This special report is the second of two parts. Read part one here.
When Quinn Capers IV took over as associate dean of admissions at the Ohio State University College of Medicine in 2009, just 13% of the entering class came from racial and ethnic groups underrepresented in medicine. A Black cardiologist, he thought the school could do better.
Capers launched a series of changes, starting with testing the 140 members of the admissions committee (like the faculty, largely white) for implicit racial or gender biases that might make them judge some applicants more harshly. The study found up to 70% of the committee held such biases, and its members — who read applications, decide whom to interview, and choose which candidates to admit — then underwent training to mitigate those biases.
The following year’s class included 17% of students from underrepresented groups, the school’s most diverse class ever. “I was hoping for it, but I was pleasantly surprised,” said Capers.
Diversifying the physician workforce, long seen as one key to reducing the nation’s racial and ethnic health disparities, was a major focus of the landmark National Academies report “Unequal Treatment,” which 20 years ago examined how systemic racism leads to poorer medical care for people of color in the U.S. STAT reported Wednesday that little progress has been made to address the problems highlighted in the report, and that, despite a pandemic that reinforced the depth and reach of these inequities, complacency could still win out, or new crises, such as the current nursing shortage, could get in the way.
“The big risk is that we’ll continue to admire the problem and won’t get to do anything about it,” said Kedar Mate, a physician who leads the Boston-based Institute for Healthcare Improvement, which is pushing health care systems to reduce disparities in the populations they serve.
But as the work of Capers and others shows, there are signs of hope. Several programs around the country — some groundbreaking and audacious in their scope, others smaller but surprisingly effective — reveal the struggle to advance health equity may be at an important inflection point. They include an effort to make a medical center fully antiracist; a program to collect and use high-quality racial, ethnic, and language data to measure and counter disparities on a statewide level; another to harness the power of electronic records to reduce pneumonia deaths in young Black patients; and Capers’ work to boost the number of Black and brown students at a top medical school.
These programs succeeded, those who created them said, because they had buy-in from the top leadership of their organizations, and because the people implementing them were given wide-ranging authority. They did not just set lofty goals, but held specific people accountable for meeting them. The goal now is for such programs to serve as models for the nation.
Diversifying a medical school
While research has consistently shown that patients of color would fare better if there were more physicians who looked like them, the number of Black physicians has stayed below 6%, and the number of Black male students enrolling in medical school has stagnated for decades.
At Ohio State, Capers has shown big changes can be achieved with persistence. After that first year, when the proportion of first-year students from underrepresented groups reached 17%, he was surprised to see that bias training paid off in ways he’d hadn’t expected. The increased numbers were not because Ohio State accepted more students from traditionally underrepresented backgrounds, but because more of those students who were accepted decided to enroll. “Those minority students, when they interviewed at Ohio State, sensed an environment of inclusion,” he said. “Many had multiple offers and chose us.”
Capers next pushed for a holistic admissions process that didn’t focus so heavily on grades and MCAT admissions-test scores, which are known to disadvantage students from lower-income families and communities traditionally underrepresented in medicine. The idea wasn’t immediately embraced. In this case, he wasn’t facing bias in the admissions committee, but the inertia of tradition. There were also fears that such changes might lead to lower average test scores, and a drop in the all-important medical school rankings published by U.S. News & World Report. “Some people found that really difficult,” Capers said. “They thought we must be out of our minds.”
“The big risk is that we’ll continue to admire the problem and won’t get to do anything about it.”
Kedar Mate, CEO of Institute for Healthcare Improvement
Capers didn’t throw out the test scores and grades altogether; he included for consideration only students whose grades and scores were above a cutoff that predicted success. But he blinded the committee to students’ individual scores. In another surprise, the overall scores and grades of admitted applicants went up, likely again because top applicants were choosing the school. “Once the admissions committee saw that, they relaxed,” Capers said.
Another major change: Capers stopped having applications screened by just two people (the admissions dean and an assistant) and instead asked a large group of faculty to screen the thousands of applications received each year to limit bias. While many colleagues told him no faculty members would want to take on such work, Capers was surprised yet again.
