Much research has gone into studying patient race and ethnicity and their effect on the care received. The recent Medscape Physician Lifestyle Survey asked physicians to identify their race and/or ethnicity in order to explore associations with patient care, behavior, and levels of happiness. The survey again posed questions from previous years about burnout, bias, and other lifestyle factors. More than 14,000 physicians from over 30 specialties responded and provided some surprising results.
Note: Values in charts have been rounded and may not match the sums described in the captions.
The Medscape survey once again asked about burnout among US physicians, which is defined in this and other major studies as a loss of enthusiasm for work, feelings of cynicism, and a low sense of personal accomplishment. Burnout rates for all physician respondents have been trending up since 2013, the first year that Medscape asked about it, when the overall rate was 40%. This year it is 51%, over a 25% increase in just 4 years. A recent major survey supports these findings, reporting that burnout and satisfaction with work-life balance had worsened between 2011 and 2014, with more than half of physicians reporting burnout.
In this year's Medscape report, the highest percentages of burnout occurred among physicians practicing emergency medicine (59%), followed by ob/gyns (56%) and family physicians, internists, and infectious disease physicians (all at 55%). The top four are all physicians who deal directly with patients with a range of complex problems. In Medscape's 2015 and 2016 reports, emergency medicine physicians, family physicians, and internists were also within the top five. Not surprisingly, other research has found high burnout rates in these three professions.[1,3-5]
Respondents were asked to rate the severity of their burnout on a scale of 1 to 7, where 1 equals "It does not interfere with my life" and 7 equals "It is so severe that I am thinking of leaving medicine altogether." Of note, emergency and primary care physicians—groups that are consistently at the top in reporting burnout—did not have the highest rates of burnout severity. Among physicians reporting burnout, urologists had the highest average severity rating (4.6), followed by otolaryngology and oncology (both at 4.5). Surprisingly, infectious disease physicians, who were within the top five for experiencing burnout, had the lowest severity rating (3.9).
Physicians were asked to rate causes of their burnout on a scale of 1 to 7, where 1 equals "Does not contribute at all" and 7 equals "Significantly contributes." Topping the list, with an average rating of 5.3, was "too many bureaucratic tasks," followed by "spending too many hours at work," at 4.7. These two factors ranked highest the past two surveys as well. And the ratings have been trending up: In the previous report, they were 4.8 and 4.1, respectively, and in the one before that, they were 4.7 and 4.0.
"Feeling like just a cog in a wheel" and electronic health records (EHRs) also rated high as causes of burnout this year (4.6 and 4.5, respectively). In an October 2016 Medscape roundtable discussing EHRs and burnout, one of the panelists, Robert W. Brenner, MD, said, "If [EHR requirements are] implemented without a change in the workflow in the office, too much data entry falls on the physician. That is what is adding to the huge burden."
Because "insurance issues," "threat of malpractice," and "family stress" were mentioned frequently as important contributions to burnout in last year's write-in responses to this question, the options were added to our survey. They rated 4, 3.9, and 3.1, respectively.
In this year's report, as in prior years', burnout was reported at a higher percentage by female physicians (55%) compared with their male peers (45%). Percentages have trended up for both men and women since this question was first asked by Medscape in 2013. That year, 45% of women and 37% of men reported burnout. However, burnout appears to be leveling off in both women and men.
While burnout rates were highest this year in the Northwest (54%), Southwest (53%), and South Central (52%) regions and lowest in the West (49%) and North Central and Southeast (50%) regions, the small degree of variance in incidence underscores that this is a national phenomenon. There seems to be little correlation between burnout and regional compensation, which, according to the 2016 Medscape Physician Compensation Report, was highest in the North Central region and lowest in the Northeast.
Using US Census Bureau criteria, we asked physicians about their race/ethnicity. Because such classifications are not necessarily straightforward, respondents could choose more than one option, and 5% did. That being considered, the majority (69%) identified as white/Caucasian. Following in prevalence were those reporting Asian ethnicities (17%), which includes 8% of all respondents identifying as Indian and 4% as Chinese. Five percent of responding physicians reported as Hispanic/Latino and 4% as black/African American.
