The following story diagram—or Storygram—annotates an award-winning story to shed light on what makes some of the best science writing so outstanding. The Storygram series is a joint project of The Open Notebook and the Council for the Advancement of Science Writing. It is supported in part by a grant from the Gordon and Betty Moore Foundation. This Storygram is co-published at the CASW Showcase.
We all know—or need to know—that race intersects with every facet of American life, from the mundane to the momentous. Where you sleep at night, what you eat, where you send your kids to school, who you’re friends with: Whether you realize it or not, decades of racial segregation and inequitable social design have played an often-invisible hand in shaping the options available to you. Options are at the heart of Annie Waldman’s data-driven investigation into maternal harm at hospitals that disproportionately serve black mothers. When a black woman in America goes into labor, which hospital she chooses for her delivery—or where the ambulance takes her, sometimes against her wishes—can determine whether she and her baby will leave the hospital together and alive.
Waldman’s reporting, part of ProPublica’s award-winning Lost Mothers series, makes the invisible visible, and it does so with data. Since race, class, geography, and poverty are so deeply intertwined in America, it can be hard to tell where one strand ends and another begins. That’s why ProPublica started with a large dataset—67,000 cases of women who experienced serious complications in the delivery room—from three diverse states: New York, Florida, and Illinois. Waldman’s team identified hospitals in those states that are “black-serving,” meaning higher proportions of their maternal patients are black. Even when accounting for factors like education, income, and overall health, they found that women who deliver at “black-serving” hospitals experience higher rates of serious complications. According to their analysis, “black mothers who are college-educated fare worse than women of all other races who never finished high school. Obese women of all races do better than black women who are of normal weight. And black women in the wealthiest neighborhoods do worse than white, Hispanic and Asian mothers in the poorest ones.”
Of course, each data point represents someone’s reality: a mother-to-be whose long-anticipated delivery turned into a preventable nightmare; a family hollowed by grief; a child denied the chance to grow up with a healthy mother—or at all. The best data journalists spend as much time encouraging people to speak as they do staring at spreadsheets. Waldman spent weeks traveling to meet women who experienced the high-risk complications seen in this dataset, women who survived but lost their babies, or the loved ones left behind when women died. Their stories and their questions—Why did this happen? Why didn’t anyone stop it? Why does this keep happening?—take this problem out of the realm of academic journals and, hopefully, into the realm of public outrage and prevention.
Waldman’s investigative report expertly weaves numbers and testimony, and shares a methodology for other reporters across the country to replicate her results and identify harmful hospitals in their own backyards. It also points to field-tested delivery room protocols that are proven to work in high-risk births, and could be adopted by more hospitals. Finally, Waldman counters the notion that women who are living with poverty, lower rates of education, and pre-existing health complications should not have the same access to safe, healthy maternal care as women with greater resources.
I’ve spent much of my own journalism career digging into huge, complex datasets and stories about race in America. It was an honor to delve deeply into Waldman’s award-winning story and reverse-engineer her reporting through my annotations. I hope to apply her methodology and metrics to my adopted hometown in Smith County, East Texas—a region with alarming and disproportionate rates of maternal harm among black women. Waldman has provided a blueprint for reporters like me to help make this problem clearly visible across America.
Story Annotation
“How Hospitals Are Failing Black Mothers”
A ProPublica analysis shows that women who deliver at hospitals that disproportionately serve black mothers are at a higher risk of harm.
By Annie Waldman, ProPublica
Published December 27, 2017
(This article was originally published by ProPublica and is republished here under a Creative Commons license.)
NEW YORK — When Dacheca Fleurimond decided to give birth at SUNY Downstate Medical Center earlier this year, her sister tried to talk her out of it.So much of this story is about agency, and what women who arrive at hospitals to give birth ought to be able to expect about their time there. I appreciate that this story explores what happened to Dacheca Fleurimond, who intentionally opted to give birth at a hospital that meant something special to her and her family, despite a loved one trying to talk her out of it. By choosing to start her investigative report with Fleurimond’s case, Waldman avoids the trap of delivering up the “perfect” victim, someone who seemingly “did everything right” and was beset by tragedy despite their unimpeachable status. No woman should have to do things “perfectly” to receive proper healthcare.
Her sister had recently delivered at a better-rated hospital in Brooklyn’s gentrified Park Slope neighborhood and urged Fleurimond, a 33-year-old home health aide, to do the same.
But Fleurimond had given birth to all five of her other children at the state-run SUNY Downstate and never had a bad experience. She and her family had lived steps away from the hospital in East Flatbush when they emigrated from Haiti years ago. She knew the nurses at SUNY Downstate, she told her sister. She felt comfortable there.In our interview, Waldman spoke of spending weeks gaining the trust of Dacheca Fleurimond’s loved ones and helping them feel comfortable enough to share her story. Details like these, which force us to understand and empathize with Fleurimond’s decision, are why.
She didn’t know then how much rode on her decision, or how fraught with risk her delivery would turn out to be.
It’s been long-established that black women like Fleurimond fare worse in pregnancy and childbirth, dying at a rate more than triple that of white mothers. And while part of the disparity can be attributed to factors like poverty and inadequate access to health care, there is growing evidence that points to the quality of care at hospitals where a disproportionate number of black women deliver, which are often in neighborhoods disadvantaged by segregation.This is the heart of the challenge that Waldman and her team had to address with this piece. Medical experts and laypeople alike will try and explain what happened to women like Fleurimond, and countless others represented in the data Waldman analyzed, as problems of class, our universally broken health care system, or related health epidemics. But race, and racial segregation, are squarely at the heart of what’s really happening here. It’s important to unflinchingly say so—and, further down, to spell out how the data reveal that fact, even if it means getting more “in the weeds” than your reader expects.
