Analysis of Physician Pay for Performance

Spending

Whether or not widespread adoption of physician pay for performance (P4P) will affect health care spending depends strongly on program design:

  • Physician P4P design features that may affect spending include the financial incentive structure and funding and the behaviors or actions rewarded. Read more below
  • Estimates of the effects on spending must reflect the administrative costs of P4P programs. Read more below
  • P4P could change the distribution of spending across subgroups of physicians. Read more below
  • The only published study linking physician P4P with spending shows a savings, but this result might not generalize to other P4P programs. Read more below

Physician P4P design features that may affect spending include the financial incentive structure and funding and the behaviors or actions rewarded.

Design features likely to reduce spending include (1) rewards for activities that decrease use of expensive services, such as hospital readmissions and certain diagnostic and therapeutic technologies, and that reduce hospital length of stay; (2) rewards for system investments, process redesign, and better coordination of care delivery that result in efficiency gains or improvements in patient safety (Centers for Medicare & Medicaid Services [CMS] would have to revise Diagnosis-Related Group payment rates to hospitals to recoup savings accrued through improvements in efficiencies); and (3) a system design that leads to reductions in morbidity and mortality risks, which in turn lead to lower spending (e.g., increased use of medications that prevent a recurrence of a problem that would likely lead to hospitalization).

Design features likely to increase spending include (1) rewards for increasing use of necessary and appropriate services—particularly those that have previously been under-provided—that do not lead to future reductions in health service use; and (2) adding new money to pay for incentives. The UK pay-for-performance program for primary care physicians financed its incentive payment through new funding: In the first year of the program, the National Health Service spent $1.8 billion in new funds for incentive payments (a 20 percent increase in spending for family practice) (Galvin, 2006; Doran et al., 2006).

No change in spending would result if the program is designed to be budget neutral (i.e., if the incentive payment is financed by redeploying existing funds in the system).

Estimates of the effects on spending must reflect the administrative costs of P4P programs.

New infrastructure is frequently required to run P4P programs. The resulting administrative costs include collecting and analyzing data; auditing, developing, and maintaining performance measures; and supporting providers. These ongoing costs are not trivial and vary with the size and scope of the P4P program.

P4P could change the distribution of spending across subgroups of physicians.

P4P programs could redistribute payments across different subgroups of providers (e.g., by region or by size of practice), depending on how performance differs across subgroups and how the program is designed. For example, in the Integrated Healthcare Association P4P program in California (based on unpublished research by Damberg et al., discussed during personal communication with Cheryl Damberg in 2008), larger, more integrated medical groups achieved higher performance scores and received larger payments on a per capita basis than did smaller, less integrated groups, such as Independent Practice Associations.

Additionally, national health plan health management organization (HMO) Healthcare Effectiveness Data and Information Set (HEDIS) scores for clinical effectiveness and patient experience are lower in the Western United States than in the Northeast; consequently, a design that used national benchmarks as a basis for determining which entities receive the incentive payment could potentially shift spending between regions. While the use of financial incentives implies redistributing dollars, using regional benchmarks is one strategy that could work to mitigate the redistribution of resources across geographic regions.

The only published study linking physician P4P with spending shows a savings, but this result might not generalize to other P4P programs.

Curtin et al. (2006) evaluated a private sector P4P program sponsored by Excellus Health Plan and conducted by the Rochester Individual Practice Association (RIPA). Using data from 2003 and 2004, the authors tracked the projected spending trend for diabetes care and found an average net savings of $2.4 million per year associated with providing more reliable diabetes care. The savings estimates accounted for new spending to provide underused services for managing patients with diabetes. The largest savings came from reducing hospitalizations, followed by physician cost, pharmacy, and outpatient spending. This study has not been replicated either within RIPA or in other P4P programs, so it is unclear whether the results obtained from this intervention would generalize to other settings.

The CMS Web site provides some data on the experience of ten physician groups participating in the CMS Physician Group Practice Demonstration (CMS, 2008). Four physician groups experienced favorable financial performance under the demonstration's performance payment methodology-which rewarded groups through a gain-sharing arrangement if they demonstrated reductions in spending.