A 2015 report based on data from the Association of American Medical Colleges on medical school graduates' racial self-identification, published by the Kaiser Family Foundation, found 7% of respondents choosing multiple races, 58.8% noting white/Caucasian, 19.8% Asian, 5.7% black/African American, and 4.6% Hispanic/Latino. Such findings might suggest a decline in white/Caucasian physicians and a slight increase in black/African American physicians.
In this year's Medscape report, black/African American respondents included the highest percentage of female physicians (63%). The lowest, 38%, occurred among whites/Caucasians.
This finding mirrored a survey by the American Association of Medical Colleges, which concluded that among physicians of all ages, female black/African American physicians were the only non-white group in which there is currently a higher percentage of women than men. Among physicians age 29 and younger, however, there are also more female than male Asians and Hispanics/Latinos. Furthermore, the survey concluded that among black/African American medical school applicants, fully two thirds are female.
This year's report shows some relationship between race/ethnicity and burnout, with the highest rates of burnout in physicians who self-identify as Chinese (56%) and other Asian (53%). Vietnamese and white/Caucasian ethnic groups tied for third (52%). The lowest percentages occurred among respondents who defined themselves as Asian Indian (46%), Japanese (47%), and black/African American (48%).
A 2007 study of medical students found that 47% experienced burnout, and the rate was lower among non-whites than whites. Notably, however, non-white students who had experienced adverse treatment because of their race had a higher burnout rate than their non-white peers. (There was no difference, however, in the percentages of those who reported that they were depressed.) A study on race and work-related stress found that, in general, non-white physicians faced a more demanding patient base than did their white peers. Nevertheless, Hispanic/Latino and black/African American physicians reported no difference in stress versus their white peers, although Asians and Pacific Islanders reported higher average stress than white physicians.
In this year's Medscape survey, when physicians were asked if they believed that they had biases toward specific types or groups of patients, 50% said they did. Physicians who identified themselves as Korean or Vietnamese most frequently reported that they do (63% and 60%, respectively). The lowest percentages occurred among other Asian Indians and other Asians (34% and 41%, respectively).
One limitation of any survey involving race and ethnicity is implicit bias, which occurs without conscious awareness. Implicit bias is frequently at odds with one's reported feelings and beliefs, and it is more likely that attitudes toward whites will be more positive while negative biases are more often found directed towards non-whites.[12,13]
Among respondents who admitted biases, more than half (51%) of male physicians, compared with 42% of female physicians, cited overweight as a patient factor that elicited bias. Emotional problems in patients were cited most frequently by female respondents (51%), nearly matched by their male counterparts (50%). Perceived low intelligence ranked third for both genders (40% of men and 38% of women), followed by language differences (28% of men and 26% of women). Nearly a quarter of men (21%) but a relatively smaller percentage of women (15%) reported bias toward patients who lack insurance. Bias toward patients of a different race from the respondent's own was reported by 10% of male and 9% of female physicians, and gender bias by only 4% of female and 5% of male physicians.
When asked if their bias affects treatment, 14% of Korean and 12% of both Chinese and Filipino physicians acknowledged that it did. Respondents least likely to report that their treatment is affected by bias described themselves as Asian Indian (5%) or white/Caucasian, black/African American, and Hispanic/Latino (all at 8%). Some research suggests that implicit bias may play a role in treatment. In one study, although implicit bias did not have an effect on treatment recommendations, physicians were more likely to find white patients "cooperative" than black patients and said that this bias could have influenced their decisions.
Only 16% of all physicians who admitted bias said that it affected their treatment. Our survey asked that group whether the effect of their bias was positive (eg, extra time spent, friendlier manner) or negative (eg, less time spent, less friendly manner), and respondents could answer "yes" to both questions. The highest percentages of physicians who admitted to negative effects on treatment of patients as a result of bias cited language differences (61%) and emotional problems (58%). Half acknowledged a negative disposition toward those who are overweight, 49% toward those who are perceived to have low intelligence, and 45% toward those who lack insurance. The only bias that half of respondents said leads to positive treatment is older age. Of note, one analysis of studies on the relationship between patients' race/ethnicity and care found that there was greater overuse of care among white patients though the reasons for this are unclear.
When bias was examined by age group, physicians under 45 years of age tended to admit to more biases than their older peers. Of particular note, more younger physicians admitted bias toward patients with emotional problems and perceived low intelligence compared with older peers (54% and 46%, and 42% and 34%, respectively) Bias toward overweight patients, in contrast, was reported at the same rate (47%) in both groups.