Researchers have found that women who deliver at these so-called “black-serving” hospitals are more likely to have serious complications — from infections to birth-related embolisms to emergency hysterectomies — than mothers who deliver at institutions that serve fewer black women.
Still, it’s difficult to tell from studies alone how this pattern plays out in real life. The hospitals are never named. The women behind the numbers are faceless, the specific ways their hospitals may have failed them unknown.I think this piece serves as a much-needed supplement to the growing body of medical research on race-based disparities in maternal and infant health outcomes. Without the work of the researchers Waldman mentions in the previous graf, it would be much harder to establish the role of race in this problem. But telling the stories of real women—and giving the loved ones they left behind a chance to share their own testimony—takes this issue out of the realm of academic journals and into public awareness.
ProPublica did its own analysis, using two years of hospital inpatient discharge data from New York, Illinois and Florida to look in-depth at how well different facilities treat women who experience one particular problem — hemorrhages — while giving birth.
We, too, found the same broad pattern identified in previous studies — that women who hemorrhage at disproportionately black-serving hospitals are far more likely to wind up with severe complications, from hysterectomies, which are more directly related to hemorrhage, to pulmonary embolisms, which can be indirectly related. When we looked at data for only the most healthy women, and for white women at black-serving hospitals, the pattern persisted.Another instance of Waldman and her editors anticipating the questions—and even the skepticism—of some readers, and getting ahead of it through smart data analysis. Truly effective data reporting is knowing not just how to crunch numbers, but what questions to ask.
Beyond this bird’s-eye view, our analysis allowed us to identify individual hospitals with higher complication rates, to look at what kinds of protocols they have and to examine what went wrong in specific cases.
We found, for example, that SUNY Downstate, where 90 percent of the women who give birth are black, has one of the highest complication rates for hemorrhage across all three states. On average, 34 percent of women who hemorrhage while giving birth at New York hospitals experience significant complications. At SUNY Downstate, it’s 62 percent.
SUNY Downstate officials defended the hospital’s handling of obstetric hemorrhages, saying it has extensive protocols for responding to them and gets exemplary results despite handling deliveries involving mothers with higher-than-average numbers of health problems like diabetes, obesity and high blood pressure. They would not comment on Fleurimond’s case, citing patient privacy.
Fleurimond was admitted to Downstate on Aug. 9.
Pregnant with twins, her doctor noticed she was in preterm labor at her 34-week checkup and prepped her for an unplanned cesarean section. When they cut into her womb to deliver the babies, Fleurimond’s uterus didn’t fully contract as it should have. She began to bleed. By the time the doctors controlled the hemorrhage, she had lost more than a liter of blood, requiring two transfusions.Here and elsewhere, I appreciate the straightforward nature of medical descriptions. It’s an old editorial rule-of-thumb: The more intrinsically dramatic the scene you’re describing, the more plainly and simply you’re required to tell it. It would be deeply disrespectful to Dacheca Fleurimond and her loved ones to describe what happened to her any other way. It would also be a disservice to readers to gloss over the details in case of squeamishness. For most women, childbirth is a medical procedure, and we all need to get far more comfortable thinking and talking about it that way.
At first, it seemed she’d be fine. She awoke the following morning thinking the worst was over, eager to see her new sons.
She wouldn’t survive the day.
Every year in the United States, between 700 and 900 women die from causes related to pregnancy and childbirth. For every woman that dies, dozens more experience severe complications, which affect more than 50,000 women annually.In my own work as a data editor on teams reporting on gun violence, I learned the importance of talking about survivors of violence and trauma, not just those who lose their lives to it. It’s not just about the enduring pain and suffering of wounded survivors; it’s also about the lasting toll on their careers, families, communities, and even the wider economy. When we talk about violence and trauma only in terms of death, we miss most of the picture. That’s why I was so glad to see Waldman include statistics on survivors of severely complicated deliveries.
The U.S. rate of maternal mortality is substantially higher than those of other affluent nations and has risen over the past decade. Outcomes for black women have led the way, intensifying efforts by medical experts and academics to understand what’s driving the racial disparity.
A complicating factor in understanding how hospital care figures in is that hospitals take on different proportions of tough cases — patients who have less access to consistent, quality prenatal care or have chronic health issues, like diabetes or heart disease, that make pregnancy and childbirth riskier.
Some prominent researchers are using a methodology for analyzing birth outcomes that attempts to even the playing field.So very, very smart to tap an unaffiliated third-party research organization and apply its field-tested methodology to ProPublica’s dataset. As data journalists, it can be tempting to “go it alone” and resist the use of third-party analyses. By adopting and applying another organization’s research algorithm, ProPublica is further reinforcing the validity of its investigative report, and encouraging best practices in the wider field of data-driven journalism.
The California Maternal Quality Care Collaborative, which studies maternal deaths and develops techniques to prevent them, looks at how well hospitals respond to obstetric hemorrhage, typically defined as losing more than 500 milliliters of blood during a vaginal birth or a liter of blood during a cesarean section. Why hemorrhages? Because women of all races experience them at roughly the same rates and their likelihood is less affected by factors like race or economic status, said CMQCC medical director Dr. Elliott Main.Another benefit of relying on this outside research organization’s methodology: It helps narrow the focus of this story to a single complication, which happens to occur at alarmingly high rates at one particular hospital (SUNY Downstate), and happened to be the reason Dacheca Fleurimond, a black woman delivering her child at SUNY Downstate, lost her life. Of course, none of this is coincidental, and CMQCC’s methodology undoubtedly informed Waldman’s decision to focus on SUNY Downstate in the first place. When she introduces us to this research organization here, it further legitimizes, in the reader’s mind, all of the editorial choices she made higher up in the piece.
