Men in St. Paul with early-stage prostate cancer are nearly four times as likely to have surgery as men with the same condition 150 miles south in Mason City, Iowa.

In fact, patients' chances of having elective surgery for breast cancer, gallbladder trouble or enlarged prostate can vary as much as tenfold, depending on where they live, according to a new report by Dartmouth University researchers.

The report, released Wednesday, highlights a trend scholars have been studying for 20 years: how medical treatment differs dramatically from city to city and state to state.

The differences in treatment, said Shannon Brownlee, one of the authors, depend "not on how sick they are, not on who they are, but on where they live."

The report "really reveals some of the most troubling aspects of our health care system," said Dr. David Goodman, a co-author and one of the leaders of the Dartmouth Atlas Project, which tracks variation in health care. Often, he said, even physicians "have no inkling that the pattern of care they deliver can be quite different" elsewhere.

Prostate cancer surgery

St. Paul, for example, had the highest rate of prostate cancer surgery (3.1 per 1,000 Medicare enrollees) in the seven-state "Great Plains" region, compared with 0.7 in Mason City. Rochester, Minn., had the region's lowest rate of a procedure called carotid endarterectomy, used to prevent stroke: 1.2 per 1,000 Medicare enrollees, compared with 3.3 in Wichita, Kans.

In general, Minnesota had significantly higher rates of knee replacement and hip replacement surgery than the national average, the report found. Nationwide, the lowest rates were found in Hawaii. The Dartmouth study is based on Medicare data on millions of Americans ages 65 and older from 2008 to 2010.

Brownlee said researchers picked surgeries considered elective -- meaning "there's no one right treatment for that condition." That included procedures for breast cancer, prostate cancer, back pain, chest pain, artery disease, knee and hip arthritis and gallstones.

Physician preference

In many cases, the approaches to treatment can depend on where doctors were trained, as well as differences of opinion about risks and benefits, the authors said.

But they say the regional patterns tend to override the preferences of patients, who often lean toward less aggressive treatment.

"Their preferences are often not taken into account," Brownlee said. "It's often the physician's preference, the physician's opinion, that ends up winning the day."

Goodman said it isn't always clear whether high or low rates of a given procedure are better for patients. "Part of this is because the evidence is not always perfect," he said. Often, he said, it's a case of "well-meaning doctors assuming they know what's best for patients."

Involving patients

Instead, he said, doctors should do more to involve patients in those decisions.

Brownlee said women with early-stage breast cancer, for example, "aren't always aware of the fact that they have a choice, and they aren't always informed about what the trade-offs are." A mastectomy can be more disfiguring, she noted, while a lumpectomy can involve radiation treatment and a higher risk of recurrence.

Giving patients a stronger voice, Goodman said, would not eliminate all variation in the rates. But it would "reflect patient preferences," he said, not their doctors'.

Maura Lerner • 612-673-7384