For patients with diabetes or coronary artery disease, Medicare expenditures grew more slowly. The four physician groups--Dartmouth-Hitchcock Clinic, The Everett Clinic, Marshfield Clinic, and the University of Michigan Faculty Group Practice--earned $13.8 million in performance payments as their share of $17.4 million in Medicare savings for improving the quality and cost efficiency of care. This compares to two physician groups that earned $7.3 million in performance payments under the first year of the demonstration (CMS, 2008).

References

Centers for Medicare & Medicaid Services (CMS), U.S. Department of Health and Human Services, Web site, "Physician Groups Earn Performance Payments for Improving Quality of Care for Patients with Chronic Illnesses," news release, August14, 2008. As of October 31, 2008: http://www.cms.hhs.gov/apps/media/press_releases.asp (search keywords: physician groups earn)

Curtin K, Beckman H, Pankow G, Milillo Y, Greene RA, "Return on Investment in Pay for Performance: A Diabetes Case Study," Journal of Healthcare Management, Vol. 51, No. 6, November/December 2006, pp. 365-376.

Doran T, Fullwood C, Gravelle H, Reeves D, Kontopantelis E, Hiroeh U, Roland M, "Pay-for-Performance Programs in Family Practices in the United Kingdom," New England Journal of Medicine, Vol. 355, No. 4, July 27, 2006, pp. 375-384.

Galvin R, "Pay-for-Performance: Too Much of a Good Thing? A Conversation with Martin Roland," Health Affairs, Web Exclusives [Epub September 5, 2006], Vol. 25, No. 5, September/October 2006, pp. w412-w419.

Back to top

Consumer Financial Risk

Not applicable

Back to top

Waste

Little is known about how physician pay for performance (P4P) programs will affect waste:

  • The published studies that show reductions in use and spending after a multifaceted intervention are from a single location. Read more below

The published studies that show reductions in use and spending after a multifaceted intervention are from a single location.

Theory suggests that waste can be reduced by decreasing the overuse of services and in improving efficiencies in the delivery of care. However, few performance measures exist that address overuse, and broader resource use measures to identify inefficient providers are only beginning to be tested and applied.

In a study by Greene, Beckman, and Mahoney (2008) that examined relative resource use (i.e., efficiency indexes) measures for physicians in the Rochester Individual Practice Association (RIPA), the authors identified key factors that explained variation in resource use among physicians within selected conditions. For hypertension management, the use of name brand, higher cost angiotension receptor blocking (ARB) medications rather than generic, lower cost angiotension converting enzyme (ACE) medications was a key factor in the variation; RIPA reduced the use of ARBs to manage hypertensives and accrued substantial savings.

Reductions in expenditures were also achieved through RIPA physicians' decreased use of fiberoptic laryngoscopy between 2004 and 2007. A study by Greene et al. (2004) found that a multifaceted intervention that included education, physician profiling, and a financial incentive led to a 20 percent reduction in inappropriate radiology use for management of acute sinusitus. This study compared use rates before and after the intervention for approximately 900 physicians in RIPA. The financial incentive was a withholding of 15 percent on each physician service. Unfortunately, neither of the studies by Greene allows any determination of whether there is a discrete P4P effect, and it is unclear which aspect of the intervention was the key driver of change.

Curtin et al. (2006) evaluated a private sector P4P program sponsored by Excellus Health Plan and conducted by RIPA. The program increased the use of appropriate care for diabetes and achieved reductions in waste from avoiding hospitalizations. However, this study has not been replicated either within RIPA or in other P4P programs, so it is unclear whether the results obtained from this intervention would generalize to other settings.

In an effort to reduce wasteful spending, the Centers for Medicare & Medicaid Services (CMS, 2008) currently test various approaches to measuring the relative resource use of providers within the Medicare program. These measures may be incorporated into future value-based purchasing efforts within the Medicare program.

References

Centers for Medicare & Medicaid Services (CMS), U.S. Department of Health and Human Services, Web site, "Physician Groups Earn Performance Payments for Improving Quality of Care for Patients with Chronic Illnesses," news release, August 14, 2008. As of October 31, 2008: http://www.cms.hhs.gov/apps/media/press_releases.asp (search keywords: physician groups earn)

Curtin K, Beckman H, Pankow G, Milillo Y, Greene RA, "Return on Investment in Pay for Performance: A Diabetes Case Study," Journal of Healthcare Management, Vol. 51, No. 6, November/December 2006, pp. 365—376.