CMQCC evaluates hospitals by calculating what percent of women who hemorrhage during birth wind up with major complications. Researchers count both the complications more directly related to hemorrhages, like hysterectomies and blood transfusions, and those that could be indirectly related, including embolisms, blood clots, heart attacks, kidney failure, respiratory distress, aneurysms, brain bleeds, sepsis and shock. Ultimately, this approach measures how often doctors prevent complications when a hemorrhage occurs, and when looked at over time, can show if a hospital has been able to improve.
ProPublica used the metric to analyze inpatient hospital discharge data collected by New York, Illinois and Florida for 2014 and 2015, examining all obstetric cases that were coded as involving hemorrhages — about 67,000 cases in all.Data journalism is all about making choices, and opting to look at two years of data from three big, diverse states gives Waldman both a large number of cases to examine and also a statistically viable number of cases involving black women.
We also put each hospital into a category based on the concentration of black mothers who gave birth there, defining facilities as low, medium or high black-serving. We crafted our analysis so that it reflected the racial distribution of mothers delivering in each state. In New York, if black mothers represented roughly a third or more of the deliveries at a hospital, we considered the hospital high black-serving. In Florida, we considered a hospital high black-serving if about 40 percent of the mothers were black. In Illinois, we considered a hospital high black-serving if at least half of its mothers were black.Speaking of choices, I was so excited to read this transparent, detailed, process-oriented graf. Editors sometimes shy away from spelling this sort of stuff out to readers, worrying it’ll “bog the reader down” and lose them “in the weeds.” I totally disagree. By giving the reader the opportunity to explore our processes and logic, we build trust, and we build data literacy.
In New York, we defined a hospital as low black-serving if less than eight percent of the women delivering there were black. In Illinois, the cutoff was 14 percent. In Florida, it was 18 percent.
Across the three states, about one in 10 hospitals in our analysis was high black-serving — in some cases, extremely high. Ninety-nine percent of the mothers who gave birth at Jackson Park Hospital and Medical Center in Chicago were black.
While a handful of low black-serving hospitals had high complication rates, our analysis found that, on average, outcomes at hospitals that served a high number of black patients were far worse.
In New York, on average, high black-serving hospitals had complication rates 21 percent higher than low black-serving hospitals. In Illinois and Florida, high black-serving hospitals had complication rates 11 percent higher.
When we limited our patient pool to only mothers of average birthing age — between 25 and 32 — who did not have any chronic conditions like heart disease or diabetes, the pattern remained largely the same. This bolstered the notion that differences in care, along with patient characteristics, affected outcomes.
Deeper analysis of the data for each state underlined this finding. At low black-serving hospitals in New York, just under a third of the women who hemorrhaged had complications. At high black-serving hospitals, that rate climbed to about half.
Dr. Elizabeth Howell, a professor of obstetrics and gynecology at the Icahn School of Medicine at Mount Sinai Hospital, has taken a more refined look at racial disparities among New York City’s hospitals. She found black mothers were twice as likely to suffer harm when delivering babies than white mothers, even after adjusting for patients’ differing characteristics, suggesting that some of the racial disparity may be due to hospital quality. In a separate study, she estimated that the rate of harm for black women would fall by nearly 50 percent if they gave birth at the same hospitals as white women.
She’s also considered the same dynamic nationally. Because three quarters of black mothers deliver in about a quarter of the country’s hospitals, Howell believes that racial disparities could be reduced if hospitals that disproportionately serve black women improved their care.
There is clear evidence hospitals can make such improvements.As a big fan of the work of the Solutions Journalism Network, which empowers journalists to “tell the rest of the story” by highlighting what works, not just what’s broken, I love that Waldman dug into field-tested protocols that are proven to work. I also love that a few paragraphs down, she asks administrators at problematic “black-serving” hospitals why they aren’t adopting these protocols.
In California, complications related to obstetric hemorrhage decreased by about 20 percent in hospitals that adopted protocols promoted by Main’s group, which include keeping carts stocked with supplies to stave off massive bleeding and holding drills to simulate severe hemorrhage events. “It creates improvement in the team, increases communication and improves your response to all emergencies,” Main said.
Still, Main’s protocols haven’t been universally adopted in California, let alone elsewhere in the U.S., and many hospitals go their own way.
The spokesperson for SUNY Downstate — where more than 14 percent of women hemorrhage during birth, an average of one mother every other day — said the hospital “has already developed their own ‘best practice’ protocols for hemorrhage that other hospitals should be following.” These include a special “Code Mom” that details steps doctors and nurses need to take when responding to a hemorrhage. And women with placental problems are monitored by ultrasound, so that doctors can anticipate the most complex cases before beginning cesarean surgeries.
According to public documents posted in an online repository of the hospital’s policies, the obstetric and gynecology department’s emergency response policy on hemorrhage does not explicitly follow some of Main’s recommendations, such as having pre-fab kits to respond to hemorrhages and doing staff drills to prepare for them. SUNY Downstate did not respond to questions about these differences.
Dr. Ovadia Abulafia, the chair of the hospital’s department of obstetrics and gynecology, noted that SUNY Downstate serves a particularly “underserved” and “high-risk” population. More than 80 percent of women who deliver there are obese, a spokesperson said, and the hospital sees a higher incidence of diabetes, blood pressure disorders and placental separation problems compared to the rest of the nation.
But Dr. Allison Bryant Mantha, a high-risk obstetrician and health care disparities researcher at Massachusetts General Hospital, said hospitals shouldn’t use demographics or patient characteristics to excuse poor outcomes.Including the voice of this researcher helps elevate this issue out of the defensive back-and-forth posturing of hospital administrators and into the realm of what it means to truly care for every patient—meeting them “where they’re at.” Instead, they should hone their practices to deliver the care their patients need.
“Hospitals have to own the conditions that women walk in with,” Bryant said. “You have to give patients what they need to get to a quality level of care. We are doing a good job of equal care, but not adjusting for needs.”
