3. Measure quality at the level of the organization, rather than the clinician.

Historically, the physician has been viewed as the leader of medicine, with responsibility for the care and outcomes of patients; in iconic photographs and paintings, the physician is seen as a lone, heroic figure. Such a view has led to natural interest in the measurement of individual physicians’ performance. It is therefore not surprising that some information brokers, including the U.S. News and World Report and many city magazines like the Washingtonian, provide ratings of “top doctors,” often based mostly on reputation, warranted or not.

However, this focus on the individual is flawed for most measures of quality and presents substantial technical challenges. Systems-based care is emerging as a key value within health care and a vital component of high-quality care, while the notion that an individual health professional can be held accountable for the outcomes of patients in isolation from other health professionals and their work environment is becoming an outdated perspective. For example, better intensive care unit staffing sometimes mitigates the evidence that surgeons who perform more procedures achieve better outcomes [21].

The communication and coordination of services across providers is required to ensure that patients, many of whom have multiple conditions, are assisted through various health care settings [22]. For some aspects of care, such as diagnosis errors and patient experience, measuring at the individual physician level might be considered. Nevertheless, focusing measurement on an individual runs counter to our goals in promoting teamwork and “systemness” as core health care delivery attributes.

For some professionals whose individual performance does matter, such as a surgeon in the operating room, there are rarely meaningful and valid process measures that reflect their individual performance anyway. In contrast, surgical outcomes depend crucially on the performance of the entire surgical team and the facility in which the procedure takes place.

It is also plausible that individuals respond differently to payment incentives than do organizations; assessment and pay-for-performance at the organizational instead of the individual level should be less likely to crowd out health professionals’ intrinsic motivations to provide high-quality care.

In addition to the conceptual issues with measuring an individual clinician’s performance, technical and statistical issues are also prominent. The attribution of a particular care process or outcome to a particular clinician is often difficult, if not impossible, to make. For example, several specialists, hospitalists, nurses, technicians, and others will typically care for a patient with a heart attack. Good estimates of performance require that the individual or group being evaluated have a sufficient number of observations to make inferences about their performance that are precise enough to be meaningful.

Yet, many physicians and other health care professionals often lack sufficient volumes of certain types of patients to permit valid inferences about their performance. By focusing assessment on the organization, hospital unit, or clinic, rather than the individual clinician, measures can assess and promote team-based care while addressing many of the technical issues that can undermine the value of measurements. For virtually every performance measure evaluated (e.g., safety culture, patient experience, hand hygiene, infection rates, process measures) there is usually substantially greater variation among units within a hospital than among hospitals. The unit or clinic is therefore often the most effective focus for improvement.

While measuring at this level is conceptually right and technically easier than measuring a single individual’s performance, it nevertheless presents challenges. For example, it makes strategic sense to measure the quality of ACOs, especially to guard against the possibility that ACOs would stint on care as they receive increasing incentives to limit spending. Yet, recently, 31 Pioneer ACOs participating in a major CMS demonstration sent CMS a letter criticizing both the agency’s use of measures that “are not yet mature” and the way in which CMS determined the thresholds for acceptable performance.23 We expect they will work through the differences and arrive at a reasonable result.

Finally, measuring at the level of the organization does not mean that substandard individual performance should be tolerated. CMS and its contractors should aggressively use performance measures to identify such unacceptable performance and sanction or otherwise limit the ability of these practitioners to serve Medicare beneficiaries. But the role of measurement for “policing” the performance of individuals is different from public reporting to inform patient choice or to provide financial incentives to improve performance.

Robert A. Berenson, MD is an institute fellow at the Urban Institute.

Peter J. Pronovost, MD, PhD is the director of the Armstrong Institute for Patient Safety and Quality at Johns Hopkins, as well as Johns Hopkins Medicine’s senior vice president for patient safety and quality.

Harlan M. Krumholz, MD, is the director of the Yale-New Haven Hospital Center for Outcomes Research and Evaluation, director of the Robert Wood Johnson Foundation Clinical Scholars program at Yale University, and the Harold H. Hines, Jr. professor of cardiology, investigative medicine, and public health.

The authors thank Lawrence Casalino, MD, PhD, chief of the Division of Outcomes and Effectiveness Research and an associate professor at Weill Cornell Medical College, and Andrea Ducas, MPH and Anne Weiss, MPP of the Robert Wood Johnson Foundation for their helpful comments on this paper. This research was funded by theRobert Wood Johnson Foundation, where the report was originally published.


21. Pronovost PJ, Jenckes MW and Dorman T. “Organizational Characteristics of Intensive Care Units Related to Outcomes of Abdominal Aortic Surgery.” Journal of the American Medical Association, 281(14): 1310-1317, 1999.

22. Pham HH, Schrag D, O’Malley AS, et al. “Care Patterns in Medicare and Their Implications for Pay for Performance.” New England Journal of Medicine, 356: 1130-1139, 2007.

23. Allina Health, Atrius Health, Banner Health Network, et al. “Letter to Rick Gilfillan and Hoangmai Pham, Center for Medicare and Medicaid Innovation.” Feb. 25, 2013,www.washingtonpost.com/blogs/wonkblog/files/2013/03/2013–Quality–Benchmarks.pdf (accessed April 2013).

1 reply »