If there’s one thing everyone agrees on, it’s that preventive care is always a good thing. Well, I’m a doctor and I’m afraid of preventive medicine.

The theory behind preventive medicine is sound. It is better to treat prevent disease than to treat it. It is better to refrain from smoking and never get lung cancer than it is to treat lung cancer. It is better to refrain from alcohol abuse than to treat alcoholic cirrhosis of the liver. In each of those cases, avoiding a behavior known to cause the disease is highly effective in reducing the incidence of the disease.

But not all preventive medicine is about avoiding behaviors known to cause diseases. Preventive medicine has held out the possibility of avoiding naturally occurring diseases by correcting hormone, mineral or other imbalances through eating specific foods, taking supplements or using medication. Many of these preventive efforts have not only been unsuccessful, they have created problems of their own, sometimes the very problems they were meant to prevent.

The paradigmatic example is estrogen replacement therapy. Menopause, characterized by a lack of estrogen, is associated with increased risk of a variety of health problems including heart disease and osteoporosis. The reasoning behind estrogen replacement therapy was that if heart disease or osteoporosis are associated with decreased estrogen, replacing that estrogen will reduce heart disease and osteoporosis. There was some experimental evidence supporting that theory, but not a lot. Nonetheless, estrogen replacement therapy became the standard of care well before large scale, long term studies could be completed. It was preventive therapy; how could it cause any harm?

Merely replacing a missing hormone is not as simple as it sounds. Hormones, like many other substances in the body, are involved in more than one system. Indeed, lots of substances play multiple roles in multiple systems. Adding back the missing hormone can have an impact far beyond the system it was designed to protect and that impact can be harmful. The data is not final, but it appears that adding back estrogen increases the risk of breast cancer. And while estrogen replacement did have a beneficial effect on bone health, large scale, long term studies have not delivered the promised benefit of reducing the risk of heart disease. Routine postmenopausal estrogen replacement is no longer the standard of care; it is reserved only for specific situations in which the benefit is judged to be worth the risk.

With routine estrogen therapy contraindicated, the search continued for non-hormonal methods of preventing osteoporosis. Biphosphonates appeared to promote bone health without the side effects of estrogen. Again, large scale, long term studies were lacking, but it was preventive therapy; how could it cause any harm? Unfortunately, it has turned out that biphosphonates may not promote bone health, but may weaken bones. The biophosphonate Fosamax has already been linked to osteonecrosis (bone destruction) of the jaw, and now it appears that long term use of Fosamax may result in severe weakening of the femur bone (the thigh bone) leading to fractures that during activities as simple as walking.

Reversal of deficiencies associated with aging is not the only place where preventive medicine has gone wrong. Preventive medicine also rests on the premise that early diagnosis is better than late diagnosis, and that anything that increases the likelihood of early diagnosis must be beneficial. The most spectacular example of that faulty reasoning is the PSA (prostate specific antigen) test. Since increased PSA is associated with prostate cancer, doctors began recommending routine PSA screening, despite the fact that there were no large scale, long term studies demonstrating benefit. It was preventive medicine; how could it cause any harm?

Listen to what the test’s inventor, Dr. Richard Ablin, has to say about its use: “I never dreamed that my discovery four decades ago would lead to such a profit-driven public health disaster.”

According to Dr. Ablin:

… [T]he test is hardly more effective than a coin toss. As I’ve been trying to make clear for many years now, P.S.A. testing can’t detect prostate cancer and, more important, it can’t distinguish between the two types of prostate cancer — the one that will kill you and the one that won’t.

Instead, the test simply reveals how much of the prostate antigen a man has in his blood. Infections, over-the-counter drugs like ibuprofen, and benign swelling of the prostate can all elevate a man’s P.S.A. levels, but none of these factors signals cancer. Men with low readings might still harbor dangerous cancers, while those with high readings might be completely healthy.

Millions of men have been subjected to unnecessary biopsies, and harmful treatments, and billions of dollars have been wasted on this failed exercise in preventive care.

What can we learn from these and other similar debacles? We need to reexamine the basic premises of preventive medicine. Sure it is better to prevent disease than to treat it, but that does not mean that reversing the metabolic changes that accompany a disease will prevent it or will prevent it without causing serious unforeseen complications. Sure it is better to treat early stage cancer than late stage cancer, but a screening test that makes lots of mistakes can be worse than no screening test at all.

Most importantly, we must never forget that preventive medicine is a branch of medicine, in the exact same way that cardiology and neurology are branches of medicine. As such preventive medicine must be held to the same standards; any treatment, even a preventive treatment, must be tested in large scale, long term studies before being put into routine clinical use. Preventive medicine, like other branches of medicine, has the power to harm as well as to help. We ignore that fact at our own peril.

Amy Tuteur is an obstetrician-gynecologist who blogs at The Skeptical OB.

Submit a guest post and be heard.