Raising The Stakes To Advance Equity In Black Maternal Health

Baby Zahiri: Shortly after delivering her baby, Zahiri, Talicia Williams showed signs of hemorrhage. A nurse issued an alert signaling an obstetric emergency and initiating a set of protocols that encompasses evidence-based recommendations and has been shown to reduce mortality rates.

Photographs by Megan Tidmore, courtesy of Penn Medicine

Under a blanket of gray clouds, a stream of people strode out of the medical center onto Franklin Field at the University of Pennsylvania. The doctors and trainees in white coats, nurses in blue scrubs, and office workers and other health care staff were ever mindful of the pandemic emergency that they confronted together. They stood six feet apart, wearing masks, some climbing into the stands of the football stadium to get extra space.

Yet they felt a common purpose. It was June 5, 2020, a warm and humid afternoon, when they collectively paused from their work to gather in outrage over the killing of George Floyd by a Minneapolis, Minnesota, police officer.

Kevin Mahoney, CEO of the University of Pennsylvania Health System, stood among them, joining this remarkable moment of solidarity. But Mahoney’s thoughts went beyond the campus to the everyday struggles in poor Philadelphia neighborhoods, and more broadly to the arc of history. His generation, coming of age in the wake of the Civil Rights movement, had celebrated advances toward racial justice, but he saw with new clarity how progress had stalled.

Florencia Polite, chief of the Division of General Obstetrics and Gynecology at Penn Medicine, prepared her colleagues for a silent tribute to Floyd—kneeling for about nine minutes, the time that officer Derek Chauvin pressed his knee against Floyd’s neck. “We are going to be uncomfortable,” she said. “Our knees are going to hurt, and that is the point. We have been comfortable for far too long.”1 She challenged them to move from the symbolic action that day to the real work of dismantling structural racism in society.

“It really was a moment in time and a reckoning for me,” Mahoney recalls. “Watching that clock in the football stadium tick down—it was very emotional and very hard.”

A few days later, Mahoney received a letter from the Alliance for Minority Physicians, a Penn organization that supports the recruitment and career development of physicians and medical students who are underrepresented minorities. Signed by more than 130 Penn physicians, it called on leaders of Penn Medicine and the Children’s Hospital of Philadelphia “to act quickly and decisively to establish a new institutional culture—relationships, policies and practices—that is anti-racist.” It suggested specific steps to “eliminate longstanding institutional racism.”

Mahoney took the letter to heart. Along with J. Larry Jameson, executive vice president of the University of Pennsylvania and dean of the Perelman School of Medicine, he committed to building “a more inclusive and just community.”2 Although the health system already had some vice chairs of diversity, equity, and inclusion, Mahoney established the same position in every clinical department. He required unconscious bias training for all employees and set up a system of reverse mentoring, in which executives spend time with community health workers to learn about their work responding to the needs of patients in disadvantaged communities. Health equity is “a complex, hard issue that we need to be intentional about fixing every day,” he says.

Mahoney also decided to address one of the nation’s most immutable and troubling health disparities: Black women in the US are more than three times more likely to die in pregnancy, childbirth, and the postpartum year than White women—a gap that persists regardless of income or education.3 The Black-White disparity in maternal death has been recognized for more than a hundred years, since authorities first began collecting data.4 In Philadelphia the nation’s first city-based Maternal Mortality Review Committee analyzed 110 pregnancy-associated deaths during the period 2013–18. It found that 73 percent of deaths related to pregnancy occurred among Black women, although they accounted for only 43 percent of live births in that period.5

It was against the backdrop of these sobering statistics that Mahoney decided to give Black maternal health the highest possible spotlight: For fiscal year 2021, which began September 1, 2020, he added health equity to the Penn Medicine team goals. The goals determine between 10 percent and 40 percent of compensation for more than 600 senior leaders, including Mahoney. Whether Penn Medicine could reduce “major maternal morbidity” (serious complications of pregnancy, birth, or postpartum) and maternal mortality among Black women would be factored into determining executive pay for the health system and medical school.

A second health equity goal targets screening for colorectal cancer among Black patients. University of Pennsylvania president Amy Gutmann, Jameson, and the university’s Board of Trustees approved the equity aims.