Greene RA, Beckman H, Chamberlain J, Partridge G, Miller M, Burden D, Kerr J, "Increasing Adherence to a Community-Based Guideline for Acute Sinusitis Through Education, Physician Profiling, and Financial Incentives," American Journal of Managed Care, Vol. 10, October 2004, pp. 670—678.

Greene RA, Beckman HB, Mahoney T, "Beyond the Efficiency Index: Finding a Better Way to Decrease Overuse and Increase Efficiency in Physician Care," Health Affairs, Web Exclusives [Epub May 20, 2008], Vol. 27, No. 4, July/August 2008, pp. w250—w259.

Back to top

Reliability

Studies provide mixed results about the link between physician pay for performance (P4P) programs and the reliability with which appropriate care is provided. It is unknown what the effect of P4P would be on improving the reliability of appropriate care absent other interventions that typically occur simultaneously (e.g., quality improvement, public reporting):

  • Studies that examine how physician P4P affects the reliability with which appropriate care is delivered show mixed results. When the results are positive, the gains are relatively modest. Read more below
  • Multiple interventions occur simultaneously with P4P programs, making it difficult to determine the discrete effect of physician P4P. Read more below
  • Physician P4P may have unintended consequences, depending on program design features. Read more below

Studies that examine how physician P4P affects the reliability with which appropriate care is delivered show mixed results. When the results are positive, the gains are relatively modest.

The published research on use of incentives in health care during the late 1990s shows a mix of findings (Petersen et al., 2006). None of the seven most rigorously designed studies (i.e., those using randomized, controlled trials) showed a clearly positive effect on reliability: Four studies reported mixed results (Fairbrother et al., 1999, 2001; Kouides et al., 1998; Roski et al., 2003), and three reported no effect (Grady et al., 1997; Hillman et al., 1998, 1999). The least rigorously designed studies tended to report positive results for at least one aspect of the programs examined (Amundson et al., 2003; Armour et al., 2004; Fairbrother et al., 1997; Francis et al., 2006; Gandhi et al., 2002; Greene et al., 2004; Levin-Scherz, DeVita, and Timble, 2006; Morrow, Gooding, and Clark, 1995; Rosenthal et al., 2005; Pearson et al., 2008; Young et al., 2007).

The results of these studies should be interpreted with caution. Because the studies are based on early versions of P4P, they do not mirror current P4P design structures, and thus the results may not generalize to current P4P program structures or those being proposed for application in Medicare. The early P4P experiments tended to have short durations of six months to a year, dealt with only a handful of measures (1—5), and involved small financial incentives (a few dollars per service or less than $1,000 per physician).

Evidence emerging from the newer, more robust P4P experiments in the United States and abroad shows either modest gains or no additional gains above secular trends in improvement when compared with groups not exposed to the P4P intervention. The five quasi-experimental studies of more recent P4P interventions in the United States and the United Kingdom report mixed findings, and when gains are observed, they are modest (1—3 point gains on a 0—100 scale) (Levin-Scherz, DeVita, and Timble, 2006; Pearson et al., 2008; Rosenthal et al., 2005; Young et al., 2007).

P4P has produced moderate improvements for some clinical measures (Campbell et al., 2007; Rosenthal et al., 2005). The Centers for Medicare & Medicaid Services (CMS) Physician Group Practice Demonstration, which rewards physicians for improving the quality and cost efficiency of health care services delivered to a Medicare fee-for-service population, was the first P4P initiative for physicians under the Medicare program. It found that at the end of the first performance year all ten participating physician groups improved the clinical management of diabetes patients. Results at the end of the second year of the demonstration show that physician groups increased their quality scores by an average of 9 percentage points for diabetes, 11 percentage points for heart failure, and 5 percentage points for coronary artery disease.

Pearson et al. (2008) found that P4P contracts in Massachusetts were not associated with greater improvements in clinical quality measures compared with improvements occurring in the broader community of providers that were not exposed to financial incentives. Thus the improvement among groups participating in P4P was indistinguishable from improvements seen in the non-P4P groups.