“We had 80 people sign up,” he said. “We had to turn people away.” He also put women, faculty from marginalized groups, and younger faculty on the admissions committee, allowed voting on candidates to be anonymous, and removed photos from applications for final voting.
The work wasn’t always easy, Capers said. There was pushback at times, and some faculty members were extremely uncomfortable discussing race. The key, he said, was support from top leadership so he could make the changes he thought necessary. Now a professor of internal medicine and associate dean for faculty diversity at UT Southwestern Medical Center in Dallas, Capers is working on enhancing diversity and inclusion for faculty and helping guide medical schools around the country that hope to make changes like those at Ohio State.
“With every tweak, we saw enhanced diversity,” said Capers. By 2016, 26% of the medical school class included students from underrepresented backgrounds. In a new U.S. News list — one based on how diverse medical schools are — Ohio State ranks seventh in the nation.
A ‘nudge’ to curb deaths
Cone Health, a large, multi-hospital nonprofit health care network in North Carolina has been nationally recognized for its efforts to reduce disparities, which include carefully analyzing data on health outcomes and mortality by race.
While examining non-Covid pneumonia deaths in recent years, a Cone Health team found a disturbing trend. Black patients being treated for pneumonia were, on average, 11 years younger than white patients with the disease. And despite being younger, they had the same number of comorbidities as older patients and were at higher risk of death. “We immediately saw a disparity,” said Monica Schmidt, the network’s executive director for health economics. Further analysis showed Black patients were experiencing longer waits before being given antibiotics or being admitted to the intensive care unit.
“We knew we could do better,” said Karen Bartles, a respiratory therapist who directs cardio-pulmonary services at Alamance Regional Medical Center. “Once we started drilling into the data, we saw opportunities around the appropriate level of care.”
What the team decided to do seems surprisingly simple. It added a “nudge” to the electronic health record system: For any patient from a historically marginalized group who was diagnosed with pneumonia, their physician would automatically receive a message saying, “Our data suggest that (patient name) has been identified at increased risk of severe complications from pneumonia. Consider a higher level of care if clinically indicated.”
The change led to dramatically decreased wait times for both drugs and ICU admissions for Black patients, said Schmidt, and what appears to be decreased mortality. The analysis does not include enough patients to be statistically significant yet, she said, but showed a drop in mortality from 7% to 6% in just a matter of months.
“When we saw that, we all got so excited as a team,” said Bartles. “It’s inspirational.”
One reason the electronic nudges appear to be working, team members said, is that they send a neutral message to practitioners and don’t trigger defensiveness. “We’ll never know what’s in a physician’s mind,” said Schmidt. “We’re not pointing any fingers by showing the data. We’re just saying, ‘This exists and we need to fix it.’”
That’s crucial, says the Institute for Healthcare Improvement’s Mate, who has been working with Cone Health and 21 other systems in a partnership to reduce health disparities. Mate said systems need to educate physicians that these problems are systemic so that when they are presented with data showing health disparities exist in their practice, they can respond without getting personally defensive. “You have to be prepared for people saying, ‘This doesn’t happen here.’” he said. “Come prepared with data.”
One challenge, said Sendil Krishnan, an internist who leads Cone Health’s efforts to reduce inpatient mortality, is that physicians have been trained to look at everyone the same and not consider race in the care they provide. And now, to address the effects of systemic racism, they’re being asked to take patients’ race into account. “That’s part of the problem,” he said. “We have not viewed care through an equity lens.”
Mate has seen the importance of such education firsthand. An internist, he recently stopped using the race-based measures of kidney function he’d relied on for 20 years that many have denounced as flawed and dangerous. “I had no clue that racism and racist science were baked into those numbers,” he said.
Hospitals also need to acknowledge their own past racism, Mate said, which is something Cone Health has done. In 2016, the system’s leadership apologized to Alvin Blount, the last living member of a group of Black physicians who in 1962 sued two of the system’s hospitals in an effort to be able to treat Black patients there. (The historic suit, Simkins vs. Moses H. Cone Memorial Hospital, eventually led to the elimination of segregated care.) Cone Health now helps fund a scholarship honoring Blount and other plaintiffs.
