Print-Article

Last reviewed by Faculty of Harvard Medical School June 27, 2013

In recent years, a succession of studies and news articles have come out that have turned conventional medical wisdom on its head. As a result, people are learning that many things their doctors have been teaching them are now considered wrong, or potentially wrong. Among the "retractions" and possible retractions are:

Hormone therapy will reduce the risk of heart disease and stroke for women who have gone through menopause. Unless you have been in cave in Tibet, you have read by now that the Women's Health Initiative has shown that postmenopausal hormone therapy actually increases risk of cardiovascular disease, stroke and some cancers. The main thing hormone therapy is now considered useful for is controlling symptoms of menopause, such as hot flashes.
A low-fat diet is good for you. Yes and no. Fat has a high caloric density and reducing dietary fat can help cut calories if you are trying to lose weight. But the real problem is not the amount of fat you eat, but the type of fat. Monounsaturated fats (olive and canola oil) and polyunsaturated fats (vegetable, safflower and sunflower oils) are healthy fats and should be part of a balanced diet. Limit saturated fats and avoid trans fats completely. Also, low-fat diets tend to leave people hungrier, and they fill the void with calorie- and sugar-rich carbohydrates. In response, their bodies crank out insulin, which makes them even hungrier. The result may be a greater tendency to gain weight, and higher risks of high blood pressure and diabetes.
People with atrial fibrillation should be treated with a goal of restoring a normal heart rhythm. This heart-rhythm problem, which is especially common in the elderly, makes the heart beat faster and puts the patient at risk for strokes due to blood clots. The conventional wisdom has been that electrical shocks and medications should be used to put patients back into a normal heart rhythm. But two major trials recently showed that the “cure” may be worse than the disease. In these trials, patients did just as well, and maybe a bit better, if they were left in atrial fibrillation and received drugs to prevent blood clots and control their heart rates.

Examples of the conventional medical wisdom being turned on its head are not limited to cardiology. A major research trial recently showed that a common operation for arthritis of the knee, arthroscopy, turns out to be no more effective for providing relief than a "sham" operation in which patients only had tiny cuts made in their knee. The placebo effect accounted for all of the benefit that physicians and patients have been attributing to this operation.

For physicians who have been dispensing advice about diet, hormone therapy, and how to treat atrial fibrillation, these recent findings are, well, a bit embarrassing. Doctors do not like to be wrong. Their patients, after all, assume that physicians are going to give them advice that is likely to be helpful.

So how are these recent "exposés" good news? They show that we are living in an era of evidence-based medicine. That means that we are investing time, money, resources, and courage in performing research to evaluate our assumptions about what we believe to be good for patients.

The fact is that much of what physicians recommend to patients has never been rigorously tested. The "scientific" era of medicine is really relatively brief and recent — only since the 1970s have researchers been performing large-scale randomized experiments to determine which therapies or testing strategies lead to the best outcomes for patients. Before then, physicians simply made their best guesses. Usually those guesses were right, but sometimes they have been wrong.

Why might our assumptions be wrong? In some cases, we have taken leaps of faith — without taking into account the possibility that drugs or operations might have unanticipated effects. For example, we saw that estrogen lowered cholesterol in women, and we assumed that this would reduce heart disease risk. We didn't fully take into account the fact that estrogen can also increase a person's risk of blood clots.

One small consolation: When it takes a large trial like the Women's Health Initiative to prove that a therapy is more harmful than helpful, that means that the danger to the patient is small. If the danger were large, it would become apparent with a much smaller number of patients. Thus, even though the risk from postmenopausal hormone therapy seems to exceed the benefits, the increase in risk to patients is really very small.

The bottom line: Every cloud has a silver lining. The good news is that we are getting increasingly sophisticated about how to judge the real impacts of medical interventions, and increasingly disciplined about performing these evaluations.

Thomas H. Lee, M.D., is the chief executive officer for Partners Community HealthCare Inc. He is a professor of medicine at Harvard Medical School. He is an internist and cardiologist at Brigham and Women's Hospital. Dr. Lee is the chairman of the Cardiovascular Measurement Assessment Panel of the National Committee for Quality Assurance.

When Wrong Advice Is Good News