This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

P E R S P E C T I V E S
V A R I A T I O N S
W E B  E X C L U S I V E   

7 October 2004  
  

Perspective: 
Decisions, Decisions: 
Why The Quality Of 
Medical Decisions Matters

If patients aren’t getting the right amount of care,
somebody is making incorrect decisions about
how much and what care to provide or to consume.

By David Blumenthal

ABSTRACT:

Questions about the process leading to particular medical decisions are seldom addressed in discussions of health care quality. In some cases, strict adherence to treatment protocols may overlook important elements in the clinical decision-making process, such as physicians’ judgment and patients’ preferences. Papers by Annette O’Connor and colleagues and Karen Sepucha and colleagues are welcome additions to this debate because they highlight the importance of the medical decision-making process; they suggest that the quality of this process should be considered when evaluating health care quality. I explore some key limitations and uncertainties that must be addressed before work in this area can become broadly applicable.

When the intellectual history of the health care quality movement is written, it will document the convergence of at least three major rivers of thought. At the headwaters of the first is the work of Avedis Donabedian, including his elaboration of the structure/process/outcome paradigm.¹ Hardly a quality discussion occurs now, or will in the future, without reference to these three constructs, which underlie most current approaches to quality measurement.

The second stream of ideas originates with the industrial quality theorists who began their work early in the twentieth century in the Bell Laboratories of the American Telephone and Telegraph Company.² They included Harold Dodge and Walter Shewhart, followed by Joseph Juran, W. Edwards Deming, Kaoru Ishikawa, and A.V. Feigenbaum.³ These thinkers and their disciples created a robust, practical, and widely applicable theory of continuous quality improvement (CQI) that anticipated virtually every innovation in health care quality in the past fifty years.⁴

The third river of thought springs from the work of John Wennberg and his colleagues and draws upon their thirty years of epidemiological research documenting unexplained variations in rates of care received by patients in different geographic areas of the United States and in other countries.⁵ Unlike the other two intellectual traditions, Wennberg did not begin by explicitly addressing how to measure or improve quality, but that has not in any way lessened the importance of his work for quality improvement in health care. An inescapable conclusion of the variations that he and his associates have documented is that some patients are getting too much or too little care and that opportunities exist to improve quality—and potentially reduce costs.

This collection of Health Affairs papers—in particular, the papers by Karen Sepucha and colleagues and by Annette O’Connor and colleagues—are downstream applications and extensions of Wennberg’s work.⁶ The thrust of both papers is that the quality of health care can be improved by attending to the quality of decisions made by health care practitioners and their patients. The Sepucha paper makes the case that measuring decision quality is important to making it better and describes some techniques for doing so. The O’Connor paper discusses a series of technologies—patient decision aids (PtDAs)—whose application might improve decision quality. The papers’ link to the Wennberg tradition is apparent: If patients aren’t getting the right amount of care, somebody is making incorrect decisions about how much and what care to provide or to consume.

Importance of decision quality. At first glance, the idea that decision quality is important to quality of care may seem so obvious as to be trivial. After all, patients commonly visit health care professionals to get help with health care decisions. Some go expecting to have decisions made for them; others go to get information to inform their own choices. Either way, decisions are assumed to be critical to the process. For their part, physicians view decision making as fundamental to their jobs and the quality of their decisions as perhaps the most important marker of their professional competence.

As Sepucha and colleagues point out, however, decision quality has not been a particular focus of efforts to measure and improve the quality of health care up to now. Most of the quality indicators adopted by groups such as the National Committee for Quality Assurance (NCQA), the Leapfrog Group, the National Quality Forum, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), and the Centers for Medicare and Medicaid Services (CMS) measure the extent to which practitioners and health care organizations conform to well-accepted, evidence-based practices. They measure, for example, whether heart attack victims get certain drugs, whether diabetics have their blood sugar tested and controlled, whether women over age fifty get regular mammograms, and whether elderly patients get influenza and pneumonia shots. These are important attributes of care for which providers can and should be held accountable. But these measures are not good indicators of decision quality.

