Physicians Choose Different Treatments for Themselves

Jim Kling

April 11, 2011

April 11, 2011 — Physicians don't necessarily take their own advice. Faced with 2 alternatives, one with a reduced chance of death but at a greater risk for severe adverse effects, they were more likely to recommend patients take the option that reduced chance of death than they were to choose it for themselves. The study was published in the April 11 issue of Archives of Internal Medicine.

According to lead study author Peter Ubel, MD, Fuqua School of Business and Sanford School of Public Policy, Duke University, Durham, North Carolina, and colleagues, patients faced with difficult choices may ask physicians for a clinical recommendation. Studies show that patients can be swayed by such recommendations, even when it's not in their best interests.

Some have argued that physicians should only present options, not make recommendations, because patients are the best judge of what's best for themselves. Others suggest that physicians can promote the patients' best interests, at least in some cases, such as when patients' cognitive bias may lead them to choose against their own interests.

Making recommendations for others can also change one's own perspective, which could in turn reduce bias. Some studies have shown that people who make decisions for others are able to ignore extraneous factors and focus on the most important aspect of the decision.

To find out whether a physician's recommendation alters his or her own perspective on medical decisions, the researchers designed 2 randomized experiments that compared physician recommendations to a patient with what choice they would make for themselves. Both tests were designed to reveal well-established cognitive biases.

Participants were US primary care physicians comprising general internists and family medicine specialists drawn from the American Medical Association Physician Masterfile. Each participant responded to 1 of 2 clinical scenarios, each involving a choice between a treatment that improved the odds of surviving a fatal disease but at a higher chance of unpleasant complications.

Participants were randomized so that they were instructed to respond as if the choice were for themselves or as if they were making a recommendation to a patient.

Choice of Treatment for Colon Cancer Scenario

The first scenario involved colon cancer. Choice 1 involved a surgery that cured 80% of patients with no complications and failed to provide a cure in 16%, leading to death within 2 years; 4% would be cured but would need to undergo a colostomy or experience chronic diarrhea, intermittent bowel obstruction, or wound infection (1% chance of each scenario). Option 2 was a surgery that would cure 80% of patients without adverse effects but would fail to provide a cure in 20% of patients. Of 242 respondents, 37.8% chose option 2 for themselves, compared with 24.5%, who recommended it to a hypothetical patient ( χ2 = 4.67, P = .03).

Avian Influenza Scenario

In the second study, a different sample of 1600 primary care physicians were selected to consider a scenario in which a new strain of avian influenza had recently come to the United States. Without treatment, patients would have a 10% death rate and a 30% hospitalization rate. An immunoglobulin treatment would be available that reduces death rate from influenza to 5% and hospitalization rate to 15% but has a 1% risk for death and a 4% risk for neurologic paralysis (usually lower extremities).

Among the 698 participants who responded, 62.9% chose for themselves the option with the higher death rate, compared with 48.5% who recommended it for a hypothetical patient (χ2 =14.56, P < .001).

Physician Bias in Patient Recommendations

In an accompanying editorial, Eric Shaban, MD, Palliative Care Program, Department of Medicine, University of Rochester Medical Center, New York, and colleagues suggested that physicians who would confer a different recommendation for themselves should think about their own potential cognitive biases. “Exploring and trying to understand the root cause of such biases before giving recommendations can help physicians and patients make decisions that best reflect each patient's values regarding his or her medical condition,” they write.

Study authors received grant funding from the National Institutes of Health as well as the American Cancer Society career award, the National Science Foundation Graduate Research Fellowship, and the Robert Wood Johnson Health Policy Investigator Award. The editorialists have disclosed no relevant financial relationships.

Arch Intern Med. 2011;7:630-634.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: