Late in May, the National Institutes of Health, or NIH, directed that a major research study of hormone replacement therapy, or HRT, be immediately terminated because of safety concerns. Their action, while unusual, would probably not have gained much media attention did not millions of women rely on HRT to moderate the effects (immediate and long-term) of menopause. In light of new evidence, physicians and patients are now rethinking a mode of treatment that hitherto seemed to have little downside risk.

What does this have to do with dentistry? More, perhaps, than one might think. For one thing, when our patients ask our opinion of the study and its implications, we should be prepared to answer intelligently. In addition, I believe that this event, and its aftermath, teach some valuable lessons about risk: what it means, how it is managed in the research setting and how it influences our decisions about health care.

Here are the essential facts as I understand them. The Women’s Health Initiative, or WHI, is an unprecedented multiyear study of the health of some 100,000 peri- and postmenopausal American women.

This very large research program was designed to assess both the natural progression of important medical conditions (cardiovascular disease, cancer, osteoporosis and, in a small way, oral health) and the effects of certain treatments. It was an HRT arm of this study, in which some women were treated with an estrogen-plus-progestin regimen, which was terminated early. Although this was an important component of the WHI, it was certainly not the only one, and the rest of the study is proceeding.

Why did the NIH act as they did? Because, on the basis of the first five years of data, the risks of this kind of hormone replacement were deemed to outweigh the benefits.¹ Specifically, patients receiving HRT were found to experience invasive breast cancer at a higher rate than those not receiving the treatment (38 vs. 30 per 10,000 person-years, or equivalently, 0.38 percent vs. 0.30 percent annually). Similarly increased rates were found for coronary heart disease (0.37 percent vs. 0.30 percent), stroke (0.29 percent vs. 0.21 percent), and venous thromboembolism (0.34 percent vs. 0.16 percent). On the other hand, HRT appeared to reduce osteoporotic fractures and colorectal cancer.

In absolute terms, these are all small numbers, but the relative differences between groups are troubling—for example, HRT doubles the rate of venous thromboembolism. Because of the large number of women in the study, we can be quite confident that these effects are associated with HRT alone and not some other factor. (All these statistics for adverse events apply to women treated with estrogen in combination with progestin; there is, as yet, no comparable evidence concerning unopposed estrogen therapy, which continues to be studied in the WHI.)

If there is a bright side to this disappointing result, it’s the fact that the mechanisms set up to protect research subjects did operate as intended. As soon as the data and safety monitoring board detected that adverse events were occurring disproportionately in the HRT group, they compared the known risks against the potential benefits on the basis of standards established before the study began.

In this case, the former outweighed the latter, and the trial was stopped. Although a drug company was a major participant in the study, and would doubtless have wished that the results had been different, the company did not attempt to delay or obstruct either the decision or its publication. All in all, it appears that everyone acted according to the highest principles of clinical research.

While the statistics may be clear, a patient considering HRT still faces a difficult choice. The elevated risks for cancer, heart disease and thromboembolism are now established and quantified, thanks to the WHI. There also may be unknown side effects, which may not show up for decades. On the other hand, HRT offers clear benefits by reducing osteoporotic fractures, as well as by moderating the unpleasant symptoms often accompanying menopause.

The decision is complex, individual and never totally rational; but armed with facts, and a basis for comparison, most patients will be able to reach a decision based on risk rather than fear.

This leads me to reflect briefly on the nature of risk, particularly as it affects how we make decisions about our health. For such a familiar concept, risk is surprisingly hard to pin down. To the statistician, it’s merely the probability of an undesirable event—a number which, by the way, we typically grossly over- or underestimate. But clearly the gravity of the outcome and the nature of the alternatives also figure in our thinking about risk.

Decisions involving risk are hard enough when the alternatives can be reduced to dollars and cents (whether to buy lottery tickets or stocks, for example), but they get a lot more complicated when the risks and benefits involve apples and oranges, certainties and probabilities, immediate payoffs and deferred hazards.

Is another slice of pie worth a 12-second reduction in your life span? If you are a woman, would you accept a 0.08 percent increased annual risk of developing breast cancer if it would eliminate hot flashes for a year? Do you always select the car with the best safety statistics?

And of course there’s the question of amalgam restorations. Some people feel that a small but unknown risk justifies the cost, discomfort and medical risk (also small but unknown) of their removal and replacement. Are they right? What can we do to help them reach a decision they can live with? Perhaps the WHI can furnish at least a partial paradigm.

First, we must ensure that our advice is based on sound scientific underpinnings. Only studies of unimpeachable quality that produce quantitative findings will have a chance of settling the issue. All involved—researchers, clinicians and (one hopes) advocates—must be prepared to acknowledge the validity of results that disagree with their prejudices.

Second, the data should be linked to recognizable health outcomes, like cancer or stroke, rather than some other measure that is only indirectly related. (Ironically, research on amalgam is handicapped by the fact that demonstrable medical endpoints are so rare as to be beyond the reach of reasonably sized studies; hence the need to rely on blood mercury levels as a rather unsatisfactory surrogate.)

Third, laypeople must be given points of reference from which they can develop an intuitive feel for how serious a risk really is, relative to other more familiar hazards. Fourth, it is useful to identify and, if possible, quantify the risks of the alternatives. Fifth, we must acknowledge that fear is a very powerful and very individual factor in the equation: it takes an extremely low probability to offset a profound fear.

A zero-risk world exists only in the minds of fanatics and litigators. Against junk science, horror stories and testimonials, cold statistics may seem a feeble defense. Yet our patients look to us for those facts to make decisions about their own and their families’ health. Informed patients, by and large, can deal intelligently with the risks and benefits of dentistry; fear and hysteria can only flourish in a climate of ignorance.