In a study of PacifiCare network physician groups in California, Rosenthal et al. (2005) found that groups exposed to the P4P intervention saw modest improvements for the three measures tracked, although similar improvements were also observed in the non-participating Pacific Northwest groups. In this study, a statistically significant greater increase was found among the P4P groups than among the non-P4P groups for one measure (i.e., a gain of 3.6 percentage points in cervical cancer screening scores). There was no evidence of a significant difference for the other two measures.

A study by Young et al. (2007), which compared performance before and after implementation of a P4P intervention, found that absolute levels of performance increased across all measures for physicians in the Rochester Individual Practice Association (RIPA), but found that no difference occurred in improvement trends before and after the intervention. This study concluded that increases in performance were largely due to improvements occurring more broadly in the provider community and could not be attributed to P4P.

Francis et al. (2006) found a post-P4P decrease in the use of less effective/inappropriate antibiotics for management of acute otitis media in RIPA. This study examined the effect of a multifaceted intervention that included targeted education, a reporting system, a modest financial incentive (P4P), and feedback to physicians. Overall, within the Excellus-RIPA P4P program, which funded the incentive by withholding a portion of the base provider payment and where 40 percent of the P4P dollars were withheld and not returned unless improvements in efficiency occurred, pediatricians reduced their use of inappropriate antibiotics by 41.5 percent, and internists and family physicians reduced their use by 22.1 percent and 14.7 percent, respectively. However, this study does not separate the independent effect of P4P in leading to care improvements.

Primary care physicians (PCPs) in the United Kingdom are participating in one of the largest P4P experiments in the world. PCPs have been exposed to P4P since late 2004, when the British government modified contracts with physicians and allowed physicians to earn additional dollars if they met performance targets. Campbell et al. (2007) studied a small number of UK physician practices for which trend data were available (1998—2003) for three measurement domains prior to the implementation of P4P in 2004: coronary heart disease, diabetes, and asthma. Among this convenience sample of physician practices, substantial improvements were observed in clinical processes of care between 2003 (just before the start of P4P) and 2005 (post-P4P). Average practice scores (on a scale of 0—100) increased from 76.2 to 85.0 for coronary heart disease, from 70.4 to 81.4 for diabetes, and from 70.3 to 84.3 for asthma. These increases were statistically significant; however, they continued earlier large improvement trends fostered by an intensive quality improvement intervention led by the National Health Service during the 1998—2003 pre-P4P period. Although the diabetes and asthma measures showed a statistically significant faster rate of improvement (of about 2—6 percentage points) post-P4P implementation, P4P did not lead to a greater rate of increase in improvement for coronary heart disease measures than that observed during the quality improvement intervention period (1998—2003).

Because of the small number of practices involved in the study, it is unclear whether the findings generalize to most primary care practices in the UK. These results may also not generalize to the United States, as the UK program provided a large infusion of dollars and systems support to primary care physicians. The government provided an electronic health record to all physician practices, dramatically raised physician salaries prior to implementing P4P, and provided a bonus potential that represents 30 percent of a physician's income (worth approximately $30,000/year). These design features contrast with those typically found among the existing set of U.S. P4P experiments, which tend to rely on reallocating existing money to finance incentives, have financial incentives that are relatively small (approximately $1,500—$5,000 per physician), and do not provide an electronic health record for patient management or other systems support.

Multiple interventions occur simultaneously with P4P programs, which makes it difficult to determine the discrete effect of physician P4P.

Given the absence of good control groups in most studies, it is difficult to distinguish the effects of physician P4P from the effects of other quality improvement programs, public reporting, and P4P. Many of the measures currently used in physician P4P programs are incorporated into multiple quality improvement programs, so disentangling the net effects of P4P incentives is difficult.

Physician P4P may have unintended consequences, depending on program design features.

Some have expressed concern about P4P causing unintended consequences, such as "gaming," whereby physicians inappropriately avoid treating or exclude from analysis (exception reporting) patients whose poor health behavior and socioeconomic status might adversely affect their ability to perform well on the clinical measures. Few studies have explicitly examined this issue. A study of the UK P4P program, which permits physicians to exclude patients from measurement for various reasons (e.g., patient is noncompliant), found that only 1.1 percent of physician practices excluded more than 15 percent of their patients, indicating that large scale gaming did not occur (Doran et al., 2006). However, tracking the unintended consequences of P4P programs is critical to understand whether P4P leads to non-measured areas being ignored ("teaching to the test") or gaming.