Fleurimond awoke in good spirits in the labor and delivery unit on Aug. 10, the day after her delivery. Her biggest concern that afternoon was what she was going to eat. “What is Jell-O going to do for me?” she complained to her sister Merline Lamy, who responded, “This is your two-day diet, baby girl.” Fleurimond rolled her eyes.
She might not have felt it at the moment, but Fleurimond was still at risk of serious complications related to her hemorrhage, including pulmonary embolism, typically caused when a blood clot travels from a patient’s leg to a lung artery, blocking blood flow to the lungs.
Her blood was already predisposed to clotting, a biological mechanism that likely evolved in pregnant women to prevent hemorrhage during birth. Carrying twins can put extra pressure on the vessels around the uterus, further constricting blood flow. The cesarean surgery, like all surgeries, substantially increased her risk, as did the transfusions.
On top of that, Fleurimond weighed 260 pounds and was being treated for high blood pressure.
To prevent clotting, nurses had put compression boots on her legs. Just after 3 p.m., according to family members who were visiting Fleurimond, a nurse unfastened the boots, helped Fleurimond into a wheelchair and took her to visit the twins, Jayden and Kayden, in the neonatal intensive care unit. She’d held them only briefly in the operating room and craved another look. They had her round cheeks, which shone like polished apples.
Experts say compression boots lose their deterrent effect about 15 minutes after they are removed. Fleurimond spent about 90 minutes in the NICU with her aunt, who recalled her sitting in her wheelchair the whole time, her legs hanging down. Shortly after her aunt left, she complained that she felt unwell, but three hospital employees who spoke to ProPublica on the condition of anonymity say that she waited at least 40 minutes for a transport aide to wheel her back to her room. There is no evidence in her medical record that anyone came to assess her when she returned.
Doctors also did not prescribe heparin, a blood-thinning medicine being used at other hospitals to prevent pulmonary embolism in mothers with high risk factors, for whom compression boots are unlikely to be enough.
In the United Kingdom, protocols that advocate more aggressive use of blood thinners, particularly after C-sections, helped reduce embolism deaths by more than half within three years.
In the United States, a chorus of medical trade groups and maternal safety organizations have begun to promote more widespread use of blood thinners during pregnancy and childbirth, but not all hospitals have made it their practice.
“There are some experts who feel that it’s not worth the time, trouble and cost to avoid relatively rare events,” said Dr. Alexander Friedman, an assistant professor of obstetrics and gynecology at Columbia University Medical Center.
Friedman’s hospital on the edge of Harlem typically administers the drug to high-risk mothers, but Fleurimond wouldn’t have had to travel that far. Three miles away from Downstate, at a Brooklyn hospital that has a smaller concentration of black patients and a lower complication rate related to hemorrhages, Maimonides Medical Center gives blood thinners to nearly all of mothers who undergo cesarean sections or have other risk factors.Here and elsewhere, we are faced with the infuriating, overlapping, and competing impacts that geography, race, and healthcare access have on the lives of women like Fleurimond (and women unlike her). Details like this—after the stoic retelling of all of the medical decisions and missed opportunities that occurred in Fleurimond’s case up to this point—make it impossible to pretend there’s not a problem here.
Friedman, who reviewed Fleurimond’s medical records at ProPublica’s request, said she should’ve received the drug.
Dr. Douglas Montgomery, an obstetrician-gynecologist and director of the Maternal Fetal Medicine Department at California’s Kaiser Permanente Riverside Medical Center, said he would prescribe the drug to any patient who had Fleurimond’s risk factors.
At around 6 p.m., Fleurimond called the father of her twins. She sounded short of breath. She said she was in pain and asked him to come to the hospital, then hung up and waited, alone.
At about 6:25 p.m., Fleurimond screamed, medical records show. A doctor and nurse entered her room and found her gasping for air. More responders came. They couldn’t find a pulse. After more than an hour of resuscitation attempts, she was pronounced dead at 7:45 p.m.
Because Fleurimond died “during diagnostic or therapeutic procedures or from complications of such procedures,” as Downstate’s website puts it, she was referred to the New York City medical examiner’s office for an autopsy. Her cause of death, according to the autopsy report: pulmonary embolism, also known as “venous thromboembolism,” a condition that almost always has a chance of being prevented.So much going on in this paragraph: medical jargon juxtaposed with the heart-wrenching details of Fleurimond’s last moments, and the sharp use of impartial third-party sourcing to further call the hospital’s actions into serious question.
If you’re enjoying this Storygram, also check out two resources that partly inspired this project: the Nieman Storyboard‘s Annotation Tuesday! series and Holly Stocking’s The New York Times Reader: Science & Technology.
In an emailed statement, Abulafia said SUNY Downstate “follows the proven [American College of Obstetricians and Gynecologists] protocols for obstetric hemorrhage, severe hypertension and venous thromboembolism.” SUNY Downstate has not had a maternal death related to hemorrhage in the past 15 years, a spokesperson said.
Such assurances provide little solace to Fleurimond’s relatives, who have sought an attorney to represent them.
“Dacheca Fleurimond was clearly at high risk to have a blood clot and there weren’t adequate preventative measures,” said the attorney, Eleni Coffinas. “The obesity, the hypertension, and the fact that she hemorrhaged after her C-section were all high-risk factors and she needed to be monitored for that.”
New York City occupies a unique place in the discussion of racial disparities in maternal mortality as both a hub of groundbreaking research on the subject and one of the nation’s starkest examples of such gaps.Echoes of the ironic, unsustainable, ever-widening imbalances we are seeing in nearly all of the nation’s wealthiest metropolises.
In addition to the work by Howell, the New York City Department of Health and Mental Hygiene has published a couple of reports, including one documenting how, as the mortality rate of expectant and new mothers overall across the city has dropped, the disparity between black and white mothers has grown.This piece does such a great job of relying on a diverse array of information sources: in-house data analysis, academic and institutional research, medical experts, and primary testimony from those who experienced harm (and who probably would never have shared their stories publicly otherwise).
