In 2012, blacks and Hispanics made up more than 30 percent of the U.S. population, but only 14 percent of medical school graduates and less than 10 percent of practicing physicians. (iStock)

I’ve never cared for a Hmong child, but I often think about what it would be like.

The summer before we started medical school, I and other students were advised to read Anne Fadiman’s “The Spirit Catches You and You Fall Down.”

The book chronicles the illness of Lia Lee, a Hmong girl with severe epilepsy, and her family’s saga navigating the American medical system. The Hmong come from Southeast Asia.

The story has become a symbol of the sometimes devastating consequences of a cultural divide — a cautionary tale of miscommunication, misperception and mistrust, culminating in a catastrophic two-hour seizure and permanent brain damage. Lia died in 2012, after living the last 26 years of her life in a persistent vegetative state.

This is, of course, an extreme case — and there are fewer than 300,000 people of Hmong descent living in the United States. But many of the cultural, structural and language barriers that plagued Lia’s experience remain pervasive in medicine and may be more important than ever, amid growing racial and ethnic diversity in the United States.

But even as we care for an increasingly diverse patient population, we haven’t done enough to diversify the medical profession. Between 1978 and 2008, 88 percent of graduates of U.S. medical schools were white or Asian. Blacks, American Indians and Hispanics together made up the remaining 12 percent. And while in some cases we’ve made progress, in others, we’re actually moving backward.

A new working paper provides compelling evidence for how a more diverse medical workforce could improve health outcomes. Researchers recruited more than 1,300 black men in Oakland, Calif., and randomly assigned them to either a black male doctor or a nonblack male doctor in the same clinic. Before meeting their doctors, patients selected which preventive screening tests they would like, if any. These included both noninvasive tests, such as blood pressure measurements, and invasive tests, such as diabetes and cholesterol screening, which require blood samples.

Initially, all patients chose roughly the same number of screening tests. But after a conversation with their doctor, black men with a black doctor were substantially more likely to opt for every test than those with a nonblack doctor. The effect was particularly pronounced for invasive tests, which require more trust between doctor and patient: Patients with black doctors were 47 percent more likely to get diabetes screening and 72 percent more likely to get cholesterol tests.

“I was really surprised by the size of the effect,” said Marcella Alsan, lead author of the study and an associate professor of medicine at the Stanford School of Medicine. “We wanted to go beyond just documenting disparities and see if there was a way to intervene. This seemed to have a bigger impact than we expected.”

Black men paired with black doctors were also much more likely to discuss their other health problems, and black doctors wrote more notes about their patients — including about personal problems they were struggling with — than did nonblack doctors.

“I never approach patients based on race alone, but there’s no doubt it’s an important factor at times,” said Oluwaferanmi Okanlami, director of medical student success in the Office of Health Equity and Inclusion at Michigan Medicine. “Medical encounters are inherently uncomfortable, and people often yearn for a sense of familiarity. I’m much more likely to hear ‘I’m so happy to have a black doctor!’ when I’m caring for a black patient than a patient from another background.”

On average, black men in America die nearly half a decade earlier than white men, and have among the shortest life expectancies of any major demographic group. Having more black doctors might reduce deaths caused chronic disease, but many black men don’t have the opportunity to see doctors who look like them. Nor do members of other racial and ethnic groups.

During a period of large demographic shifts across the United States, changes in the physician workforce haven’t kept up. Since 1978, there has been just a four- percentage-point increase in the proportion of Hispanic medical school graduates and less than a two-point increase in black graduates. The number of black men in medical school has actually declined since the 1990s, and black women now outnumber black men 2 to 1 in medical school.

In 2012, blacks and Hispanics made up more than 30 percent of the U.S. population, but only 14 percent of medical school graduates and less than 10 percent of practicing physicians.

“There’s a severe under­representation of certain minority groups in medicine,” Alsan said. “And an increasingly strong argument to fix that.”

While individual interactions between doctors and patients are important, a more diverse physician workforce could also have broader effects. It could help us move toward a system in which the lived experience of minority groups is better understood and validated, and the barriers they face more readily identified and addressed.

Black, Hispanic and immigrant physicians are, for example, much more likely to practice in underserved areas and to care for uninsured patients and those on Medicaid. They’re also more likely to choose specialties with doctor shortages, such as primary care and pediatrics.

But many effects are more subtle: biases confronted, research questions asked, new perspectives recognized.

“This is about creating a system with more structural competency,” said Lisa Cooper, the Bloomberg Distinguished Professor in Health Equity at Johns Hopkins Medicine. “It’s about giving doctors the skills, knowledge and attitudes to understand the many societal and historical factors that affect health, instead of thinking of health as something that’s just the result of individual choices and behaviors.”

One study found that half of medical students and residents hold erroneous beliefs about biological differences between blacks and whites, such as believing that blacks have thicker skin, that their blood clots faster or that they have less-sensitive nerve endings. Perhaps unsurprisingly, black Americans are systematically undertreated for pain. Other research finds that doctors are less patient-centered and more verbally dominant with minorities and that they show less empathy in end-of-life conversations with these patients.

More diversity could help. Students who attend medical schools with more-diverse student bodies feel more prepared to care for patients from different backgrounds. And more-diverse research groups may be more likely to explore the interests of marginalized populations and to publish more impactful studies.

More broadly, research suggests that greater exposure to people of other backgrounds can reduce bias and that black patients who are treated at hospitals with more-diverse patient populations fare better — possibly because doctors are more familiar with their social and cultural circumstances.

“The argument for more diversity is really one about changing the overall climate,” Cooper said. “Diversity brings new energy and creativity. It allows physicians to learn from one another and to focus on problems that may not have been recognized before.”

Despite advances in treatments and technologies, medicine remains a fundamentally human field. The care we give and the care we receive is powerfully influenced by the interactions between doctor and patient. Progress toward a more equitable system will require focusing not only on who’s on the exam table but also on who’s beside it.

Khullar is a physician at NewYork-Presbyterian Hospital, an assistant professor in Weill Cornell’s medicine and health-care policy and research departments, and director of policy dissemination at the Physicians Foundation Center for Physician Practice and Leadership. Follow him on Twitter at @DhruvKhullar.