Physician, Inform Thyself

By Melissa Ludtke

Two 50-year-old men complaining of chest pains and shortness of breath head for their doctors' offices. In both cases, angiograms show that the patients are suffering from partly blocked arteries. But at this point the medical paths of these men, with identical symptoms but different doctors, may diverge radically. One man lives in Beverly Hills, and the chances that he will have coronary-bypass surgery are nearly twice as high as they are for the other man, who lives in Pasadena, just 20 miles away. The Pasadena patient is more likely to be treated with drugs and a modified diet.

Is it possible that where people live can determine what medical treatment they receive? Surprisingly, the answer is yes. "There is an underlying assumption that two doctors in two different places will prescribe the same treatment," says Dr. Phil Caper, who founded the Codman Research Group in Lyme, N.H., to study variations in the patterns of physician care. "That just isn't so."

Treatment patterns can vary among communities because doctors in different places have different methods. Within a given hospital, doctors tend to consult one another and reach a consensus on how to practice, but that consensus may not be the same in another city. In some areas, for example, the frequency of hysterectomies is three times as high as in other places. As discoveries like these accumulate, statistical evidence begins to raise doubts about the scientific certainty usually associated with medicine.

A major reason that medical practices vary so widely is that doctors suffer from a shortage of certain essential information. Despite the proliferation of medical reports and journals, there are few statistics on the comparative results of clinical procedures. And there is no comprehensive national collection of data concerning what treatments work best for what kind of patients. In fact, relatively little systematic research has been done on the "outcomes" of patients' treatment -- whether they get better or worse, live or die.

Steps are being taken to fill medicine's information void. In a new field of study called patient-outcomes research, hospitals, clinics, health-maintenance organizations and other medical groups are collecting data on how well various treatments work. Armed with such knowledge, doctors should be able to get better results. Dr. Paul Ellwood, chairman of the InterStudy health-policy center near Minneapolis, predicts that within a year at least 100 patient- outcomes projects will be under way, with sponsors as diverse as the Cleveland Clinic and the Maine Medical Assessment Foundation. High on the list of treatments to be studied are those for cataracts, diabetes and broken hips (the question: When is replacing the hip the best thing to do?). A report in the New England Journal of Medicine suggested that one type of prostate surgery works better than an increasingly popular alternative operation. The American Urological Association is planning an intensive comparative study of the long-term prospects of patients who undergo one of the two procedures.

Lack of information about patient outcomes has both physical and financial consequences. Not only do some patients endure unnecessary surgery, but health-care costs in the U.S. continue to increase faster than the gross national product. Observes Dr. David Eddy, professor of health policy and management at Duke University: "Current medical logic tells doctors, 'When in doubt, do it.' " One such procedure is the carotid endarterectomy, performed to remove a clot from a neck artery. Until recently no one, including doctors who perform the operation, knew how clinically appropriate this surgery was. A joint study by the University of California, Los Angeles, and the Rand Corporation concluded that just one-third of the 1,302 operations surveyed were beneficial; in fact, 6.4% of the patients later had strokes, which the surgery was supposed to help them avoid. Rand recommended that the $46,900 operation be done less frequently.

Advocates caution that outcomes research by itself is not an antidote to rising medical costs. "It shouldn't be sold as a cost-containment measure at all," says Dr. Jack Wennberg, a professor of epidemiology at Dartmouth Medical School and a pioneer in the research. "It is a scientific measure." Even the best outcomes data will never address much more fundamental questions, such as which patients should have access to heart transplants or other ultraexpensive procedures. Those dilemmas are still left to medical ethicists and society to resolve.

To some doctors the new studies may seem threatening. "Outcomes research will demonstrate that a large percentage of what we do doesn't make any difference," says Dr. Robert Brook, who oversees Rand's outcomes studies. Many doctors fear that the research will handcuff them with a "cookbook mentality" -- a dash of this, a pinch of that, and the result is known. But, says Brook, "good cooks start from a cookbook. Then they modify the recipe. In a very complex world, we shouldn't back away from starting with a certain protocol."

Most physicians, haunted as they are by the specter of malpractice suits, will probably appreciate having this information. Asserts Dr. Caper: "We say ) to doctors, 'Here's a tool that allows you to control your own destiny.' " Patients too will be able to make more informed decisions about treatments. Used wisely, patient-outcomes research will undoubtedly prove a boon to the entire practice of medicine.