Even when accounting for risk factors like low educational attainment, obesity and neighborhood poverty level, the city’s black mothers still face significantly higher rates of harm, the agency found. Of note, black mothers who are college-educated fare worse than women of all other races who never finished high school. Obese women of all races do better than black women who are of normal weight. And black women in the wealthiest neighborhoods do worse than white, Hispanic and Asian mothers in the poorest ones.So powerful and compelling. We can’t educate, gentrify, or exercise our way out of this health disparity.
The health department has even mapped where the most maternal harm occurs, dividing the city into community districts. The highest rates of complications are concentrated in a swath of land in Central Brooklyn, in an area largely untouched by the wave of gentrification that has swept other parts of the borough. Here, mothers face up to four times the complication rates of neighborhoods just a few subway stops away. Fleurimond lived in one such danger zone, in a public housing development in eastern Crown Heights.
At three medical centers in this area that deliver babies — Brookdale University Hospital Medical Center, Kings County Hospital and SUNY Downstate — more than half of mothers who hemorrhaged during delivery experienced complications, ProPublica’s data analysis shows. More than three quarters of the women who give birth at Brookdale are black, as are nearly 90 percent of the women who deliver at Kings County Hospital.
Officials at Brookdale, a private nonprofit hospital, would not respond to questions from ProPublica. The New York City Health + Hospitals Corporation, the public benefit organization that operates Kings County Hospital, gave a detailed response laying out its protocols for obstetric hemorrhages, including some recommended by Main’s group. Robert de Luna, a spokesperson for the city’s hospital operator, said in an email that while hemorrhage is a good proxy indicator for maternal harm, “some of our patients come from all over the world (self-referred), a good number coming to us too late to benefit from our prenatal care services.” (Read the full response here.)
Some of the women who deliver at these hospitals are well aware of their reputations.
Brookdale, for example, was recently rated an ‘F’ by Leapfrog, the health care quality and safety nonprofit, one of only 15 hospitals in the country to receive a failing grade.
But proximity sometimes takes precedence over choice. That was the case for Merowe Nubyahn, a 37-year-old hospice aide.
In March 2013, when Nubyahn was 24 weeks pregnant, she was overcome with intense nausea and vomiting, and unexpectedly, her water broke. When emergency medical technicians arrived at her East New York apartment, she begged them to take her anywhere but Brookdale.I appreciated that Waldman chose to lead this report with the story of Dacheca Fleurimond, a woman who opted to deliver her baby at SUNY Downstate because that was where she felt most comfortable. I also appreciate the inclusion of Merowe Nubyahn, who fought hard not to deliver at a “black-serving” hospital. By digging into both of these stories, Waldman’s reporting makes it clear that solutions—and accountability—can’t be found at the level of individual choice. She hadn’t liked what she had heard about the hospital and had been getting her prenatal care elsewhere. The ambulance took her to Brookdale anyway because it was closest.
At the hospital, she was rushed in for a cesarean section. Her daughter, delivered at what’s considered the edge of viability, barely clung to life in the hospital’s NICU. When Nubyahn awoke in the recovery room, layers of gauze covered her belly and her throat felt like sandpaper. Disoriented, she said she asked a nurse what had happened, but the words felt garbled leaving her mouth. Two of her teeth had been knocked out when she was intubated for anesthesia, according to her medical records. Nubyahn recalled that when she asked the doctor about them, he gave her an incredulous look and asked, “Are you sure you had teeth when you came in here?”I have nothing smart to say about this quote, other than how incredibly sad, scary, and sickening it was to read.
A bigger threat to her health emerged the morning after she was discharged from the hospital. As she sat in bed, she says she felt sharp cramping pains and a warm, viscous feeling. She looked down at her belly and saw dark, clotted blood — “plums and prunes” — bursting out of her cesarean incision.
Her wound had become infected — a common complication — and had begun to come apart. Still wearing her hospital bracelet, she was shuttled back to Brookdale and told she’d also developed a hematoma, a mass of blood, around her incision site.
While Nubyahn was being treated in one part of the hospital for her various complications, her baby died in another. Overcome with grief and stung by her treatment, Nubyahn checked herself out and vowed to never return. “All the horror stories that I have heard about Brookdale … I totally have my own now,” she said.
Khari Edwards, the vice president of external affairs at Brookdale, said the hospital would not comment on Nubyahn’s case due to privacy laws.
Recognizing that hospitals in Central Brooklyn have some of the highest maternal complication rates in the city, the health department has begun to target the area with services in recent years. It supports the By My Side initiative that pairs up women with doulas who can advocate for them during birth. The department also supports prenatal programs in the area based on a model of assessment, education and support, also known as CenteringPregnancy.
“We are data driven and we look to where the outcomes are the worst,” said Dr. Deborah Kaplan, the assistant commissioner for maternal, infant and reproductive health at the department.
This month, the city convened a new committee to review deaths and severe complications related to pregnancy and childbirth. One of its priorities will be to figure out why — despite years of research and attention — the city’s racial disparities have persisted and even grown.
“We used to say we are not sure why we are seeing these racial disparities. Now we say unequivocally that racism causes these problems,” said Kaplan. She emphasized that this encompassed not only health care but all aspects of life in the city, from housing to schools. “If we provide equally to everyone, we could widen the inequity. We have to prioritize putting resources in neighborhoods with the highest rates of severe maternal morbidity and the least access.”
Just three months after Fleurimond died at SUNY Downstate, another black woman died there, hours after giving birth.
Tanesia Walker, a 31-year-old flight attendant, had originally planned to deliver at New York-Presbyterian Brooklyn Methodist Hospital in Park Slope, where she had her first son. But a week before her scheduled cesarean, her doctor changed her delivery to Downstate, where he also had privileges.
