LOCAL KNOWLEDGE.(The Talk of the Town)(geography still vital when it comes to medical care)

Everyone seems to agree that the American health-care system is broken, yet no one can agree on how to fix it. Part of the problem is that there is no American health-care system. Instead, there are hundreds of local health-care systems, each with its own way of treating patients and spending money. Where you live has a profound effect on how much attention your doctor gives you, how many days you spend in the hospital, and what drugs you are (or are not) prescribed. For all the talk about the death of distance and the homogenization of the American landscape, when it comes to health care geography is still destiny.

The recognition of this fact has humble roots. In the nineteen-seventies, a doctor named John Wennberg conducted a study in his home state of Vermont and found that, even in this small and relatively homogeneous corner of the country, doctors in different areas adopted wildly different approaches to the treatment of tonsillitis. In one town, seventy per cent of children had had their tonsils removed by the time they were twelve. In the town next door, it was only twenty per cent. Statistical fluke? Apparently not. Over the next three decades, Wennberg and other researchers documented similarly dramatic geographical variations in far more serious cases--cesarean sections, spinal fusions, mastectomies, coronary-bypass surgeries.

No two patients or groups of patients are exactly alike, so people have always been skeptical of these geographical comparisons. But at this point the evidence is hard to refute. A recent set of Dartmouth Medical School studies, led by Wennberg, looked at the way top teaching hospitals treated elderly patients who were in the last six months of their lives. Patients at Mount Sinai, in New York, spent nearly twice as many days in the hospital as patients at the Mayo Clinic, in Rochester, Minnesota. Patients at U.C.L.A. spent three times as many days in the intensive-care unit as patients at Mass General. At the "high-intensity hospitals," patients saw doctors, consulted with specialists, and were given tests far more often than at low-intensity ones. What's more, the extra attention didn't allow patients to live any longer. The weeks in the hospital, the batteries of tests, the regular consultations: they did little but drive up the cost of treatment.

The problem, in other words, isn't so much that different doctors treat patients differently. It's that these differences have a major impact on the cost of health care. No one is opposed to spending more if we could get better results by doing so. But variations in treatment often mean that we're doing the opposite. Medicare, for instance, pays twice as much per patient in Miami as it does in Minneapolis. But, again, Medicare patients in Miami don't live any longer than those in Minneapolis. At a time when everyone's worried about controlling health-care ...