When President Obama announced his support for “evidence-based” medicine, I became very concerned. Might patients think our past medical practices and recommendations were not based on evidence? Recent events have validated my concerns, but for entirely different reasons. Many politicians, patients, and even physicians it appears, completely oppose basing clinical practices on sound evidence!

Consider the varied responses to the vaccine for H1N1 (“swine”) flu. While many people were concerned about the lack of adequate supplies, many others vociferously refused vaccination for themselves and, even more worrisome, for their children. Why? Because of unsubstantiated fears that the vaccine was “novel,” was inadequately tested, and might well give them swine flu. As health officials repeatedly made clear, this was utter nonsense. The vaccine was prepared in exactly the same way that all flu vaccines are prepared. Indeed, had the H1N1 virus appeared during the seasonal flu season, it would simply have been one of the components of this year’s seasonal flu vaccine. Irrational members of the public remain undeterred in their opposition. A sizeable fraction of society does not believe in the safety and efficacy of any vaccines, particularly childhood vaccines. They attribute all manner of disease to their use, from autism to type I diabetes. Their vocal opposition has needlessly confused and sowed concern among many young parents.

A second surprise was the outpouring of hostility that greeted the U.S. Task Force on Preventive Services (USTFPS) recommendation that screening guidelines for breast cancer be altered—in particular, that mammography not be routinely performed (except in high-risk individuals) prior to 50 years of age, and then only every 2 years. Why? Because early mammography provides little benefit, and generates relatively large numbers of false-positive tests that result in biopsies and sometimes more invasive interventions. The hysteria that greeted these recommendations, among advocacy groups, politicians, and even some physicians, was not new; more than a decade ago an independent expert advisory committee appointed by the then-director of the National Cancer Institute (NCI) advised against routine mammography before age 50. Within hours of the release of their report, the NCI director (who appointed them) denounced their recommendations, as did many members of Congress. The NCI’s “official” recommendation remained annual mammography beginning at age 40.

What in the world is going on here? Lots; none of which bodes well for those of us who believe that treatments and recommendations should be based upon a careful weighing of the best available evidence.

The first problem is our highly charged political environment. Those partisans opposed to health care reform pounced on the recommendations as evidence of a federal conspiracy to ration medical services. That of course is ridiculous, since the Task Force had been working on the subject for years, and its recommendations carry no regulatory force. Two decades ago the Agency for Healthcare Research and Quality (AHRQ) issued clinical guidelines questioning the value of much of the back surgery being performed. Back surgeons rose in righteous anger, while their partisans in Congress gutted the AHRQ’s funding (after failing to obliterate the agency entirely).

The second problem is the nature of evidence. Most of what physicians do is based on “shared wisdom”; interventions for which we have rock-solid, definitive data from randomized trials are very much in the minority. Since “experience” and medical “culture” vary, there are wide variations in the way perfectly well-meaning physicians treat the same condition. The USTFPS, first established in 1984, followed the example of its Canadian predecessor in rigorously rating the quality of the evidence upon which its recommendations are based. Often the quality of existing evidence is so weak that the Task Force refuses to issue any recommendation at all, as happened several years ago when it attempted to address the value (if any) of routine prostate-specific antigen (PSA) testing. The Preferred Practice Patterns (PPP) of the American Academy of Ophthalmology also rate the quality of the evidence for each of their recommendations.

A third issue concerns what, precisely, a recommendation seeks to address. The USTFPS would no doubt agree that the earlier mammography is begun, the larger the number of tumors that would be discovered. But the issue they were addressing was not the total number of tumors detected, but the outcome that earlier screening would have on people’s lives. They therefore had to consider, as well, the number of tumors mammography would miss, which would present the patient with a false sense of security; the number of false positives, which would result in unnecessary and potentially harmful follow-up investigations; and whether finding and treating small tumors (whose biology may be very different from large and aggressive ones) makes any difference in patient outcomes. A recent analysis suggests that early screening decreases mortality from colon cancer, but probably has had no impact on deaths from either prostate or breast cancer. Widespread adoption of annual PSA testing, for example, has resulted in a virtual “epidemic” of prostate cancer (and treatment), largely because it identifies many small, potentially innocuous lesions; but the incidence of late-stage, lethal lesions has not fallen correspondingly. This delicate weighing of perspectives can lead reasonable people to alternative conclusions. The United States has long recommended that women undergo Pap smears at more frequent intervals than do most countries of Western Europe.

Lastly, ordinary physicians and patients do not necessarily respond to data (and recommendations) rationally, particularly recommendations that involve subtle trade-offs. Both patients and physicians are influenced by their own prejudices, fears, traditions, and, increasingly, pharmaceutical advertising. A decade ago, definitive trials demonstrated that old-line, inexpensive thiazide diuretics were as effective for the initial treatment of systemic hypertension as new, more expensive classes of drugs. This has not had any perceptible impact on prescribing practices. Similarly, 2 highly promoted (and related) drugs for reducing the risk of cardiovascular disease, Zetia (ezetimibe) and Vytorin (which combines ezetimibe with a statin), have yet to demonstrate superior clinical outcomes to statins alone. These 2 prescription drugs nonetheless earned their manufacturers over $3.5 billion last year.

We should all welcome additional support for “head-to-head” trials of competing treatment modalities and renewed emphasis on “evidence-based medicine.” Ophthalmology is particularly well placed to obtain evidence-based advice, not only from the Academy’s PPP, but also from the Cochrane Library. The National Eye Institute supports the U.S. satellite of the Cochrane Eyes and Vision Group at Johns Hopkins, an important methodological infrastructure that assists vision scientists in developing and publishing systematic reviews of available evidence as part of a Cochrane Review. The Cochrane Library also maintains a list of all known controlled trials related to vision (over 11 000 at last count) for those wishing to undertake their own analyses and reviews.

Unfortunately, while funding and support for such studies are extremely valuable, they won’t, by themselves, broadly improve clinical practice. The British National Institute for Health and Clinical Excellence (“NICE”) makes evidence-based recommendations that guide clinical practice in the National Health Service. It has been vocally denounced by those opposed to any health care reform in the United States, and held up as the harbinger of the much maligned, and imaginary, “death panels.” Those who insist that patients and doctors need the freedom to choose whatever course of therapy they desire, and think that “more medicine” is always “better medicine,” forget the harm that can be done and that, for most people, personal finances (and their insurance policies—should they be fortunate enough to have medical insurance) rarely provide the opportunity for truly informed decisions about their care. The medical profession has a responsibility to convince its members to behave rationally. The less that politicians have to say, the better.