These two aims are among about twenty team goals, including improving patient satisfaction and reducing health care–acquired infections. The framework of systemwide goals arose in the 1990s, when the University of Pennsylvania’s medical school became integrated with its teaching hospital and physician practices into the academic health system known as Penn Medicine.

Of course, Penn Medicine seeks to reduce severe complications of pregnancy and childbirth for all women, but by specifying Black women, the team goal put racial disparities in maternal morbidity at the forefront, Mahoney says. In the first year after setting the goal, the health system was able to reduce severe pregnancy-related complications among Black women by 29.4 percent (Elizabeth Howell, Perelman School of Medicine, University of Pennsylvania, personal communication, December 6, 2021).

“I hope that we’re setting the tone within Penn Medicine that structural racism needs to be eliminated,” Mahoney says.

A happy ending: Tayale Williams was excited to welcome his little sister, Zahiri. Although his mom, Talicia, spent four days in the hospital after Zahiri’s birth, these days she stresses the happy ending to her story.

Increased Risk For Black Women

As one looks east from the University of Pennsylvania campus across the Schuylkill River at the glass towers of the city skyline, Philadelphia appears as an archetype of progress and prosperity—one of the nation’s oldest cities reaching toward the future. But to the west, beyond the gentrified area known as University City, the neighborhoods of aging row houses tell a more complicated story.

Almost one in four Philadelphians live in poverty, making it the poorest of the ten largest US cities.6 Philadelphia was once a blue-collar powerhouse, but over time, well-paying manufacturing jobs were replaced by low-paying service jobs. The city’s homicide rate, which was already among the nation’s highest, surged during the COVID-19 pandemic. One searing tragedy occurred on July 4, 2021, in West Philadelphia, when four men sprayed a holiday cookout with a hail of gunfire, killing two men and injuring a sixteen-year-old girl. The shooting, possibly linked to a neighborhood rivalry that targeted someone who attended the cookout, shocked a community that has wearily grown accustomed to violence.7 It happened just about two miles from the Penn campus.

Living in a violent neighborhood is associated with a higher risk for complications in pregnancy, childbirth, and postpartum, perhaps because of persistent stress. But strikingly, so is just being Black. University of Pennsylvania researchers analyzed 63,334 pregnancies in the health system during the period 2010–17 and found that with every 10 percent increase in people who identify as Black in a census tract, the risk for severe maternal morbidity in that tract increased by 2.4 percent.8

That correlation could reflect the experience of living in a marginalized community, says Mary Regina Boland, principal investigator of the study and a Penn assistant professor of informatics. For example, the pregnant people may lack access to fresh produce or green space, have greater exposure to environmental toxins, or encounter ongoing stressors caused by inequities, she says. “The communities may be experiencing various forms of racism both directly and indirectly,” she notes.

Talicia Williams, a twenty-five-year-old private security officer who lives in West Philly, didn’t know anything about those statistics—or, more broadly, about the increased maternal risks for Black women like herself—when she became pregnant in late 2020. Her first pregnancy had been uneventful, aside from a rise in blood pressure that led doctors to induce her delivery early, and she had no reason to think that her second one would be difficult.

Yet suddenly, it was. She developed nausea so severe that she threw up blood and even ended up in the hospital. She took antiemetics, but they didn’t make the nausea go away. She adapted with a very specific diet of foods that were more palatable—fruit cup but no fruit juice, for example. She could eat flavored ice, but only lemon. She drank water, but only if it was ice-cold.

She tried to keep a good attitude. “I was actually fine through everything,” she says. “I wasn’t emotional or anything.”

Then, as suddenly as the nausea began, it went away as she entered her third trimester. Her baby was in the breach position, but doctors were able to turn her. As she talked to her son, Tayale, who was two years old, about the baby sister on the way, she was focused on the joys of having a girl. Yet this pregnancy had more serious challenges ahead.

