One-third of physicians won't report impaired or incompetent colleagues, according to a survey, even though it's generally required by medical ethics codes and some state laws.

A 2009 survey of nearly 2,000 physicians in a wide range of specialties found only about 64 percent fully believed they had a professional obligation to tell authorities if a colleague were unable to do his or her job.

Furthermore, 17 percent had been aware of an incompetent or impaired colleague over the last 3 years, but one-third chose not to report him or her.

These findings are "troubling" because doctors checking up on each other is the first line of defense for patient safety, the authors wrote in their report, published Wednesday in the Journal of the American Medical Association.

"Self-regulation is the primary mechanism that patients have as protection against physicians who should not be practicing because their judgment is somewhat impaired," Dr. Catherine M. DesRoches, of the Mongan Institute for Health Policy at Massachusetts General Hospital and lead author of the study, told DOTmed News.

The American Medical Association Code of Ethics, the Charter on Medical Professionalism and many states require doctors to report on colleagues they suspect of being incompetent or impaired, especially through drug or alcohol abuse.

In the study, the researchers analyzed mail-in questionnaires answered by 1,891 respondents, from pediatrics, family practice, surgery, anesthesiology, psychiatry, internal medicine and cardiology.

The results found that about 64 percent of physicians said they "completely" agree they should report colleagues unable to do their job properly, with the rest saying they somewhat or don't agree.

And though similar numbers said they were prepared or able to report on colleagues, of the 17 percent who had known someone unfit for the job, only 67 percent reported on the person.

The chief reasons for not reporting were thinking someone else was already taking care of it (19 percent), not believing reporting would make a difference (15 percent) and fear of retaliation (12 percent).

Graduates of U.S. medical schools, whites and Asians and doctors working at universities, medical schools, hospitals and clinics were the most likely to say they completely agreed with reporting. Also, doctors operating in areas with low numbers of malpractice claims were somewhat more likely to say they should report on their colleagues (68 percent) compared with medium and high claim areas (60 to 63 percent), "But [the difference is] pretty marginal," DesRoches said.

Doctors with foreign medical degrees, underrepresented minorities and those who plied their trade in solo or two-person practices were somewhat less likely both to report and to say they completely agreed with reporting.

The authors suggested a variety of reasons to explain differences in willingness to report. Doctors with foreign degrees could, because of their outside status, be more fearful of retaliation, DesRoches said. And physicians in small practices depend on referrals, and might not want to risk their reputations. They could also be more isolated, and have a harder time keeping their reports anonymous.

"If you're in a small practice in a small town, it could be very easy to figure out who reported," DesRoches said.

To boost willingness to report, DesRoches said it's important for medical groups to get the message across to doctors that even if they believe someone else has reported on a problem doctor, it's still their responsibility to do so. Also, medical regulating bodies can set up confidential feedback mechanisms to let doctors know that their reports were followed up on and not just lost in a bureaucratic black hole.

"And then finally, ensuring and really beefing up the confidentiality of these systems," DesRoches said, "so physicians who do report don't have to be afraid of damage to their reputations or damage to their referral network." This could be accomplished by, for instance, educating patients, nurses and others to also report.

An editorial accompanying the article, while generally agreeing with it, disputed some of the interpretations of the data. Physicians in small practices, for example, might not report on colleagues because they're friends or don't have a clear authority to report to. Also, failing to report on a doctor because someone else is taking care of it could be a "reasonable" interpretation of the situation.

"Still, it is impossible to escape the implication that these survey results constitute a frontal assault on a basic premise of medical professionalism," writes Dr. Matthew K. Wynia, director of the Institute of Ethics at the American Medical Association. "Despite any minor flaws, this research is proof that individual physicians cannot always be relied on to report colleagues who threaten quality of care."