Before and after guidelines - clinical practice guidelines - Editorial

Journal of Family Practice, April, 1997 by Douglas B. Kamerow

Clinical practice guidelines seemingly have become ubiquitous in health care Wherever we turn, they are being created, touted, decried, implemented, revised, or (rarely) withdrawn. Unrealistic expectations frequently accompany them. Medical specialty societies hope guidelines will improve care while, perhaps not coincidentally, protecting their turf. Health care purchasers hope they will decrease their costs without compromising care. Health plans hope they will help meet the increasing demand on then to document the quality of their care. State legislatures see them as a chance to "fix" politically troublesome symptoms without addressing the root problems facing the health care delivery system. The public and many practicing physicians ate confused What's going on?

THE GROWTH OF GUIDELINES

Four factors have contributed to the enormous growth of guidelines in the United States. The first is the extraordinary increase in US health care costs. The second is the nature of the US health care delivery "non-system." The third is the well documented but unexplained variation in the delivery of health care services by region and medical specialty. And the fourth is the recent dramatic group of changes being wrought by the growth of managed care.

Through the 1980s and early 1990s, health care costs have grown at double-digit rates. As a percentage of gross domestic product, national health care expenditures grew from 8% in 1975 to 10% in 1982 to 13.6% in 1993. For comparison, in Europe the 1993 estimated growth rates were 9.8% in France, 8.6% in Germany, 7.5% in Sweden, 7.3% in Spain, and 7.1% in the United Kingdom. (1)

Unlike most European countries, the United States has a decentralized, pluralistic health care delivery and financing system, with little if any central planning and control. The federal government controls only the payment for health services to the elderly (the Medicare program), but even in this case does not actually deliver services. The individual states control the Medicaid program, for the poorest Americans, with some contribution and minimum standards set federally. Most Americans pay for their health care with insurance provided by employers, and it is in this sector that the most dramatic recent changes have taken place.

Wennberg (2) and others have clearly documented dramatic geographic and specialty variations in surgery rates, medication use, hospitalizations, and delivery of other health care services in the United States. These variations ate often unexplained by diagnosis or severity of disease. They bespeak gaps in our knowledge either of what to do or of how to implement what we do know.

In the last 10 years, employee-sponsored health insurance has changed from being an overwhelmingly fee-for-service system in which patients chose their doctors and payments were based on services rendered. It is now predominantly a managed care model, in which patient choice is limited to a specific small or large panel of physicians, services are closely scrutinized, and payment to physicians is either at a negotiated rate or capitated.

These four factors have affected the growth and use of guidelines in different directions. Geographic and specialty-specific variations in care and the growth of health care costs and of managed care have provided strong incentives for assessing and controlling clinical practices to help keep costs down while attempting to maintain high-quality care. Our decentralized, pluralistic system, however, makes consistent application of universally endorsed guidelines virtually impossible. Net result? Lots of guidelines, lots of systems to use them in, lots of frustrations.

It is thus no surprise that this issue of the Journal has no fewer than five articles and editorials on the creation and use of guidelines. Each provides us with important lessons.

IT'S ALL VARIABLE

Croft et al (3) add to the voluminous literature on unexplained practice variations. In an analysis of heart failure treatment using 1991-1992 National Ambulatory Medical Care Survey data, they find that use of the recommended angiotensin-converting enzyme (ACE) inhibitors varied widely by specialty. In addition to providing one more example of seemingly inappropriate variations in care, this article will also be used as ammunition in the "turf wars" to substantiate claims that subspecialists treat seriously ill patients with more appropriate medications than generalists. The message I get from this article, however, is the need for more research to better understand the causes of these variations and how to decrease them.

EVIDENCE, EVIDENCE, MY KINGDOM FOR EVIDENCE

Everett and Chesebro (1) fault the National Cancer Institute's interim cervical cytology management guidelines for lacking a strong evidence base. They thus can muster only a weak endorsement for implementing them. Two important points here. First, most guidelines cannot be based exclusively on evidence. We just do not know that much about most of what happens in the practice of medicine. Second, we should probably focus our energies, especially our energies for preventive care, on those interventions that have been proven effective. Before a guideline topic is chosen, criteria should be established to make sure that important, potentially implementable recommendations are likely to emerge. That is the beginning of an effective guideline. But it is by no means the end.