In a separate study by Doran et al. (2008), the researchers examined whether disparities in care had worsened under P4P for UK physician practices that served large populations of socioeconomically disadvantaged populations of patients. This study of 7,637 general practices tracked changes in the median performance by level of disadvantage, broken into five equal groups, ranging from least to most disadvantaged in terms of patient population characteristics. In the first year of the P4P program, the median performance was 86.6 percent for the group of practices with the fewest disadvantaged patients (quintile 1) compared with 82.8 percent for the practices with the highest fraction of disadvantage patients (quintile 5). Between years one and three of the P4P program, the median increase in performance was 4.4 percentage points for quintile 1 as opposed to 7.6 percentage points for quintile 5; by 2006 the gap in median performance had narrowed between these groups of practices, from 4.0 percentage points to 0.8 percentage point. This indicates that P4P in the UK program did not exacerbate disparities in care and in fact led to reductions in disparities.

It is possible that incentives based on fewer quality measures may encourage physicians to neglect other important aspects of care or provide inappropriate tests to improve overall performance. However, no evidence is currently available about this concern (Epstein, Lee, and Hamel, 2004).

References

Amundson G, Solberg LI, Reed M, Martini EM, Carlson R, "Paying for Quality Improvement: Compliance with Tobacco Cessation Guidelines," Joint Commission Journal of Quality and Safety, Vol. 29, No. 2, February 2003, pp. 59—65.

Armour BS, Friedman C, Pitts MM, Wike J, Alley L, Etchason J, "The Influence of Year-End Bonuses on Colorectal Cancer Screening," American Journal of Managed Care, Vol. 10, September 2004, pp. 617—624.

Campbell S, Reeves D, Kontopantelis E, Middleton E, Sibbold B, Roland M, "Quality of Primary Care in England with Introduction of Pay for Performance," New England Journal of Medicine, Vol. 357, No. 2, July 12, 2007, pp. 181—190.

Centers for Medicare & Medicaid Services (CMS), U.S. Department of Health and Human Services, Web site, "Physician Groups Earn Performance Payments for Improving Quality of Care for Patients with Chronic Illnesses," news release, August14, 2008. As of October 31 2008: http://www.cms.hhs.gov/apps/media/press_releases.asp (search keywords: physician groups earn)

Doran T, Fullwood C, Gravelle H, Reeves D, Kontopantelis E, Hiroeh U, Roland M, "Pay-for-Performance Programs in Family Practices in the United Kingdom," New England Journal of Medicine, Vol. 355, No. 4, July 27, 2006, pp. 375—384.

Doran T, C Fullwood, Kontopantelis E, Reeves D, "Effect of Financial Incentives on Inequalities in the Delivery of Primary Care in England; Analysis of Clinical Activity Indicators for the Quality and Outcomes Framework," Lancet, Vol. 372, No. 9640, August 30, 2008, pp. 728—736.

Epstein AM, Lee TH, Hamel MB, "Paying Physicians for High-Quality Care," New England Journal of Medicine, Vol. 350, No. 4, January 22, 2004, pp. 406—410.

Fairbrother G, Friedman S, Hanson KL, Butts GC, "Effect of the Vaccines for Children Program on Inner-City Neighborhood Physicians," Archives of Pediatric and Adolescent Medicine, Vol. 151, No. 12, December 1997, pp. 1229—1235.

Fairbrother G, Hanson K, Friedman S, Butts G, "The Impact of Physician Bonuses, Enhanced Fees, and Feedback on Childhood Immunization Coverage Rates," American Journal of Public Health, Vol. 89, No. 2, February 1999, pp. 171—175.

Fairbrother G., Siegel M, Friedman S, Kory P, Butts G, "Impact of Financial Incentives on Documented Immunization Rates in the Inner City: Results of a Randomized Controlled Trial," Ambulatory Pediatrics, Vol. 1, No. 4, July—August 2001, pp. 206—212.

Francis DO, Beckman H, Chamberlain J, Partridge G, Greene RA, "Introducing a Multifaceted Intervention to Improve the Management of Otitis Media: How Do Pediatricians, Internists and Family Physicians Respond?" American Journal of Medical Quality, Vol. 21, No. 2, March 2006, pp. 134—143.

