(CNN) If you are overweight, have emotional problems or have difficulty with English, there's a good chance your doctor could be judging you because of it.
That troubling finding comes from a survey conducted over email of 15,800 physicians across the United States from more than 25 specialties, including emergency and family medicine and psychiatry. The survey was part of an annual report on bias and burnout among doctors that Medscape, a news and education resource for health care professionals, carries out every year.
The survey found that 40% of physicians reported having biases toward certain groups of patients. For a handful of specialties, closer to half of doctors said they harbored biases, including 62% of emergency medicine doctors, 50% of orthopedists, 48% of psychiatrists and 47% of family medicine doctors and OB-GYNs.
Biases were least common among radiologists, cardiologists and pathologists, of whom 22%, 22% and 10% respectively reported these feelings.
The most common reason that doctors said they stereotyped patients was because of their emotional problems, which elicited biases among 62% of physicians, followed by their weight, which 56% of male and 48% of female physicians said provoked biases for them. Other triggers were patients' intelligence, language differences, insurance coverage, age, income level, race and attractiveness.
"It's troubling, but not altogether unexpected," said Dr. Joseph R. Betancourt, director of the Disparities Solutions Center at Massachusetts General Hospital and Harvard Medical School. Betancourt was not involved in the survey.
A growing amount of research shows that doctors hold prejudices and that these prejudices can influence the kind of care that patients get. The Institute of Medicine and the National Academy of Sciences urged the need for more research on the issue in a 2003 report.
"I think 40% is probably an underestimation, because there is a component of subconscious bias," said Betancourt, who was one of the authors of the Institute of Medicine report. The current survey did not include tests, such as the implicit association test developed at Harvard, to determine whether doctors held biases that they were not even aware of.
"Human beings in general have these (subconscious) stereotypes, so it's hard to believe it couldn't be 80% or 90%," Betancourt said.
Betancourt also suspects that doctors in the survey might have been reluctant to admit prejudices that were less socially acceptable, such as biases toward racial groups. Only 16% of male and 14% of female physicians reported feeling biases toward patients based on race, but that is probably "exceedingly low compared to what the reality is," Betancourt said.
In the face of these findings, we could at least take some consolation in the fact that fewer than 15% of doctors, regardless of specialty, said that their biases affected the treatment they gave patients. But we probably shouldn't rest too easy.
"That (rate) is completely erroneous in my opinion, and built on the 'not me' phenomenon. Doctors say, 'I don't do it, not me,'" Betancourt said. "Doctors have been trained to believe that we make decisions based on symptoms and science and that personal characteristics don't impact us," he added.
The preponderance of evidence suggests that doctors do give different care to different groups of patients. For example, many clinicians have negative views of people with obesity and this can affect how they communicate with and treat these patients.
Doctors who have biases are not going to say that they affect the care they give, because "that is so antithetical to what they do as physicians," said Louis Penner, professor of oncology at Wayne State University. Nevertheless the care they give overweight patients is "potentially not as good," Penner said.
When it comes to race, research suggests that many physicians have implicit or subconscious biases, even if they do not have explicit biases that they are aware of. "The insidious thing about implicit racial bias is that it occurs among physicians who believe they're not racist, but patients pick up on it and (it) affects how they feel about the treatment," said Penner, who has conducted research on the effect of implicit biases on health care.
One recent study found that black children were less likely than white children to be given medication to manage their pain from appendicitis.
The types of doctors that the survey found are most susceptible to biases also does not come as a surprise to Betancourt. "In emergency medicine, (doctors) have to make quick decisions with a limited amount of information; it is high risk and high intensity. Those are the hallmarks of stereotype activation," Betancourt said.
Even doctors who do get to know their patients, such as family medicine doctors and OB-GYNs, are still susceptible to biases, though, and it could be getting worse because of the growing amount of stress they feel. Among the doctors that reported burnout, most commonly due to factors such as long work hours and more computerization of work, 43% reported having biases, compared with 36% of those who said they were not burnt out.
The good news is that some progress is being made, at least among the newer generation of physicians, Betancourt said. In the last 10 years, many medical schools, including at Harvard, University of California-San Francisco and Johns Hopkins University, have started teaching students about stereotypes, how to be aware of them and try to prevent them, he said.
There has also been growing demand for online modules to train health care professionals about biases, said Betancourt, who is co-founder of a company called Quality Interactions that creates these modules.
"My concern is there will be a generational effect here, where individuals that lived in a certain time and are used to certain behaviors and the old way of doing things have stereotypes that will be hard to break," Betancourt said.