Walker grew nervous after reading negative reviews of the hospital online, family members said. They tried to calm her down. Hers was not a high-risk pregnancy, they reminded her. She wasn’t overweight and her blood pressure was fine.
Walker seemed okay after a C-section at SUNY Downstate on Nov. 27, holding her newborn son Tyre close to her chest as her family spent the afternoon with her. Then, at 2 a.m., she sent a text message to her fiancé saying she had a pain in her side, he told ProPublica.
A few hours later, she was dead.
As family members trickled into the hospital that morning, shocked and confused, doctors couldn’t say why she died, said her father Junior Walker. They mentioned the possibility of blood clots in her lungs, he added.
The family has requested her medical records from the hospital. As with Fleurimond, the city medical examiner’s office has done an autopsy, but has not yet released its report.
Walker’s death haunts her younger brother Dwayne, who kept in touch with her nonstop as she traveled. He can’t stop thinking, why her? She was educated, had a criminal justice degree from John Jay College. She was healthy, didn’t drink or smoke, ran track in high school. She was financially stable, quit a management job at Chase Bank to see the world aboard American Airlines.By presenting us with three very different stories of three very different women—and corresponding data analysis—Waldman’s reporting helps us understand that factors like education, lifestyle, and income can’t necessarily inoculate black women from serious, and even deadly, childbirth complications. This reporting loudly but respectfully offers an answer to this anguished question: Why her? It’s because Tanesia Walker was black, and because she was forced to deliver her baby in a hospital that serves black women.
“I just want to know why she died,” he said, eyes wet with tears. He keeps sending her text messages, even now that she is gone. “She was a healthy woman who shouldn’t have died from a cesarean section.”
Fleurimond’s family is doing its best to survive without her.
Her sister, Merline Lamy, took in Fleurimond’s six youngest children, blending them into her own household, but that meant squeezing 12 people into a three-bedroom apartment. The landlord threatened to evict them.
Fleurimond’s brother and his wife have tried to collect money for the children on GoFundMe, but so far have only raised about $250. (ProPublica reporter Nina Martin, who was not involved in the reporting or preparation of this story, donated $100 three months ago.)
Fleurimond’s 58-year-old mother has become the principal surrogate parent — changing diapers, cooking dinners and breaking up sibling spats. She sleeps no more than a couple of hours each night, her eyes permanently rimmed with dark shadows.
The kids, too, are struggling to settle into their new life.
On a recent evening, Joshua, 9, tried to tune out the noise in Lamy’s packed apartment and concentrate on his math homework. Berlynda, 10, comforted a twin in each arm. Aiden, 2, climbed on the couch with a runny nose.
Like all toddlers, his mood teeters between buoyancy and despair. But when he calls for “mama,” his siblings have to remind him she will not come.It was imperative to close this report with scenes of traumatized families left to pick up the pieces: children raising children, grandparents denied their earned stage of life. Truly effective investigative journalism shows us not only what’s broken but also how it could be fixed if we decided it was important enough—and whose lives could be transformed if we did.
ProPublica Illinois reporters Duaa Eldeib and Jerrel Floyd contributed to this report.
ProPublica is a Pulitzer Prize-winning investigative newsroom. Sign up for The Big Story newsletter to receive stories like this one in your inbox.
A Conversation with Annie Waldman
Tasneem Raja: Let’s start at the beginning. What made you decide to take on this story, and why now?
Annie Waldman: ProPublica had been working on what we called our Lost Mothers series for several months at the point that I came on to the project. My expertise is really looking at data, analyzing huge data sets that normally are very unwieldy for traditional reporters, and integrating those data sets into stories. I was brought on about halfway through the project in order to provide a deeper data understanding of maternal mortality and how hospitals played a role in the deaths of mothers.
Earlier on, we had started collecting data from a number of different hospitals across the country. The idea was to look at the data and start to see what kinds of patterns we could find. We had a theory, of course, before we started diving into the data, that if you look at hospitals that serve a high percentage of mothers who are mothers of color, you might see higher complication rates. After we analyzed the data, that’s exactly what we found.
There have been a number of stories looking at maternal mortality outside the United States, specifically in less industrialized countries or less developed countries. That lens hadn’t been put on the United States, despite the fact that if you look at the statistics, they’re quite troubling.
In the U.S., we have an incredibly high complication rate, and an incredibly high number of women who die in childbirth. Every single year, 700 to 900 women die from causes related to pregnancy and childbirth. There are significantly more who experience severe complications—about 50,000 women annually. In the U.S., there are significantly higher rates of maternal mortality than in other more industrialized or wealthier nations around the world. Not only that, but the rate of mortality has risen over the past decade. So this was a surprising statistic for us to look at.
Now, if you unpack that data, you see that black women are more affected by severe complications of childbirth and maternal deaths than women of other races. Not only journalists but also academics and other medical experts have really been trying to understand why this is the case. Why are the outcomes for black women leading the maternal mortality rate in the United States? That’s one of the drivers for why we wanted to look at this project.
TR: I want to come back to race and unpack that more. But taking a step back, you examined nearly 70,000 cases involving hemorrhages in Florida, New York, and Illinois in 2014 and 2015. Tell me why you focused on those states and those years. Was it simply a matter of what was available?
AW: When you choose data for a project, it’s always a combination of a number of things. For us, it’s what was available, but it’s also that we wanted to look at urban centers, or areas where there was a higher percentage of black mothers. When we looked at demographic data, it seemed that those three areas were going to provide us with enough data points that we wouldn’t be looking at small numbers and trying to extrapolate statements from that.
TR: I’m always interested in hearing more about the paper trail. Can you tell me exactly how you got access to the medical records you analyzed?
AW: New York State actually allows a lot of this data to be publicly available; it’s a dream of mine that all of this data could be publicly available across the country.