Respondents were asked if they are socially conservative or liberal. The most liberal physicians identified themselves as Japanese (69%) or black/African American (68%). Filipino (46%) or "other Asian" (49%) physicians were the least likely to identify as liberal. About 60% of all of the other groups reported being socially liberal.
This year's lifestyle survey, like previous years', asked whether physicians are happy at work and outside of work. Respondents were asked to rate their happiness on a scale of 1 to 7, with 1 being "extremely unhappy" and 7 being "extremely happy." Among those who said they are very to extremely happy, there was little difference between male and female physicians' happiness levels outside of work (69% and 67%, respectively). When looking at happiness at work, however, those percentages were lower, with men being happier (45%) than women (39%).
In this year's report, the highest percentages of physicians who said they are either very or extremely happy at work were seen among dermatologists (43%) and ophthalmologists (42%). Dermatologists and ophthalmologists were also the top two happiest at work in the 2016 and 2014 reports. In 2016, however, the percentages for these two specialties were lower (39% and 38%), respectively. At 76%, urologists were at the top of the list in reporting being happy outside of work, though ophthalmologists and dermatologists also placed high, tied for second place at 74%.
At the bottom of the list, rheumatologists and nephrologists were the least happy, both at work (both 24%) and outside of work (61% and 62%, respectively).
Among this year's respondents, about two thirds of those who identified themselves as Japanese (69%), Filipino (65%), or Vietnamese (64%) reported being very to extremely happy outside of work. The lowest percentages occurred among those who described themselves as other Asian (50%) or Chinese or black/African American (both 55%).
The highest percentages of physicians reporting happiness at work this year occurred among those who identified themselves as Asian Indian (37%) or Hispanic/Latino (36%), but the percentage range for nearly all groups from top to bottom was only 13%. At 24%, only those respondents who described themselves as Chinese were lower than 30%, but all of the percentages were discouragingly low.
A 2007 study on the relationship between physician race and stress found that Hispanic/Latino—and, to a lesser degree, black/African American—physicians reported higher work satisfaction levels than their white peers. The study showed significantly lower job satisfaction among Asian physicians than white physicians, but unlike Medscape's survey, it did not break down Asian groups.
Race/ethnicity does not appear to have any strong relationship to frequency of exercise among physicians. The respondents most likely to exercise at least twice a week identified themselves as white/Caucasian (69%) or Japanese (67%). The lowest percentages were found among respondents who described themselves as other Asian (58%), and Korean or black/African American (both at 59%).
Race does appear to have a relationship to physicians' weight. The highest percentages of respondents who said they are overweight or obese were among self-identified black/African American (52%), Hispanic/Latino (49%), and white/Caucasian (44%) physicians. These numbers echo the findings of a 2016 study of all American adults by the Robert Wood Johnson Foundation, although it should be noted that the study did not break out rates among Asian Americans. In the Medscape survey, self-described Asian respondents were the least likely to report that they are overweight or obese, with the lowest percentages seen in Vietnamese (21%) and Chinese (28%) physicians.
In general, reported alcohol use among physicians is low. Respondents who most often indicated that they drink one or more alcoholic beverages a day identified themselves as white/Caucasian (22%). Japanese and Korean physicians followed at 17% and 15%, respectively. Physicians who described themselves as other Asian or Filipino were least likely to report alcohol use, at 8%.
In spite of the ongoing income disparity between male and female physicians, in this year's report, there was almost no difference between genders in the amount of reported debt. Sixty-two percent of all respondents said their debt is manageable, and only 7% of men and 9% of women characterized it as unmanageable. More than a quarter of both men and women (29% and 27%, respectively) said they have no debt at all.
In this year's report, a slightly higher percentage of male versus female physicians (57% and 53%, respectively) believe their income and assets are sufficient to support their goals. Conversely, 12% of men and 17% of women have no hope that they will ever be sufficient. Thirty percent of both men and women say that their income and assets aren't enough right now, but that they expect them to improve.
The highest percentages of respondents who believe that their finances are sufficient to support life goals occurred among those who described themselves as Japanese (72%). They were followed by white/Caucasian (59%), Filipino (57%), and Chinese (54%). Only 43% of self-identified Korean and 46% of black/African American respondents reported sufficient income and assets, although a relatively high percentage of the latter group (42%) believed the situation would improve.