The reason is that the medical literature has made absolutely clear that the measured care should be provided (except in rare circumstances). The decision is a given, made by the science of medicine, and failures to deliver the measured services constitute oversights or lapses in vigilance: in other words, medical errors. In this sense, commonly used measures of quality capture providers’ ability to perform as expected (in industrial terms, to conform to specifications) and are best viewed as indicators of “performance” quality rather than “decision quality.” One reason why some physicians are frustrated with current quality measurement efforts is that they seem not to assess what physicians feel they were trained to do and take the most pride in: to make countless daily decisions about diagnosis and treatment using copious, incomplete, confusing, and changing information, under time pressure and in the face of an ambiguous medical literature. Such decisions may involve things as comparatively simple as how often to tell a patient with multiple chronic illnesses to return for follow-up or as complex as what to do for patient who has just had a heart attack and now needs abdominal surgery for an unrelated problem.

The work of these researchers, therefore, is a welcome addition to the quality literature that, one hopes, will eventually make quality measurement and improvement much more relevant and understandable to the average patient and health professional. However, as is so often the case, the early work in a novel and important field raises as many questions as it answers and leaves much to be done before it can be broadly applicable. In this regard, a couple of key limitations and uncertainties regarding these Health Affairs papers need to be kept in mind.

Limitations and uncertainties. First, the papers are concerned above all with driving home the important point that good medical decisions must be responsive to patient preference. In this, the authors are reflecting and extending a substantial literature on how to make optimal decisions in the face of uncertainty. When science does not make decisions for doctors and patients—when there is no clear right or wrong—then patients’ feelings about which risks to take, how to play the medical lottery, become especially pertinent to making good choices. It is patients, after all, who have to live with the consequences of such decisions. Involving patients is also consistent with principles of patient-centered care, which is increasingly emphasized in the quality literature.

However, incorporating patients effectively into decision making is only one criterion for good decisions, and many other criteria remain to be elaborated. Thus, the use of PtDAs would not in itself guarantee high-quality decisions, nor would Sepucha and colleagues’ measurement strategy capture all critical aspects of health care decision making. For example, physicians know that some of their colleagues make better decisions than others. Physicians-in-training quickly identify these expert decisionmakers among their teachers and look to them as role models. What isn’t clear, and has never been effectively measured or described, are the precise attributes that make these master clinicians—the doctors chosen by other doctors to care for themselves and their families—so effective. The best decisionmakers have an excellent store of medical knowledge. They also have the ability to process and organize large amounts of information. They exhibit good judgment and something more ineffable: wisdom. Ultimately, improving decision quality in medicine will require understanding, measuring, teaching, and supporting these and other critical but elusive capabilities of superb physician decisionmakers.

A second point to keep in mind about these papers is that the techniques they describe are still at an early stage of development. A number of PtDAs have been extensively tested in clinical trials and show considerable promise. However, these technologies tend to focus on a limited number of elective or semi-elective decisions, mostly concerning procedures. Important as these decisions are, they constitute only a small proportion of those facing health professionals and their patients. As O’Connor and colleagues note, the usefulness of PtDAs in helping patients and their caretakers manage the many small and large choices associated with major chronic illnesses over time remains to be tested. Incorporating these tools into the care of common problems may actually require much reengineering of processes of care throughout the health care system.

