MONDAY, April 11 (HealthDay News) -- When faced with a choice of treatments, primary care doctors often choose a different option than they would recommend to their patients, a new study finds.

In fact, physicians who were surveyed in the research more often chose a treatment in which they were more likely to have long-term repercussions, such as paralysis, than to die.

But when it came to advising their patients, doctors were more likely to urge the treatment with a greater chance of death vs. the one with non-lethal adverse effects, say researchers publishing their findings in the April 11 issue of the Archives of Internal Medicine.

"It's a pretty clear message: There's a discrepancy between what doctors recommend to their patients and to themselves," said Dr. Timothy Quill, author of an accompanying editorial in the journal.

"We need to try to reconcile that and make sure decisions are guided by patient values. We need to know what patients' values are, particularly in value-laden decisions," added Quill, who is director of the Center for Ethics, Humanities and Palliative Care at the University of Rochester Medical Center. "Decisions need to be guided by patients' values, not doctors'."

The study's lead author agreed. "Doctors and patients both need to just be aware that advice giving is not as neutral as they think," said Dr. Peter A. Ubel, a physician and behavioral scientist at Duke University in Durham, N.C. "This is not just about experts handing down expertise. It fundamentally changes the way people weigh risks and benefits. A better approach would be for both of the groups, doctors and patients, to have a better discussion of what matters to the patient."

In the study, Ubel and his colleagues presented one of two different treatment-decision scenarios to primary care physicians in the United States.

For the first scenario, 242 doctors were told that either they or a patient had just been diagnosed with colon cancer and could choose one of two surgeries. Both options had a cure rate of 80 percent, but one had a higher death rate and fewer side effects. The other had a lower death rate but patients were more likely to need a colostomy or to have chronic diarrhea, intermittent bowel obstruction or wound infection.

When it came to their own bodies and lives, 37.8 percent of physicians chose the option with higher mortality over fewer side effects, but only 24.5 percent thought that patients should choose this course.

The second scenario involved being infected with a fictitious new strain of avian flu. The flu itself had a 10 percent death rate and would send 30 percent of patients to the hospital for an average of one week.

The one treatment would halve the rate of adverse events, but would kill 1 percent of recipients and cause permanent paralysis in 4 percent.

Of the nearly 700 doctors completing this survey, 62.9 percent decided that they personally would endure the flu rather than get this treatment. For their patients, however, only about half (48.5 percent) recommended foregoing treatment in the event of infection.

"People think that doctors should not be telling people what to do, but [just] laying out the risks and benefits," Ubel said. "There isn't a lot of research on what happens when people put on these different hats -- decision-maker vs. advice-giver. There's a concern that doctors will often give bad advice because they're biased by their own financial interests or specialty or style. A surgeon thinks you need surgery, radiation oncologists says radiation."

Ubel says that feelings tend to pull people in one direction and thinking in the other. And doctors may put thinking first when it comes to their patients.

There's a lot more decision-making in health than sometimes thought, added Ubel. "Even taking a cholesterol pill can be a judgment call. There are potential side effects," he said, and "the benefits are tiny if you're at low risk for a heart attack."

More information

There's more on the doctor-patient relationship at the American Medical Association.