Mortality Rates Continue To Rise

The great majority of childbirth stories are happy ones, even amid discomfort, pain, and tense or frightening moments. But in the US, Black women are more likely to encounter problems. Their cesarean section rate in 2019, 35.9 percent of live births, was higher than the 30.7 percent rate for White women.9

Black women are also more likely than White women to experience severe outcomes of postpartum hemorrhage—needing a transfusion, for example, or even a hysterectomy to control excessive bleeding—according to an analysis of 360,370 cases of postpartum hemorrhage between 2012 and 2014 identified in the National Inpatient Sample, using data collected by the Agency for Healthcare Research and Quality.10

Reducing US maternal mortality has long been a national priority, yet it has been an elusive goal.

Reducing US maternal mortality has long been a national priority, yet it has been an elusive goal. In the ten-year preventive health initiative known as Healthy People 2020, the Department of Health and Human Services aimed to reduce the rate from 12.7 maternal deaths per 100,000 live births in 2007 to 11.4 by 2020.11 Instead, the rate rose, reaching 20.1 maternal deaths per 100,000 live births in 2019.12

To reduce the risk, hospitals typically focus on preventing or quickly responding to complications of pregnancy or childbirth. The Centers for Disease Control and Prevention (CDC) provides reports on twenty-one indicators of severe maternal morbidity. The agency found that the proportion of births requiring women to have blood transfusions because of hemorrhage in 2014 was five times higher than in 1993. Other serious complications such as kidney failure or respiratory distress syndrome rose as well, although they remained much less common than blood transfusions, which occur after significant blood loss in delivery.13

“It’s a movement that is going to take concerted effort and momentum and a little bit of time,” says Sindhu Srinivas, director of obstetrical services at the Hospital of the University of Pennsylvania and physician lead of the Penn Medicine Women’s Health Service Line, of the effort to reduce maternal complications. “But the time is now. The need is urgent.”

Finding A Metric

When Kevin Mahoney settled on the idea of a team goal of improving Black maternal health, he needed a way to track the progress. The decision was fraught—widely used metrics focus on public health trends, not hospital outcomes. So Mahoney turned to Elizabeth Howell, the newly named chair of the Department of Obstetrics and Gynecology in the Perelman School of Medicine. She came to Penn from the Icahn School of Medicine at Mount Sinai in New York City, where she gained renown as a leading expert in racial and ethnic disparities in maternal and infant health.

Two years before, journalists from USA Today had calculated rates of severe maternal morbidity for hospitals in thirteen states, based on deidentified patient discharge records from public agencies. The 2018 investigative piece asserted, “Hospitals know how to protect mothers. They just aren’t doing it.”14

USA Today provided comparisons based on the CDC indicators, but the newspaper acknowledged that the results weren’t risk-adjusted. Medical centers that specialized in treating high-risk pregnancies would be expected to have higher rates of complications than those that didn’t. The newspaper reported that the Hospital of the University of Pennsylvania’s rate of severe maternal morbidity for Black women was slightly higher than the state average—2.2 percent versus 2.0 percent—and below the national rate of 2.6 percent. Pennsylvania Hospital, another Penn Medicine hospital, located across the Schuylkill River in Philadelphia’s Center City, had a rate equal to the state average, and Lancaster General Health, which had recently become part of the health system, had too few deliveries among Black women to calculate a rate.15

Although variation in patient risk factors could make it unfair to compare hospitals directly, the CDC indicators are still useful to track patient outcomes and identify opportunities for quality improvement, Howell says. Still, setting a metric for the team goal posed other challenges. Severe maternal complications, although much more common than maternal deaths, are still rare events, which means that it would be difficult to discern improvements in care. The CDC estimates that in 2014 more than 50,000 US women experienced severe pregnancy-related complications13 out of about four million births.16 (In 2019, 754 women died of “maternal causes” in the US, according to the National Center for Health Statistics.)12

Howell worked with Srinivas, who is also vice chair for quality and safety at the Hospital of the University of Pennsylvania. Under their direction, Penn Medicine created a composite quality metric that includes measures for hemorrhage, infection, and embolism provided by the health care analytics company Vizient, which provides performance benchmarks for hospitals; the CDC severe maternal morbidity metrics; and medical record codes for HELLP (hemolysis, elevated liver enzymes, and low platelets) syndrome, a life-threatening complication related to preeclampsia.