But progress on eliminating disparities nationally remains uneven. For every success story he’s seen, “there are plenty of health systems that would rather see this go away,” said Mate. “They don’t want to see the data. They’re afraid of the data.”
Meanwhile, the data team at Cone Health isn’t stopping its work. “It’s very much like drinking out of a firehose,” said Krishnan. “Once you start, you want to fix all of it.”
Measuring widespread health inequities
The “Unequal Treatment” report found race and ethnicity data were not standardized in health care or even always collected. That problem continues today: There is no unified, national effort to gather high-quality health data broken down by race and ethnicity.
“There’s a saying, ‘Where there’s no data, there’s no problem,’” said Daniel Dawes, a health policy researcher who directs the Satcher Health Leadership Institute at the Morehouse School of Medicine and recently launched an effort to collect such data and pinpoint where major gaps exist. “It’s like flying a plane with 50 different instrument panels since each state collects data differently.”
Risa Lavizzo-Mourey, a professor emerita at the University of Pennsylvania and former president and CEO of the Robert Wood Johnson Foundation, helped lead the committee that wrote the report. “It was an uphill battle to get reporting by race, ethnicity, language,” she said, adding that it is still sorely needed. “If you look at a clinic and there are longer waiting times by race or provider, that kind of granular information can allow for quality improvement and get providers to understand and address their own bias.”
While the collection of racial and ethnic data has largely lagged, one state — Minnesota — has quietly charged ahead.
MN Community Measurement is a nonprofit organization founded in 2002 to harness data to improve health care. The group collects data from partners that include most of the state’s primary health care providers and nearly all of its major health plans — some 1,300 practice sites — and then shares that data so individual clinics, hospitals, and health care systems can make improvements.
“The collection of data can be very fragmented,” said Julie Sonier, the group’s president and CEO. “We were founded as a neutral place to aggregate data for everyone’s purposes.”
Recognizing that deep health disparities existed in the state, the group in 2008 decided to start systematically collecting and analyzing data on patients’ race, ethnicity, language, and country of origin. It was difficult at first. Some providers didn’t have electronic health records at the time, some didn’t feel they had the resources to collect the data, others said some staff were uncomfortable asking patients about their identities.
To help, MN Community Measurement created a handbook on best practices for data collection. “One way is to tell patients, this isn’t about you, it’s about us,” said Sonier. A key to success, she said, was having the leadership of the various health care organizations buy in to the need for collecting such data. The group is now seeing strong support overall, she said.
Collecting data was not the only hurdle. Sonier’s group had to make sure the data, coming from a variety of different sources, was consistent and of high quality so that the pooled data could be trusted. Lots of discussion — and continued education about best practices — helped.
“I don’t think you can effectively tackle disparities, or measure if actions are having an impact, without data. It seems obvious in hindsight.”
Julie Sonier, president and CEO of MN Community Measurement
The data have already been used, Sonier said, to close large deficits in colorectal screening for Hispanic patients in the state and for analysis of which populations have higher rates of diabetes. The group publishes an annual report on disparities. “I don’t think you can effectively tackle disparities, or measure if actions are having an impact, without data,” she said. “It seems obvious in hindsight.”
The work of MN Community Measurement was recognized as one of the best in the nation in a recent RAND report on measuring health equity. “It can’t be done overnight,” said Sonier. “But it can be done.”
Creating an antiracist medical center
Before the pandemic and the racial reckoning that followed the murder of George Floyd, work on health equity was already well underway on the campus of Vanderbilt University Medical Center in Nashville. School leaders had launched an office of health equity in 2019, were developing a program for medical students to earn a certificate of health equity, and were having conversations about how to curb disparities. But after the summer of 2020, Consuelo Wilkins, a professor of medicine and a gerontologist who serves as the medical center’s senior vice president for health equity and inclusive excellence, knew it wasn’t nearly enough.
When Vanderbilt, like many other medical centers, pledged to confront racism that summer, Wilkins was asked to lead the charge. She started by making sure the organization’s leadership would support her vision for change. And she insisted health care workers could no longer distance themselves from racism by hiding behind the nobility of their profession.
