Thursday, March 18, 2010

Eminence-based medicine and prostate cancer treatments

Until the arrival of evidence-based medicine (EBM) movement in the 1980s, medical training and continuing medical education in the U.S. were guided by what might be called the other EBM, "eminence-based medicine" - that is, physicians made decisions about patients based largely on what their attending physicians taught them to do in medical school, what therapies respected colleagues told them to do, and what review articles written by those gray-haired eminences instructed them to do, regardless of what actually worked. By the time I entered training in the late 1990s, a few things had changed for the better, and useless or harmful therapies such as estrogen for menopause, high-dose chemotherapy with bone marrow transplanation for breast cancer, and drugs to suppress cardiac arrhythmias were exposed for what they were in randomized clinical trials.

But sadly, much of medicine continues to operate in an evidence-based vacuum, and nowhere is this state of affairs more painfully evident than with treatments for clinically localized prostate cancer (prostate cancer that is detected via a blood test and confirmed by microscopic examination of a biopsy sample, but causes no symptoms in the patient). Not only is there little reliable data to support choosing one type of therapy over another, for many men the option of deferring therapy (called "watchful waiting" or "expectant management") may be superior to all of them.

So how do men decide what therapy to choose for this condition? They consult physicians, which would be a sensible thing to do if different types of physicians actually agreed what therapy to recommend in specific situations. Unfortunately, a recent study published in the Archives of Internal Medicine confirms what I had already suspected from my own experience: eminence-based medicine is alive and well. Rather than rationally selecting therapies based on their personal health status, tolerance of uncertainty, and preferences for certain side effects over others, patients with prostate cancer overwhelmingly choose therapies based on the type of physician they happen to see. Men who visit a urologist between the time of cancer diagnosis and treatment are most likely to choose surgery, men who visit a radiation oncologist are most likely to choose prostate irradiation. Meanwhile, men who also visited their primary care physician were most likely to defer therapy regardless of what specialist they saw, but only 1 in 5 men in the study actually had a primary care consultation.

While the health care bill pending in the House of Representatives will not solve this problem, it does contain some essential elements: support for primary care, research on the effectiveness of common therapies, and support for unbiased organizations that evaluate this research to decide what is or is not a worthwhile use of limited health care resources. These are the necessary first steps to ending the era of eminence-based medicine.