’ attention from important but unmeasured areas of clinical care.

Concerns about public reporting. Although a large majority of respondents supported financial incentives for quality, a large majority opposed public reporting, especially reporting of individual physicians’ performance. To our knowledge, this is the first survey to report this gap. We do not have data to explain this finding; it is possible that respondents believed that they were unlikely to lose much from having financial incentives for quality but feared that a poor public quality rating would be humiliating and might lead to losses of patients and of peer approval.

Other relevant studies. To our knowledge, this is the first national survey to be published of physicians’ views on P4P and public reporting. Some relevant studies have been conducted in more limited settings. In an e-mail survey of nearly 6,000 Massachusetts physicians (response rate was 7 percent; 29 percent were internists), the Massachusetts Medical Society (MMS) found that respondents were open to quality measurement but did not believe that current measures were accurate. Four physician surveys related to public reporting have been conducted; each surveyed either cardiologists or cardiac surgeons in New York or Pennsylvania, where data on cardiac procedures are publicly reported.The results were similar across the surveys: The majority of respondents stated that the measures were not adequately risk-adjusted and believed that public reporting was leading physicians to avoid doing bypass surgery or angioplasty on high-risk patients.

Possible rationales for survey results. The fact that nearly three-quarters of the general internists we surveyed support financial incentives for quality could be surprising, given their concerns about unintended consequences, possible fear of the unknown, and literature on physician professionalism suggesting that physicians prefer not to subject their performance to external oversight.12 Respondents might simply have seen such incentives as a way to gain more income; this belief might have outweighed any resistance they had to external oversight.

An alternative, or complementary, explanation is that physicians would like to improve the quality of care and believe that P4P would finally give them a "business case" for investing in improving quality.Respondents who reported having financial incentives for quality were more likely to support such incentives: Their experience with incentives might have been positive, although it is also possible that physicians who are open to financial incentives for quality are more likely to join practices that have such incentives.

A third possible explanation is that physicians who oppose P4P were less likely to respond to the survey and that our results therefore overstate the degree of support. However, nonrespondents were very similar to respondents, except that nonrespondents were less likely to be board-certified. Since non-board-certified respondents were equally or slightly more likely to favor financial incentives for quality, this potential source of nonrespondent bias seems unlikely to greatly alter our results.

Possible study limitations. Several possible limitations should be considered in evaluating our results. First, the results are not representative of all U.S. physicians; rather, they are limited to general internists in twelve metropolitan areas. Although these twelve areas are not precisely statistically representative of U.S. metropolitan areas, they represent a broad cross-section of such areas in the country and have been used as the sampling frame for major studies of health care quality.
Second, our response rate was modest, though not atypical for physician surveys; it appears, for the reasons detailed above, that nonresponse bias is unlikely to have greatly affected our results. Third, the survey asked about "financial incentives for quality" in general, not about any specific incentive program, so different physicians might have had different programs in mind when responding.

Implications for quality improvement in public and private programs. Our findings suggest three main implications for Medicare and for other public and private policymakers seeking ways to improve the quality of health care. First, there is a large potential reservoir of physician support for P4P, at least among general internists. However, respondents’ concerns about unintended consequences, and their lack of trust that health plans and government will work hard to make quality measures accurate, suggest that physician support could disappear very rapidly if these consequences do occur. Recent experience with such highly touted innovations as utilization review and primary care gatekeeping suggest that potentially useful policies might generate a strong backlash if they are not framed and implemented with attention to physicians’ concerns.

Second, respondents’ concerns suggest that evaluations of P4P and public reporting programs should be explicitly designed to assess possible unintended consequences on disparities in health care, on physicians who practice in areas of low socioeconomic status, and on the quality of care in important areas of physician practice not included in the program being evaluated.

Third, if further research replicates the large gap we found between physician support for P4P and support for public reporting, policymakers might want to consider whether and how to deal with this difference when designing and sequencing their programs. Although respondents expressed a lack of trust in health plans and in the government, they did not simply oppose change: The majority supported financial incentives for quality, although they conditioned their support on measures’ being accurate. They doubted that measures are adequately risk-adjusted for patients’ medical conditions or socioeconomic status—doubts that are supported in the literature.They were considerably more supportive of public reporting at the medical-group level than at the individual-physician level—a reservation that is supported by research demonstrating the difficulty of creating reliable and valid quality measures for individual primary care physicians.And they were concerned about possible unintended consequences of measuring and rewarding quality—a concern that is supported by economic theory and by experience in other industries.Policymakers might wish to pay close attention to physicians’ concerns both to increase physician support for programs and because these concerns could be quite important for improving the quality of care.
EVALUATIONS OF P4P and public reporting programs should be explicitly designed to assess possible unintended consequences on disparities in health care delivery, on physicians who practice in areas of low socioeconomic status, and on the quality of care in important areas of physician practice not included in the program being evaluated.




MARKETWATCH
General Internists’ Views On Pay-For-Performance And Public Reporting Of Quality Scores: A National Survey
Lawrence P. Casalino, G. Caleb Alexander, Lei Jin and R. Tamara Konetzka

Very little is known about rank-and-file physicians’ views on pay-for-performance (P4P) and public reporting. In a national survey of general internists, we found strong potential support for financial incentives for quality, but less support for public reporting. Large majorities of respondents stated that these programs will result in physicians’ avoiding high-risk patients and will divert attention from important types of care for which quality is not measured. Public and private policymakers might avoid a physician backlash and better succeed at improving health care quality if they consider these concerns when designing P4P and public reporting programs.

Health Affairs, 26, no. 2 (2007): 492-499
doi: 10.1377/hlthaff.26.2.492