Gandhi TK, Francis CE, Puopolo AL, Burstin HR, Haas JS, Brennan TA, "Inconsistent Report Cards: Assessing the Comparability of Various Measures of the Quality of Ambulatory Care," Medical Care, Vol. 40, No. 2, February 2002, pp. 155—165

Grady KE, Lemkau JP, Lee NR, Caddell C, "Enhancing Mammography Referral in Primary Care," Preventive Medicine, Vol. 26, No. 6, November 1997, pp. 791—800.

Greene RA, Beckman H, Chamberlain J, Partridge G, Miller M, Burden D, Kerr J, "Increasing Adherence to a Community-Based Guideline for Acute Sinusitis Through Education, Physician Profiling, and Financial Incentives," American Journal of Managed Care, Vol. 10, October 2004, pp. 670—678.

Hillman A, Ripley K, Goldfarb N, Nuamah I, Weiner K, Lusk E, "Physician Financial Incentives and Feedback: Failure to Increase Cancer Screening in Medicaid Managed Care," American Journal of Public Health, Vol. 88, No. 11, November 1, 1998, pp. 1699—1701.

Hillman A, Ripley K, Goldfarb N, Weiner J, Nuamah I, Lusk E, "The Use of Physician Financial Incentives and Feedback to Improve Pediatric Preventive Care in Medicaid Managed Care," Pediatrics, Vol. 104, No. 4, October 1999, pp. 931—935.

Kouides R, Bennett N, Lewis B, Cappuccio J, Barker W, LaForce F, "Performance-Based Physician Reimbursement and Influenza Immunization Rates in the Elderly: The Primary-Care Physicians of Monroe County," American Journal of Preventive Medicine, Vol. 14, No. 2, February 1998, pp. 89—95.

Levin-Scherz J, DeVita N, Timble J, "Impact of Pay-for-Performance Contracts and Network Registry on Diabetes and Asthma HEDIS Measures in an Integrated Delivery Network," Medical Care Research and Review, Vol. 63, No. 1, Suppl., February 2006, pp. 14S—28S.

Morrow RW, Gooding AD, Clark C, "Improving Physicians' Preventive Health Care Behavior Through Peer Review and Financial Incentives," Archives of Family Medicine, Vol. 4, No. 2, February 1995, pp. 165—169.

Pearson SD, Schneider EC, Kleinman KP, Coltin KL, Singer JA, "The Impact of Pay-for-Performance on Health Care Quality in Massachusetts, 2001—2003," Health Affairs, Vol. 27, No. 4, July/August 2008, 1167—1176.

Petersen LA, Woodard LD, Urech T, Daw C, Sookanan S, "Does Pay-for-Performance Improve the Quality of Health Care?" Annals of Internal Medicine, Vol. 145, No. 4, August 15, 2006, pp. 265—272.

Rosenthal MB, Frank RG, Li Z, Epstein AM, "Early Experience with Pay-for-Performance: From Concept to Practice," Journal of the American Medical Association, Vol. 294, No. 14, October 12, 2005, pp. 1788—1793.

Roski J, Jeddeloh R, An L, Lando H, Hannan P, Hall C, Zhu S-H, "The Impact of Financial Incentives and a Patient Registry on Preventive Care Quality: Increasing Provider Adherence to Evidence-Based Smoking Cessation Practice Guidelines," Preventive Medicine, Vol. 36, No. 3, March 2003, pp. 291—299.

Young GJ, Meterko M, Beckman H, Baker E, White B, Sautter KM, Greene R, Curtin K, Bokhour BG, Berlowitz D, Burgess JF, "Effects of Paying Physicians Based on Their Relative Performance for Quality," Journal of General Internal Medicine, Vol. 22, No. 6, June 2007 pp. 872—876.

Back to top

Patient Experience

The effect of physician pay for performance (P4P) on patient experience is uncertain:

  • No published studies evaluate the effect of physician P4P on patient experience; one unpublished study found no incremental effect of P4P on improving patient experience. Read more below

No published studies evaluate the effect of physician P4P on patient experience; one unpublished study found no incremental effect of P4P on improving patient experience.