Getting decision aids into use. This last point raises a general problem that confronts health care innovations in our confusing and confused health care system: how to get them used more widely. Like all approaches to quality improvement, the use of PtDAs or other devices for improving decision quality will have economic consequences. The economics of the interventions described by these papers now favors third parties and employers and threatens providers. The threat to providers flows from two sources. First, incorporating patients’ preferences into decisions seems to reduce use, which also reduces providers’ incomes in a fee-for-service environment. Second, fitting PtDAs into the care process may require disruptive changes in the way health care professionals and institutions organize their work. One may anticipate, therefore, much provider resistance to the kinds of changes envisioned in these papers. Third parties and employers are mobilizing to beat back providers’ opposition on a number of fronts, and they may ultimately prevail with regard to PtDAs and the measurement of decision quality. However, the history of the managed care struggle suggests that doctors’ influence should not be underestimated when external parties attempt to directly intervene in the relationships between health care professionals and patients.

It would certainly be preferable, therefore, if the benefits—economic and otherwise—of incorporating patients’ preferences into medical decisions could be shared among all affected parties. O’Connor and colleagues’ suggestion that third parties pay for the dispensing of information—information therapy—is interesting and deserves further study. The question is whether paying a surgeon or a gastroenterologist to dispense information therapy will adequately compensate for a 25– 30 percent decline in their hernia operations or colonoscopies. It should come as no surprise that the use of PtDAs has flourished outside of fee-for-service settings: in group- and staff-model health maintenance organizations (HMOs) and in the British National Health Service (NHS). As in so many areas of quality improvement, creating a business case for the kinds of changes suggested by these researchers will likely require that providers share the financial risks and gains associated with the innovation. If only managed care and capitation had not gone so badly off the rails in the 1990s!


Looking forward, the challenge confronting policymakers is to pick their battles and try to change prevalent incentives. The quality improvement movement is at a stage where technical innovation and social policy must proceed hand in hand; otherwise, neither will be effective. The question of how to improve decision quality deserves to be on the scientific agenda for quality improvement investigators. Using the results of that scientific progress, if it occurs, will depend on whether insurers, employers, professionals, and patients can reach agreement on policies that will get us off the barricades and back to the real work of health care.

This work was supported in part by grants from the Commonwealth Fund and the Agency for Healthcare Research and Quality (5 R01 HS 013099-01A1).

NOTES

1. A. Donabedian, The Definition of Quality and Approaches to Its Assessment (Ann Arbor, Mich.: Health Administration Press, 1980).
2. M. Walton, The Deming Management Method (New York: Perigee, 1986).
3. K. Ishikawa, What Is Total Quality Control? The Japanese Way (Englewood Cliffs, N.J.: Prentice Hall, 1985); W.E. Deming, Out of the Crisis (Cambridge, Mass.: Massachusetts Institute of Technology Center for Advanced Engineering Studies, 1986); A.V. Feigenbaum, Total Quality Control, 3d ed. (New York: McGraw Hill, 1983); and J.M. Juran, Juran on Leadership for Quality: An Executive Handbook (New York:. Free Press, 1989).
4. D. Blumenthal and A.C. Scheck, Improving Clinical Practice: Total Quality Management and the Physician (San Francisco: Jossey-Bass, 1995).
5. See, for example, K.R. Sepucha, F.J. Fowler Jr., and A.G. Mulley Jr., “Policy Support for Patient-Centered Care: The Need for Measurable Improvements in Decision Quality,” Health Affairs, 7 October 2004, content.healthaffairs.org/cgi/content/abstract/hlthaff.var.54.
6. Ibid.; and A.M. O’Connor, H.A. Llewellyn-Thomas, and A.B. Flood, “Modifying Unwarranted Variations in Health Care: Shared Decision Making using Patient Decision Aids,” Health Affairs, 7 October 2004, content.healthaffairs.org/cgi/content/abstract/hlthaff.var.63.

David Blumenthal (dblumenthal{at}partners.org) is the Samuel O. Thier Professor of Medicine and a professor of health care policy at Harvard Medical School and director of the Institute for Health Policy at Massachusetts General Hospital and the Partners Health Care System, in Boston.

DOI: 10.1377/hlthaff.var.124
©2004 Project HOPE–The People-to-People Health Foundation, Inc.