Study Finds Patients Prefer Doctors Who Share Their Same Race/Ethnicity

PHILADELPHIA – Patients who shared the same racial or ethnic background as their physician were more likely to give the maximum patient rating score, according to a new analysis of 117,589 patient surveys from 2014 to 2017. The findings, from a team of Penn Medicine researchers, were published today in JAMA Network Open.

“What it comes down to is that patients who see physicians of their own race or ethnicity are more likely to rate their physicians higher than patients who see physicians of a different race or ethnicity,” said the study’s corresponding author Junko Takeshita, MD, PhD, MSCE, an assistant professor of Dermatology and Epidemiology in the Perelman School of Medicine at the University of Pennsylvania.

The study’s findings underscore the need for health systems to address issues of implicit biases and to diversify the physician workforce, she added. “In a healthcare setting, racial and ethnic biases can affect health care delivery and ultimately, lead to health disparities.”

In order to improve patient-centered care, health systems have made strides to both measure and publicly share patient ratings for individual physicians, sometimes linking these ratings to promotion and compensation decisions. The Press Ganey Outpatient Medical Practice Survey is one such tool used by many practices to evaluate the patient experience. Despite its wide use, the relationships between the patient experience and patient, physician, and clinical encounter characteristics are not completely understood, particularly when it comes to whether racial/ethnic concordance leads to higher Press Ganey scores.

To explore the relationships between race/ethnicity, as well as gender, and patient experience, the research team examined 117,589 Press Ganey surveys collected following adult outpatient visits across various medical specialties within the University of Pennsylvania Health System between July 2014 and May 2017. The patients were predominantly white (81.6 percent), followed by Black (12.8 percent), Asian (3.4 percent), and Hispanic (2.3 percent).

The surveys query patients on six domains: access, moving through your visit, nurse/assistant, care provider, personal issues, and overall assessment. Patients can rate their experience in each domain, ranging from one (very poor) to five (very good). For their study, the researchers focused on the responses for the “likelihood of your recommending this care provider to others” in the care provider domain of the survey.

They found that 87.6 percent of physicians from racially/ethnically concordant patient-physician pairs received the maximum score for this question, while only 82.1 percent of physicians from discordant pairs received the maximum score.

At the level of specific patient-physician racial/ethnic pairs, among Black patients, for example, the average ratings for white physicians were 0.03 points lower than those for Black physicians. According to Takeshita, these seemingly small differences are significant.

“Even a 0.02 difference in the mean score for the Press Ganey survey could take a physician from the 100^th percentile down to the 70^th percentile in terms of rankings,” Takeshita said.

Along with examining racial/ethnic concordance, the authors also investigated gender concordance as a factor that might affect the patient experience. However, unlike the racial/ethnic concordance data, analyses incorporating gender did not reveal statistically significant differences in ratings.

While a valuable tool and reflection of the patient experience, there are limitations to the Press Ganey survey, said study senior author Deirdre Sawinski, MD, an associate professor of Renal-Electrolyte and Hypertension in the Perelman School of Medicine. In particular, it was not meant to be used to evaluate and compare individual physicians to one another, she said.

“Misapplication of these evaluations on an individual level can disadvantage some physicians in terms of survey data being used for promotions or bonuses,” Sawinski said. “Misinterpreting the data can contribute to physician burnout or job dissatisfaction and not get at the heart of what’s important for the patient experience. This study is an important step towards understanding how different factors contribute to create the patient experience, but clearly there is still more work to do. Most of all, we do not want our findings to discourage physicians from caring for a diverse patient population.”

Even so, the study’s findings suggest that racial/ethnic concordance does in fact play a role in patient-physician interactions, and the team hopes this will generate discussion and action to improve cultural competency among all physicians, so that they are prepared to care for a diverse patient population.

“Our data highlights why it’s more important than ever to have a diverse physician workforce who looks like all the different types of patients we take care of, including different genders and different races,” Sawinski said.

Penn authors Shiyu Wang, Alison Loren, Nandita Mitra, Justine Shults, and Daniel Shin, also contributed to the study.