MILWAUKEE – More than 40% of white physician trainees demonstrated racial bias in medical decision making about treatment of low back pain, as did 31% of nonwhite trainees. However, just 6% of white residents and fellows, and 10% of the nonwhite residents and fellows, reported that patient race had factored into their treatment decisions in a virtual patient task.
The 444 medical residents and fellows who participated viewed video vignettes presenting 12 virtual patients who presented with low back pain, wrote Alexis Grant of Indiana University–Purdue University Indianapolis and her colleagues. In a poster presentation at the scientific meeting of the American Pain Society, Ms. Grant, a doctoral student in clinical psychology, and her collaborators explained that participants agreed to view a series of 12 videos of virtual patients.
The videos presented male and female virtual patients who were black or white and who had jobs associated with low or high socioeconomic status (SES). Information in text vignettes accompanying the videos included occupation, pain etiology, physical exam findings, and pain intensity by self-report.
After viewing the videos and reading the vignettes, participating clinicians were asked to use a 0-100 visual analog scale to report their likelihood of referring patients to a pain specialist or to physical therapy and of recommending opioid or nonopioid analgesia.
“Next, they rated the degree to which they considered different sources of patient information when making treatment decision,” Ms. Grant and her coauthors wrote. Statistical analysis “examined the extent to which providers demonstrated statistically reliable treatment differences across patient race and SES.” These findings were compared with how clinicians reported they used patient race and SES in decision making.
Demonstrated race-based decision making occurred for 41% of white and 31% of nonwhite clinicians. About two-thirds of providers (67.3%) were white, and of the remainder, 26.3% were Asian, 4.4% were classified as “other,” and 2.1% were black. The respondents were aged a mean 29.7 years, and were 42.3% female.
In addition, Ms. Grant and her coauthors estimated provider SES by asking about parental SES, dividing respondents into low (less than $38,000), medium ($38,000-$75,000), and high (greater than $75,000) SES categories.
and similar across levels of provider SES, at 41%, 43%, and 38% for low, medium, and high SES residents and fellows, respectively. However, the disconnect between reported and demonstrated bias that was seen with race was not seen with SES bias, with 43%-48% of providers in each SES group reporting that they had factored patient SES into their treatment decision making.
“These results suggest that providers have low awareness of making different pain treatment decisions” for black patients, compared with decision making for white patients, Ms. Grant and her colleagues wrote. “Decision-making awareness did not substantially differ across provider race or SES.” She and her collaborators called for more research into whether raising awareness about demonstrated racial bias in decision making can improve both racial and socioeconomic gaps in pain care.
The authors reported funding from the National Institutes of Health. They reported no conflicts of interest.