Howell knew that there would be limitations to any particular metric. “Could we come up with the right measure?” she recalls wondering. “It’s not the perfectionism of the research world. This involves doing things in the real world in real time.” For example, errors in coding could create false comparisons, so it would be essential for the system’s five hospitals with labor and delivery units to code consistently. Penn Medicine aims to standardize the measurement as much as possible.

Others also have struggled with the question of measurement. In 2020 Maryland launched a pilot project with six hospitals to track and review cases of severe maternal morbidity.17 That made it the second state, after Illinois, to conduct surveillance of severe complications as well as maternal deaths. The Maryland criteria are limited to admission to an intensive care unit or transfusion of four or more units of blood, which avoids issues of hospital miscoding when multiple diagnosis or procedure codes are used to identify severe complications.

“We’re trying to understand which of these [maternal complications] were actually preventable,” says Andreea Creanga, program director of the Maryland Maternal Health Innovation Program and an associate professor at the Johns Hopkins Bloomberg School of Public Health, in Baltimore. “By learning from these events, we are better able to come out with recommendations that are specific to improving care and reducing disparities in the state.”

‘We’re Not Going To Let You Die’

Talicia Williams was thirty-seven weeks pregnant when she went into labor. That evening she had been on her feet at a cookout—her best friend’s birthday party—so at first she wasn’t alarmed about waking in the middle of the night with back pain. She grabbed a heating pad and went back to sleep.

But the pain grew worse. She went to the bathroom, and as the pain became excruciating, she ended up on the floor, unable to stand. Her mother called an ambulance to take her to the Hospital of the University of Pennsylvania. In labor and delivery, Williams received epidural anesthesia, but it didn’t completely block the pain of the contractions. She could still feel them pulsing through her left leg.

Hours went by, and her labor wasn’t progressing. A doctor broke her amniotic sac and gave her two medications to spur labor, which brought improvement. But then Williams spiked a fever, a sign of infection. She received intravenous antibiotics.

When she pushed and the baby emerged, if felt as if her body was ripping apart. “Instantly I got really cold, I was shivering,” Williams says. A nurse pressed on her abdomen near her belly button, a technique known as a fundal check to palpitate the uterus to ensure that it was contracted. At first, everything seemed normal, but then the nurse saw an unusually large pool of blood and she issued an alert signaling an obstetric emergency. What followed is an example of how recommended protocols unfold in real time—processes that were a focus of the Penn Medicine team goal.

The room instantly filled with doctors and nurses, who brought in a hemorrhage cart stocked with supplies. They counted and weighed blood-soaked pads. One doctor massaged the uterus from the outside while simultaneously reaching internally to clear out any large blood clots. A nurse administered intravenous fluids while someone else performed an ultrasound to look for any placental tissue that hadn’t been expelled.

Williams’s mother was holding the baby, Zahiri, as the scene ensued, with its urgent but coordinated activity. Williams felt blood flowing out of her, and she was aware of blood-soaked pads being changed beneath her and the crowd of medical personnel around her.

“Am I going to die?” she asked a nurse.

“No, we’re not going to let you die,” the nurse replied.

Hemorrhage is more common after cesarean section than vaginal birth, but it can occur when the uterus fails to contract after either type of delivery—a condition known as uterine atony. The medical team calculated that Williams lost 2.2 liters of blood, about 40 percent of her total blood volume.

Rapid assessment is key, says Adi Hirshberg, a maternal fetal medicine physician and associate director for obstetric services at the Hospital of the University of Pennsylvania. Williams’s doctors and nurses had recently practiced their skills in a simulation of a hemorrhage emergency, so everyone knew exactly how to respond. “She could have lost a lot more blood really quickly had we not gotten there right away,” Hirshberg says.

Inconsistent Practice

Hemorrhage is the most common serious complication of childbirth, and that made it a priority for the Penn Medicine team in addressing the health equity goal.

Hemorrhage is the most common serious complication of childbirth, and that made it a priority for the Penn Medicine team in addressing the health equity goal. Srinivas and Howell met with leadership at the health system’s five hospitals with labor and delivery services and asked them to identify a physician, a nurse, and a quality professional to participate in a monthly learning collaborative. The Women’s Health Service Line Champions began by developing a quality improvement framework.