We were looking at in-patient discharge records of hospital stays. Most states collect this data as part of their billing practices. We wanted to understand whether we could use this data to also look at things like complications in maternal deaths. For hemorrhages, for example, how much blood is actually used in transfusions is an indicator.
We also looked at other data in this set including patient diagnoses, race, age, and whether a patient had high-risk characteristics—for example, diabetes or high blood pressure. Even though there are small differences in how each state records this data, we do believe that it is a comparable data set.
TR: You write that it’s long established that black women fare worse in pregnancy and childbirth, and that there is growing evidence that race and racial segregation are driving these disparities. For people who pay attention to these things, this really isn’t in dispute at this point.
But you also talk about the fact that individual hospitals aren’t named in studies, and so from those studies, we don’t get to learn or understand the specific stories of real women and the complications and outcomes they faced. Did you think of your story as serving as a supplement to the medical research that’s out there, a way of filling the gaps?
AW: That’s exactly right. There are incredible researchers who are out there describing these trends. But as a reporter, it’s not enough to just describe the fact that black women are three to four times more likely to die in childbirth than white mothers. We wanted to say what hospitals are contributing to these figures, and where they are. It’s not just about establishing that there is a problem. It’s about revealing where these problems are, and potentially what issues in those areas could be changed so there could be real impact and change.
That’s why, when we looked at these datasets, we wanted to find individual hospitals where these problems were very severe. That’s why we turned our eye to SUNY Downstate and two others in New York. We felt comfortable naming these hospitals as places where a lot of changes could be made to improve the safety and care of women, and particularly women of color.
TR: You introduce us at the start of your piece to Dacheca Fleurimond. She’s a black woman who delivered her twin boys at SUNY Downstate. She didn’t survive her stay. Tell me how you found out about her story, and why you decided to focus on her. I’m sure you learned about a lot of similar tragic cases. Why did this story stand out to you?
AW: I’m a self-professed data nerd, but it’s never enough just to have the numbers in your story. You have to have the humans behind it. Oftentimes, when you’re working with data, as a healthcare reporter or a science reporter, or even as an academic, you forget that each data point is a human. Behind each human there are stories and relationships and connections that should be reported on and told. There’s narrative there that could illustrate something for the larger audience.
We were already looking at SUNY Downstate as a hospital that had a high complication rate. I went down to the courthouse and looked at dozens of malpractice lawsuits against SUNY Downstate and some other hospitals that had high complication rates. We were trying to find individuals with stories that really illustrated what we were finding in the data.
Coincidentally, we actually had a whistleblower from SUNY Downstate who reached out to us, somebody on the nursing staff who said that they had witnessed a woman perish on the delivery ward. She thought it was due to failures of the hospital. We had reached out to a number of midwife organizations and doula organizations and asked them to put out emails to their colleagues, saying we’re reporting and we want more information on this. There was a huge social push with our reporting, trying to get people to contact us who had stories to share. So even though you could call it a coincidence, I don’t necessarily think it was. There were a lot of problems with this hospital, and it was a matter of making sure the right people understood that we were reporting on these issues.
We communicated with that whistleblower. I met with them a number of times, and was able to connect with other people who were part of the hospital and either were there the day Dacheca passed away, or knew other people who were there. From there, once I had her name, I reached out to Dacheca’s family members. It was an incredibly difficult process to gain their trust. That can be lost as well when we focus so much on data reporting. We can get these numbers easily by putting out public records requests, and we forget that there’s this whole other step of making sure that the people you’re reporting on, who are represented in these numbers, actually trust you when you show up at their doorstep. That takes a lot of time and a lot of honesty about what you’re doing, and transparency, to make sure that they understand that what you’re doing is trying to hold the place where this very serious and sensitive thing happened accountable.
I spent many weeks meeting with family members. Dacheca’s family was based in Brooklyn, but they also had family members in central Pennsylvania. I drove out there several times to gain their trust. I had been in touch with the seasoned malpractice lawyer who had taken the case on. I had to gain the trust of the lawyer as well. I didn’t know at first whether she was going to share medical records with me. It was a lot of listening and a lot of patience, as well as crunching those numbers.
TR: Dacheca Fleurimond consciously made the choice to deliver her baby at a so-called “black-serving” hospital, because she felt more comfortable there. But you also tell the story of Merowe Nubyahn, who begged not to be taken to Brookdale University Hospital Medical Center because she knew it had a bad reputation. She ended up dealing with absolutely nightmarish harm to her own body—and her baby died in the hospital not long after delivery. Was it important to you to include a story where a patient didn’t choose to have their baby at a particular hospital, and in fact actively fought against it?
AW: It’s easy to think, “Well, if these numbers are so bad, why would women keep going to these hospitals?” It almost takes agency away from the women, right? It’s assuming that they don’t look up the same numbers that other people do, and that’s absolutely not true.
There is a whisper network among women about where to deliver your babies and where not to. I think it’s important to illustrate that a lot of women—even if they know these hospitals aren’t necessarily going to provide them with the care they deserve—don’t necessarily have a choice, because of the way our healthcare system is set up. There are public hospitals in areas that, as with our residential system or our school system, are incredibly segregated. We don’t put enough care into hospitals that predominantly serve communities of color.
TR: At several points in your story, your analysis seems to almost be trying to err on the side of caution when it comes to identifying the extent to which race is a driving factor behind these higher rates of significant complications. For example, you limit your patient pool only to mothers who are of average birthing age. You note that black mothers in New York City who are college-educated fare worse than women of all other races who never finished high school. Were you anticipating pushback along those lines? Did you feel you had to cross every “t” to show that race really is at the heart of these disparities?
AW: Some people will want to explain away certain facts or trends that they see in the data through correlating characteristics. For example, if you look at patients who have diabetes, oftentimes you’ll see a higher rate in women of color, specifically African American women. So, a lot of people then try to tie that, as a correlating factor, to maternal complications.
