Perspective

Engaging Specialists in Performance-Incentive Programs

Jeffrey O. Greenberg, M.D., M.B.A., Jessica C. Dudley, M.D., and Timothy G. Ferris, M.D., M.P.H.

N Engl J Med 2010; 362:1558-1560April 29, 2010

Article

During the debate over U.S. health care reform, there were widespread calls for increasing “value” in the health care system — calls reflecting concerns about suboptimal health outcomes and rapidly growing health care costs. The health care reform law contains funding for pilot projects to test new ways of increasing our return on health care spending, including global capitation, bundled payments, and medical homes, as well as reduced payments for readmissions. All these options would require the quality of care to be measured and incentives (financial or otherwise) provided to physicians in an effort to improve performance.

Until now, the most common form of performance-based incentives for physicians has been pay-for-performance programs devised by private payers and, more recently, Medicare. Over the past decade, such programs have grown in number and intensity, shifting their focus from clinical processes to outcomes. However, most programs have focused on primary care physicians rather than specialists. In a year-long survey initiated in July 2004, 27.8% of primary care physicians reported receiving some compensation that was based on quality metrics, as compared with only 17.8% of medical specialists and 12.6% of surgeons.1 Yet specialists are responsible for a large and growing proportion of patient care. The proportion of all office visits that were visits to generalists (physicians in family practice, general medicine, pediatrics, or obstetrics–gynecology) dropped from 66.2% in 1980 to 57.5% in 2004, whereas the proportion consisting of visits to specialists grew from 33.8 to 42.5%.2 Thus, to have a meaningful impact on the quality of care, pay-for-performance programs and newer-generation quality-incentive programs must engage more specialists.

Engaging specialists in such programs is challenging for several reasons. First is the obvious challenge of developing a menu of meaningful metrics for each of the dozens of medical, surgical, and pediatric specialties. The relevance and validity of each metric must be supported by strong evidence. National societies with broad representation from a given field should agree that each metric reflects the quality of care and is applicable to many patients. As specialties beget subspecialties, these demands become even more challenging. Control of low-density lipoprotein (LDL) cholesterol may be an appropriate performance goal for a general cardiologist but may not accurately capture the quality of the care delivered by an electrophysiologist. Glucose control in patients with diabetes may be an appropriate measure for most endocrinologists, but not for thyroid specialists. This increasing degree of subspecialization requires similar increases in the specialization of performance metrics if they are to have validity and gain acceptance. This requirement makes the task of developing metrics and engaging specialists more complex. The Physician Consortium for Performance Improvement sponsored by the American Medical Association has generated 266 performance measures for specialists to date, but many of these measures have not been endorsed by the National Quality Forum, and there remain substantial barriers to progress in developing such measures.3

Second, determining which of a patient's multiple physicians should be held accountable for particular outcomes can be difficult, especially when the patient has complex medical needs. If a patient is followed by both a primary care physician and a cardiologist, who is responsible for the control of the patient's LDL cholesterol levels? In reality, both physicians are, and both should be accountable for the coordination necessary to achieve control. As patients live longer and benefit from the evolution of medical technology, they are often cared for by many physicians, for varying durations, in multiple settings, over many years. Assigning responsibility for outcomes becomes increasingly complicated.

The third challenge in engaging specialists in performance-incentive programs is achieving adequate risk adjustment. Risk adjustment is also a concern for generalists, but the problem is amplified for specialists, especially those based at academic medical centers, where they focus heavily on complex cases and regularly provide second opinions. Without accurate risk adjustment, it would be unfair to compare the work of community-based specialists with that of specialists at academic medical centers; it might also cause harm by encouraging the “cherry picking” of healthier patients. Such “risk-avoidance creep” has been seen in several states in conjunction with the public reporting of mortality rates after coronary-artery bypass surgery and cardiac catheterization.4

Fourth, even if appropriate performance metrics are developed for specialists, measuring performance accurately requires considerable infrastructure. To minimize the cost and administrative burden, performance data should be captured electronically rather than manually, which involves expensive, time-consuming chart reviews. Ideally, data should be culled from electronic medical records (EMRs) rather than third-party claims. EMRs capture a full range of clinical data from all the patients in a provider's panel, whereas claims are limited to billing data from individual payers, which are often hard to aggregate for collective analysis. Unfortunately, a recent survey suggests that only 15% of specialists have adopted even basic EMRs.5 Furthermore, not all EMRs are configured to measure and report performance, and most are not connected to networks that can coordinate uniform and consistent data collection.

Eventually, health informatics platforms may provide a foundation for performance measurement, and ongoing work sponsored by the Office of the National Coordinator for Health Information Technology will accelerate this process. It will be several years, however, before most specialists will be using EMRs capable of feeding clinical data to the requisite analytical engines. Since the drive for payment reform and physician incentives will not wait, we must find solutions that work with our existing infrastructure while we invest in EMR adoption.

When possible, specialists should be given flexibility in choosing metrics from menus (see tableAdvantages and Disadvantages of Various Possible Pay-for-Performance Metrics for Specialist Physicians.) that have been developed by specialist societies and approved by the National Quality Forum. In addition, practices and departments should be encouraged to develop their own performance goals and targets, which may be more appropriate for their circumstances than preexisting ones — although they should be required to demonstrate that their metrics are important to patients and address a real deficit in quality or cost-effectiveness. Although such flexibility might make interpractice comparisons difficult, it would initiate physicians into the process of measuring performance and striving to improve it. The measure of quality improvement should be based on year-to-year change within a practice, not on comparisons among physicians.