Although this was the first systemwide ongoing project of its kind, it built on previous Penn Medicine and national efforts. In 2015 a collaborative of medical organizations called the National Partnership for Maternal Safety developed a “consensus bundle for obstetric hemorrhage,” the first maternal safety checklist that is now part of the Alliance for Innovation on Maternal Health, a federally funded quality initiative.18 Penn Medicine adopted the bundle, which encompasses evidence-based recommendations, including assessment of hemorrhage risk, measurement of blood loss, and readiness for quick response.

Such protocols have become important tools to reduce disparity in maternal care. In August 2021 the Centers for Medicare and Medicaid Services (CMS) added a new Maternal Morbidity Structural Measure to assess participation in a statewide or multistate quality improvement initiative in maternal health. As the agency explained in its Federal Register notice, “One of the main factors contributing to the increase in maternal morbidity and mortality is inconsistent obstetric practice.”19

Citing the “maternal health crisis” as a CMS priority, the agency said that it will ask hospitals a two-part question: “Does your hospital or health system participate in a Statewide and/or National Perinatal Quality Improvement Collaborative Program aimed at improving maternal outcomes during inpatient labor, delivery and post-partum care, and has it implemented patient safety practices or bundles related to maternal morbidity to address complications, including, but not limited to, hemorrhage, severe hypertension/preeclampsia or sepsis?”19

In a study that predated the team goal on Black maternal health, University of Pennsylvania maternal and fetal medicine physician Rebecca Hamm showed that protocols for labor induction could reduce or even eliminate disparities. From 2013 to 2015, 491 women who were in active labor but had a cervical dilation of two centimeters or less received interventions such as oxytocin on the basis of standardized recommendations. Another 364 women with a similar “unfavorable cervix” were treated on the basis of physician judgment.20

Black women treated according to the protocol were less likely to have a cesarean section than those treated on the basis of provider discretion (25.7 percent versus 34.2 percent in the control group), and their infants were less likely to have birth-related complications (2.9 percent compared with 8.9 percent in the control group). There were no significant differences among non-Black women.20 Providers don’t believe that they are treating Black women differently, “but biases play into what everyone does every day. It’s something you have to actively manage,” Hamm says.

In June 2020, as racial justice protests occurred around the country, a New England Journal of Medicine article reported on racial disparities within some medical algorithms, which embedded racially disparate treatment into the standard of care.21 Hamm realized that such a disparity existed in obstetrics: a lower hemoglobin cutoff for treating anemia in pregnancy among Black women of 10.2 grams per deciliter (g/dL) in the third trimester versus 11.0 g/dL for women of other races.22

The different cutoff stemmed from a 1993 report by the Institute of Medicine (now called the National Academy of Medicine) that cautioned against overtreating Black women for iron deficiency, as overall they had lower hemoglobin levels than women of other races.23 But Hamm and colleagues analyzed data from 1,369 women who delivered at the Hospital of the University of Pennsylvania during 2018–19 and showed that Black women with hemoglobin levels between 10.2 g/dL and 11.0 g/dL (the range untreated for them but triggering treatment for women of other races) were more likely to be anemic when they arrived for labor and delivery. Even mild anemia is associated with a greater risk of transfusion after childbirth, they found.24

The American College of Obstetricians and Gynecologists has since released a revised recommendation that removes the racial difference.25

Standardizing care with race-neutral protocols in itself doesn’t create health equity.

Standardizing care with race-neutral protocols in itself doesn’t create health equity; in California, a national leader in reducing maternal mortality, a racial gap widened even as the state’s overall maternal death rate dropped. The California Maternal Quality Care Collaborative, a public-private partnership, was launched in 2006 with maternal health toolkits and quality improvement learning collaboratives. Maternal mortality declined by 55 percent between 2006 and 2013.26 In 2018 California remained a national outlier for maternal health, with 11.7 maternal deaths per 100,000 live births compared with a US average of 17.4 deaths.27

Yet the racial disparity in California grew between 2008 and 2016, reaching a rate among Black women in 2014–16 that was six times higher than that among White women, at 56.2 versus 9.4 deaths per 100,000 live births within one year postpartum, according to a 2021 report of the California Pregnancy Mortality Surveillance System.28

“Our feeling now is that structured protocols are really important for reducing maternal mortality, but you need more than that if you are really going to close the gaps,” says Elliott Main, an obstetrician-gynecologist at Stanford University and medical director of the California Maternal Quality Care Collaborative, which has created a birth equity collaborative to focus on equity and best practices.