Research by Damberg et al. (unpublished; discussed during personal communication with Cheryl Damberg in 2008) on the Integrated Healthcare Association's P4P program shows that participating physician groups saw no incremental increase in year-to-year improvements in patient experience scores after the introduction of P4P, compared with improvement trends from previous years, in which only public reporting of quality scores occurred. Year-to-year average improvements statewide across all 162 physician groups stayed constant at around 0.5 to 1.0 percentage point per year, both before and after implementation of P4P, for individual and composite measures of patient experience. However, some groups within the 162 physician groups saw much larger increases (on the order of 4—6 percentage points). These groups had implemented effective communication with patients and had instituted system changes (e.g., providing next day lab results online), both of which were based on targeted interventions with physicians.

Back to top

Health

No studies directly link physician pay for performance (P4P) programs to improved health outcomes:

  • Theory suggests physician P4P could improve health by increasing the reliability of delivered care. Read more below
  • Most P4P programs do not track effects on health, partly because of the long time horizon necessary to see some of the health gains. Read more below

Theory suggests P4P programs could improve health by increasing the reliability of delivered care.

P4P programs typically are aimed at promoting the use of evidence based practices that have been shown empirically to improve health outcomes (Shekelle, 2007). If the P4P program does this successfully, it may lead to improved health by reducing both morbidity (e.g., reduced risk of other diseases, fewer complications of chronic illness, earlier detection of diseases) and mortality.

Most P4P programs do not track effects on health, partly because of the long time horizon necessary to see some of the health gains.

Although very little literature has been published on the health benefits of physician P4P programs, the health benefits could be modeled. The modeling could be based on estimates of reliability improvements and health outcome data from controlled trials that are linked to those improvements.

References

Shekelle P, "Medicare's Hospital Compare Performance Measures and Mortality Rates," Journal of the American Medical Association, Vol. 297, No. 13, April 4, 2007, pp. 1430—1431; author reply, p. 1431.

Back to top

Coverage

Not applicable

Back to top

Capacity

Not applicable

Back to top

Operational Feasibility

Broad implementation of physician pay for performance (P4P) is technically feasible but will require substantial investments in infrastructure and resources to support implementation and ongoing operations. Implementation is especially challenging in small physician practices because of the absence of dedicated staff to assist with implementation and adherence:

  • Physician P4P programs require significant infrastructure to collect and analyze data, manage appeals, provide performance feedback, educate providers, and calculate and administer payouts. Read more below
  • Physician P4P programs are operationally feasible, but we do not know the level of resources required for successful operation. Read more below

Physician P4P programs require substantial infrastructure to collect and analyze data, manage appeals, provide performance feedback, educate providers, and calculate and administer payouts.

Major infrastructure requirements include data warehousing; data aggregation, programming, and analysis; data auditing; appeals management and data correction; performance feedback; physician communication and education; measures maintenance; and payout computation and distribution (Damberg et al., 2007; Sorbero et al., 2006). Some of this infrastructure may already exist in areas in which performance measurement and reporting activities are already under way. Operational issues can arise not only for the program sponsor but also for those whose performance is being assessed. Investments will need to be made by physicians (and group practices) to create data systems and tracking mechanisms that will allow them to generate the data used to produce performance measures and to engage in quality improvement activities.

Physician P4P programs are operationally feasible, but we do not know the level of resources required for successful operation.

There are more than 100 physician P4P programs in operation; in addition, the Centers for Medicare & Medicaid Services has both a physician group P4P and individual physician pay-for-reporting initiatives under way. This indicates that such programs are operationally feasible. We also know that programs lacking sufficient resources may be less successful in engaging program participants. However, to date, P4P programs have not measured or published the level of resources required for operations.

References

Damberg CL, Sorbero ME, Mehrotra A, Teleki S, Lovejoy S, Bradley L, An Environmental Scan of Pay for Performance in the Hospital Setting: Final Report, Washington, D.C.: Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, WR-474-ASPE/CMS, November 2007. As of May 15, 2009: http://aspe.hhs.gov/health/reports/08/payperform/PayPerform07.html

Sorbero ME, Damberg CL, Shaw R, et al., Assessment of Pay-for-Performance Options for Medicare Physician Services: Final Report, Washington, D.C.: Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, WR-391-ASPE, May 2006.

Back to top