Until more robust EMR systems are in place to capture the subtle data from multiple sources that are needed to appropriately adjust for severity of illness, we should focus on measures that reflect the processes involved in high-quality care, not just the outcomes. Many of these measures are not unique to specific subspecialties. Examples include the effective use of electronic records for documentation, prescribing, and communication with patients and the use of evidence-based checklists before procedures are performed.

When outcome measures are used, we should consider holding practices and institutions accountable collectively rather than holding specific physicians accountable individually. Measuring the collective performance of physicians encourages communication and collaboration — hallmarks of high-quality care. If a primary care physician and a cardiologist are both held accountable for the control of LDL cholesterol levels in a shared patient, they are more likely to work together to ensure proper treatment.

Finally, we must continue to encourage and provide incentives for the adoption of interoperable EMRs. The American Recovery and Reinvestment Act of 2009 included a $20 billion down payment on EMR adoption and the collection and reporting of performance data. Further incentives could come from differential reimbursements from public and private payers for the use of EMRs that can feed clinical data into systems for quality measurement. The Physician Quality Reporting Initiative of the Centers for Medicare and Medicaid Services is the beginning of an attempt at “pay for reporting.”

If appropriate incentives are a cornerstone of quality improvement, they must follow patients through the health care system — and touch all physicians. The sooner we develop and implement a framework for engaging specialists, the sooner we can truly pursue our goal of improving the quality of care for everyone.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

This article (10.1056/NEJMp1000650) was published on April 21, 2010, at NEJM.org.

Source Information

From the Department of Medicine, Brigham and Women's Hospital, and the Brigham and Women's Physicians Organization (J.O.G., J.C.D.); the Department of Medicine, Massachusetts General Hospital, and the Massachusetts General Physicians Organization (T.G.F.); and Harvard Medical School (J.O.G., J.C.D., T.G.F.) — all in Boston.

References

References

1. 1

   Reschovsky J, Hadley J. Physician financial incentives: use of quality incentives inches up, but productivity still dominates. Issue Brief Cent Stud Health Syst Change 2007;108:1-4
   Medline

2. 2

   Health, United States, 2008 (with chartbook). Hyattsville, MD: National Center for Health Statistics, 2009.
   

3. 3

   Ferris TG, Vogeli C, Marder J, Sennett CS, Campbell EG. Physician specialty societies and the development of physician performance measures. Health Aff (Millwood) 2007;26:1712-1719
   CrossRef | Web of Science | Medline

4. 4

   Resnic FS, Welt FG. The public health hazards of risk avoidance associated with public reporting of risk-adjusted outcomes in coronary interventions. J Am Coll Cardiol 2009;53:825-830
   CrossRef | Web of Science | Medline

5. 5

   DesRoches CM, Campbell EG, Rao SR, et al. Electronic health records in ambulatory care -- a national survey of physicians. N Engl J Med 2008;359:50-60
   Full Text | Web of Science | Medline

Citing Articles (8)

Citing Articles

1. 1

   Castigliano M. Bhamidipati, Damien J. LaPar, George J. Stukenborg, Charles J. Lutz, Margaret C. Tracci, Kenneth J. Cherry, Gilbert R. Upchurch, John A. Kern. (2012) Transcatheter arterial revascularization outcomes at vascular and general surgery teaching hospitals and nonteaching hospitals are comparable. Journal of Vascular Surgery
   CrossRef

2. 2

   Jha, Ashish K., Joynt, Karen E., Orav, E. John, Epstein, Arnold M., . (2012) The Long-Term Effect of Premier Pay for Performance on Patient Outcomes. New England Journal of Medicine 366:17, 1606-1615
   Full Text

3. 3

   David L Carpenter, Sara R Gregg, Daniel S Owens, Timothy G Buchman, Craig M Coopersmith. (2012) Patient care time allocation by nurse practitioners and physician assistants in the intensive care unit. Critical Care 16:1, R27
   CrossRef

4. 4

   Sonali P. Desai, Jinoos Yazdany. (2011) Quality measurement and improvement in rheumatology: Rheumatoid arthritis as a case study. Arthritis & Rheumatism 63:12, 3649-3660
   CrossRef

5. 5

   John D. Freedman, Alice Bendix Gottlieb, Paul F. Lizzul. (2011) Physician performance measurement: Tiered networks and dermatology (An opportunity and a challenge). Journal of the American Academy of Dermatology 64:6, 1164-1169
   CrossRef

6. 6

   Asaf Bitton. (2011) Who is on the Home Team? Redefining the Relationship Between Primary and Specialty Care in the Patient-Centered Medical Home. Medical Care 49:1, 1-3
   CrossRef

7. 7

   Rodney Johnson, Chitra Dinakar. (2010) Pediatric Pay-for-performance in Asthma: Who Pays?. Current Allergy and Asthma Reports 10:6, 405-410
   CrossRef

8. 8

   Robert Lawrence Reed. (2010) How to Achieve High-Quality Low-Cost Trauma Care. Annals of Surgery 252:2, 223-224
   CrossRef