I think it’s really important to address those factors, of course. But I talked with numerous people who work in hospital safety at a number of hospitals around the country, and in New York City, and they said that at the end of the day, these people are still showing up in your emergency rooms. These patients are still showing up to deliver their babies at your hospitals. The fact that they’re walking into your hospitals with complicating factors is not an excuse for poor care. If doctors are aware of the fact that a patient has characteristics that might put them in harm’s way more than other women of different races—like high blood pressure, diabetes, obesity, or anything like that—then doctors have to make sure that they’re doing things to protect women from those higher-risk factors.
That’s what was so shocking about what happened to Dacheca Fleurimond. They recognized that she was an incredibly high-risk patient. She ticked off a number of factors. Despite that, they did not provide her with the extra care that she not only required, but that she deserved. That’s why she had that embolism. In situations like that, you can run a regression with the data and you can explain it away by saying two things are correlated. But that doesn’t excuse the fact that, as a society, we have to make sure that we provide care for individuals with high-risk characteristics.
TR: I know from my own work in data reporting that states like New York, Illinois, and Florida tend to have rich and robust data to draw from. For someone who’s interested in doing a similar analysis in a less open state, a smaller locale, or a more rural region—for example, I report out of East Texas now—how would you recommend trying to tackle reporting on this issue?
AW: We provided a methodology for how we measured complications. The idea behind why we produce these methodologies at ProPublica is to provide a potential template for other reporters who are out there—whether national or local—to reproduce our work. It’s really important that we actually reveal the insides of our reporting and how we came to our conclusions.
In-patient data on hospital stays aren’t required only in the areas that we looked at; it is possible to get this data in other states because of Medicaid and Medicare laws. There is pressure from the federal government on state governments to provide this information so that people understand billing practices across the country. If you go to your state and you ask for in-patient discharge records of hospital stays, it is possible to get this data. Then, if you reach out to the California Maternal Quality Care Collaborative, it’s possible to get the metric that we used (which is hyperlinked in my methodology) and apply similar techniques in order to understand what the complication rate might be for a certain hospital in your state.
TR: Let’s talk timeline. How long did this reporting take you, and does the end result map neatly with the story and the analysis that you expected to undertake, or did you end up going in surprising directions?
Surprisingly, it was actually short—for us—meaning that sometimes we can spend over a year on a single investigative piece. This piece took me about three months, a relatively short amount of time for how much data analysis went into it, finding sources, confirming things, getting records, and following up with whistleblowers.
I did have a thesis going into this, but I didn’t expect to connect with so many actual cases of harm. Especially with maternal harm—but really with any kind of medical harm—people are frequently embarrassed about what happened to them, or feel that it was very private grief that they experienced.
When our social team was reaching out to women across the country on Facebook, Twitter, and other social platforms to try and get them to talk about the harm they had experienced, it was a very difficult task. Frequently, when you do an interview about a woman who has died in childbirth, it’s reliving that trauma for that family. When we started looking at this one hospital, SUNY Downstate, I was surprised by how many women actually felt comfortable coming out to talk, and how quickly they realized the importance of sharing their story for furthering the accountability in their case.
TR: You include context and insights from medical experts in OB/GYN and maternal health from all over the country. You must have anticipated that you weren’t going to get a lot from the top-level administrators and doctors at the hospitals you focus on in the piece. Did you feel it was especially important to assemble a squad of impartial outside experts for that reason?
AW: Exactly. I was assuming that most of the hospitals that we were looking at would probably not go on the record to talk to us. I knew it was important to have those outside experts, who could say, “This is what normally happens, this is what should happen, and this is what shouldn’t happen.”
As I was doing my data analysis, looking at the hospitals, and reaching out to families, I was also reading as many studies as I possibly could around the themes I was looking at—on pulmonary embolism, on hemorrhage and childbirth—and looking at the researchers who had already looked at similar numbers. Then I reached out to all of them and said, “Even if it’s off-the-record or on background, would you be willing to comment on this case I’m looking at?” I had the opportunity to share medical records of individuals that we profiled in our piece with our experts, so that I wasn’t misinterpreting anything.
Investigative journalists always have to be careful about confirmation bias. Our lens is to always be looking for harm, always looking for areas where we can provide accountability. We have to make sure that somebody else who’s objective, who’s not doing the project with us, can tell us, “Yes, that is indeed harm, and that’s something that shouldn’t be done.”
Annie Waldman is a reporter at ProPublica covering education. She graduated with honors from the Columbia Graduate School of Journalism and the School of International and Public Affairs at Columbia, where she was the recipient of the Pulitzer Traveling Fellowship and the Brown Institute Computational Journalism Award. Her stories have been published in The New York Times, The Atlantic, Vice, BBC News, The Chronicle of Higher Education, and Consumer Reports. Her documentary short film on the lives of homeless high school students after Hurricane Katrina appeared in the Sundance Film Festival and was later broadcast nationally on PBS. Her story “How Hospitals Are Failing Black Mothers” was part of ProPublica and NPR’s Lost Mothers series, which won a Peabody Award and a National Academies of Sciences, Engineering, and Medicine Communication Award. Follow her on Twitter at @anniewaldman.
Tasneem Raja is the executive editor of The Tyler Loop, a digital magazine that explores policy, history, and demographics in Tyler, Texas. She is an award-winning journalist who has reported for NPR, The New Yorker, The Atlantic, Mother Jones, and other national outlets. A former senior editor at NPR, she launched a popular podcast exploring issues of identity and race with NPR’s Code Switch team. At Mother Jones, she specialized in data journalism and led a team that built the first-ever database of mass shootings in America. She’s a pioneer in the field of data-driven digital storytelling, a frequent speaker on issues of digital journalism, and a die-hard fan of alt weeklies, where she got her start as a local reporter. She lives in Tyler with her husband, her stepson, and two imperious terriers. Follow her on Twitter @tasneemraja.