Late one afternoon in August 2021, Penn team members from all five of the system’s hospitals with labor and delivery services signed into a monthly Zoom call to share their progress on the team goal. They each took a few minutes to share the work of their hospital: an improvement in hemorrhage response, an updated cart of emergency supplies, revamped coding to make the data more accurate, revised labor induction and anemia protocols.

For months the teams had dissected their data, and now they had positive results to celebrate. For example, Kelly LaMonica, obstetric safety and quality specialist at Princeton Medical Center, in New Jersey, used a monthly dashboard to raise awareness about serious maternal complications at her hospital’s obstetrics departmental meetings. She was able to demonstrate the benefits of changing practice, using strategies shared among the systemwide team. “It’s just been great having the support of all the Penn entities,” LaMonica tells her colleagues.

Getting buy-in from private providers, including obstetricians, family practice doctors, and nurse midwives, also was a key factor in making changes, says Alyssa Livengood Waite, director of nursing at Lancaster General Health. “There’s a lot of energy around this topic, so there’s really engaged participation,” she says.

Beyond the data and protocols, Penn Medicine took other steps to advance health equity.

Beyond the data and protocols, Penn Medicine took other steps to advance health equity. The Department of Obstetrics and Gynecology created a strategic plan for diversity, equity, and inclusion and a multidisciplinary health equity group. Open to physicians, nurses, and any interested staff members, the group addresses concerns about health inequities in maternal, child, and gynecologic health at the Hospital of the University of Pennsylvania.

An equity-based survey conducted by Hospital of the University of Pennsylvania team leaders working on the Penn Medicine goal revealed discomfort about the common use of the acronym FOB for “father of baby,” instead of referring to a patient’s partner or husband. Was FOB used more often to refer to the partners of Black women? Was the term disrespectful? The discussions led the department to stop use of the acronym. After all, equity requires rethinking care from the patient’s perspective, says Abike James, an obstetrician and vice chair for diversity, equity, and inclusion in Penn Medicine’s Department of Obstetrics and Gynecology. “We started looking at different policies that could contribute to any feeling of mistrust,” she says.

In fiscal year 2022 Penn Medicine expanded its maternal health equity team goal to encompass Black, indigenous, and people of color, known as BIPOC. Chester County Hospital, in West Chester, Pennsylvania, had already begun reaching out to Hispanic women, interviewing them to learn more about their childbirth experiences. “That will help us get a better understanding of where we might have avenues for improvement,” says Fran Doyle, director of maternal child services at the hospital.

Mahoney is proud of the progress, which occurred even as the Penn Medicine staff met the challenges of a pandemic and the health system opened the Pavilion, a 1.5-million-square-foot facility at the Hospital of the University of Pennsylvania, one of the largest hospital projects in the US. He remains committed to the team goal, which he says is likely to continue through at least fiscal year 2023. “We want to get it to a point where it’s hardwired in and [the health equity work] is happening regardless of whether it’s in the printed goals,” he says.

‘The Best Thing In The World’

When Talicia Williams woke in the delivery room, she began swatting the air around her face. She had the illusion that tiny gnats were flying around her. A nurse explained that the phenomenon was likely related to low blood pressure from her blood loss. She received a blood transfusion as well as blood clotting factors and intravenous iron to raise her hemoglobin.

Yet she didn’t feel as weak as the nurses seemed to expect. Finally, they placed her daughter, Zahiri, swaddled and sleeping, in her arms. Williams was struck by how much the baby looked like her older brother. “It was the best thing in the world,” she recalls. “I couldn’t believe I had a girl.”

Williams spent four days in the hospital, but when she went home, those scary moments soon receded. Now when she tells her childbirth story, she stresses the happy ending. “I just want to let others know what I went through,” she says. But her message to other new moms is reassuring: “Don’t be scared. The nurses will make sure nothing worse happens.”



Leave a Reply

Your email address will not be published.