She didn’t want a small task force made up of a few hospital leaders that met sporadically. She wanted a large task force that included employees who had experienced racism and lower-wage workers who had traditionally been left out of high-level discussions. She wanted to run the project the way she saw fit and set a goal of becoming nothing less than an antiracist medical center, which means full recognition that the center is part of a culture that has oppressed and systematically disadvantaged many racial and ethnic groups, and that it is committed to eliminating those injustices.
“I said, ‘I need you all to get people out of my way so I can do this work,’” Wilkins said. “I couldn’t do that if I didn’t have the support of senior leadership and if I wasn’t already at the table.”
Her task force ended up including more than 100 employees, among them people from food services, environmental services, and the campus police. An early challenge was making sure everyone felt comfortable speaking openly, or even speaking at all, during meetings.
“We had a lot of conversations early on about how we were going to balance power,” Wilkins said. “We told leaders, the physicians in particular, that when you go to this meeting, if others are being called by their first name, you’re going to be called by your first name too. We said, don’t show up at this meeting with your white coat on, you’re not seeing patients here.”
Everyone’s opinion mattered, she said. “They may be leaders in their community, or deacons in their church,” Wilkins said of the lower-wage workers. “They know how to run a meeting.”
“I said, ‘I need you all to get people out of my way so I can do this work.’ I couldn’t do that if I didn’t have the support of senior leadership and if I wasn’t already at the table.”
Consuelo Wilkins, Vanderbilt University Medical Center senior vice president for health equity and inclusive excellence
The group spent five months studying data and conducting surveys, interviews, and listening sessions to come up with a list of 187 actions that Vanderbilt could take to become an antiracist medical center. But a list of recommendations wasn’t enough for Wilkins. She’d learned that some of her group’s recommendations had been made before, but never acted on. “I spent a lot of time thinking about why nothing happened,” she said. “A big eye-opener for me was there was no accountability.”
Instead of giving a long list of recommendations to a dean or CEO, who have much on their plates, Wilkins’ task force listed a specific person or group to be accountable for each recommendation. A more diverse board? That’s the chair of the board’s responsibility. More scholarships for students of color? VP for development. Better treatment of students of color using the library at night? Chief of police. Better race, ethnicity, and language data in health records? Health IT. “There has to be a name next to every action,” she said.
Such work — in a red, Southern state like Tennessee — could prove challenging. “Some of these sessions, when we started to talk about white supremacy, the shock and withdrawal started to creep in pretty quickly,” Wilkins said. “People tried to tie health disparities to poverty. The myth of meritocracy was pretty strong.”
One thing that helped was having all senior leadership, including the board, have a full day of antiracism training, with the invitation and opening remarks coming from top leadership. “I said, ‘This is not going to be me and my team pushing this, it has to come from you,’” she said.
Even after accomplishing so much, Wilkins faces a difficult balancing act. “We have people who are all in on this work who think we are not moving fast enough, and who say, ‘Those recommendations were a year ago and you have not wiped out racism yet, what are you doing, sleeping?’ And then we have people who say, ‘You’re moving way too fast, I’m so tired of hearing about racism,’” she said. “In the midst of a pandemic and Great Resignation, particularly of nurses, there are so many competing demands. It’s easy for this work to get lost.”
Major goals include offering more education, training, and health and wellness opportunities for the center’s lower-wage workers, who are disproportionately people of color; acknowledging that race is a social construct and is not tied to biology or genetics; and ensuring health care workers understand the health impacts of structural racism.
The school has already made progress: The first-year medical school curriculum now contains antiracism training, and the minimum wage has been raised to $15 an hour. In an effort to help Vanderbilt meet another goal — confronting racism in its past — the school has changed the name of Dixie Place, a street that runs through campus, to Vivien Thomas Way.
Thomas was a pioneering Black surgeon who developed a technique to treat a birth defect affecting the heart called tetralogy of Fallot. Despite his medical success, Thomas, the grandson of a slave, spent his time at Vanderbilt classified — and paid — as a janitor. At the time, that was the only employment category available to Black employees.
This is part of a series of articles exploring racism in health and medicine that is funded by a grant from the Commonwealth Fund.