ProPublica is a nonprofit newsroom that investigates abuses of power. Click here to read their biggest stories as soon as they’re published.
Lissbeth Hernandez holds her twins in the NICU at Tufts Medical Center in Boston on May 13. Hernandez became severely ill with COVID-19 and was intubated shortly before delivering. (Erin Clark/The Boston Globe via Getty Images)
In late June, after three months of near silence on the topic, the Centers for Disease Control and Prevention finally weighed in on a question of critical importance to millions of American women and families: How dangerous is the coronavirus for pregnant women and new mothers?
The CDC had been asserting that pregnant women don’t seem to be at higher risk for severe complications from the virus. As recently as late May, a spokesperson told ProPublica, “Current evidence shows pregnant women have the same risk of severe illness from COVID-19 as adults who are not pregnant.”
Then, the agency abruptly changed
its tone. In its first examination of U.S. data on COVID-19 in pregnancy, the CDC found
that expectant mothers with the virus had a 50% higher chance of being admitted to intensive care and a 70% higher chance of being intubated than nonpregnant women in their childbearing years.
Pregnant Latina and Black women were infected at significantly higher rates than white women, researchers reported. As of July 2, at least 30
expectant and new mothers with the virus had died.
That news was sobering enough. But what many experts found really worrisome were the glaring gaps in data that the study exposed.
The CDC acknowledged that crucial health information was missing for about
three-quarters of pregnant women with the virus, including whether they had preexisting conditions or required an ICU stay or mechanical ventilation. For the vast majority of U.S. women of reproductive age who tested positive — about 326,000 women through June 7 — there was no information about pregnancy status at all.
Researchers couldn’t even say how many of the hospitalized mothers-to-be — 31.5% of the pregnant women in the study — had been admitted because of COVID-19, versus how many were in the hospital for other reasons, such as giving birth.
The flawed CDC report highlights a problem that OB-GYN providers and researchers in the U.S. have been fretting about since the pandemic began. Because emerging diseases can have catastrophic consequences for pregnant women and their babies, close monitoring of new illnesses in this vulnerable population is important. So is rapid communication with providers trying to keep their patients safe.
But the U.S. public health system’s efforts to understand the impact of the coronavirus in mothers and babies have been flat-footed, scattershot and agonizingly slow.
The holes in the CDC’s data are “startling. I won’t lie,” said Cindy M. P. Duke, an OB-GYN who runs the Nevada Fertility Institute
in Las Vegas. “It’s shocking to realize that we do not have a uniform system in place” for collecting and analyzing basic maternal and infant health information during times of crisis.
“By the time we get the best data available,” said Christina Han
, a maternal-fetal specialist who is on the clinical faculty at the University of California, Los Angeles, medical school, “this pandemic will be over.”
To track the coronavirus in pregnant women and their babies, the CDC is instructing local health departments to check a box on the standard Case Report Form
if a patient is currently pregnant. Jurisdictions can also complete an “optional, supplemental form
” providing information about disease severity and outcomes in mothers and babies, a spokeswoman said.
It’s a system that guarantees there will be huge data inconsistencies, time lags and gaps. “We’re talking about millions of cases that are having to be reported back to the CDC right now,” Han said, “And the clinicians who are taking care of these patients don’t have the time to be filling out the full case report forms.”
The CDC has “initiated COVID-19 pregnancy surveillance,” according to its study, working with local health departments to improve data collection going forward. But researchers didn’t give any details.
“We need to be set up prior to an outbreak to capture information about how any new or emerging or reemerging pathogen affects pregnant women and their infants,” said Denise Jamieson
, who spent 20 years at the CDC studying reproductive infectious disease and now heads the OB-GYN department at Emory University medical school in Atlanta. “And I think it’s clear … that [the CDC] are not well poised to do that.”
It Doesn’t Have to Be This Bad
When COVID-19 arrived in Europe this winter, researchers in the United Kingdom were ready.
Caught unprepared by the H1N1 outbreak of 2009, and determined not to let that happen again, the National Institute for Health Research issued a call for studies in 2011 that could be up and running as soon as the next pandemic hit. Nine proposals
were chosen, all focusing on flu but adaptable to other types of outbreaks; then they were “hibernated” — put on hold, the protocols periodically reviewed and revised, pending government action.
When the sirens began blaring this February, eight of the studies launched, including one by the U.K. Obstetric Surveillance System
, a research platform devoted to rare disorders of pregnancy. “I was told to activate [it] on a Friday,” said Marian Knight
, a professor of maternal and child population health at the University of Oxford, “and by Monday, we were collecting the data.”
The UKOSS design was ambitious — a comprehensive, prospective study of COVID-19-infected pregnant women at the National Health Service’s 194 obstetric hospitals; it was also straightforward, based on a two-page electronic form
, with all patient information kept anonymous. The careful planning allowed researchers to move quickly without being overly rushed, Knight wrote in a blog post
for The BMJ journal, “minimizing the risks of publication of, at best, misleading, or, at worst, erroneous, information” — problems
that have bedeviled other high-profile COVID-19 studies.
By mid-May, UKOSS released
its first findings from an analysis of every pregnant woman hospitalized in Britain from March through mid-April who tested positive for the virus, 427 in total. Mothers-to-be did not seem to get as sick overall from COVID-19 as they had from H1N1 flu and SARS, but Black, Asian and Middle Eastern women were far likelier to be hospitalized with the virus than white women were. About 1 in 10 women became so ill they required respiratory support. Five mothers and five babies died.
The study provided the kind of high-quality data that doctors and policymakers value most: populationwide, scientifically rigorous, immediately actionable. British medical organizations quickly issued new guidance
about the heightened risks for women of color.
Medical groups also reaffirmed earlier recommendations
that women in their third trimester should avoid job settings, such as working as a doctor or nurse on the front lines, that could expose them to the virus. That’s a much more conservative position than American medical organizations
have taken, citing a lack of research that would suggest a different approach. Knight said the UKOSS findings support stringent social distancing in late pregnancy. “That’s where we see the majority of women with critical disease. The good news is that women now know when they should start to be particularly vigilant.”
Public health experts say UKOSS’ proactive, big-picture approach is exactly what was needed in the U.S., which has nearly 4 million births annually — about five times as many as Britain — and wide racial disparities
and infant outcomes that COVID-19 has likely only widened. By one estimate
, more than 16,000 American women could be infected with the coronavirus when they give birth this year.
But the U.K. has universal health care, which gives the government access to vast amounts of patient data, making large-scale research like the UKOSS study more feasible. The U.S. health care system is fragmented and inefficient, much of its data hidden in black boxes.
The two nations also have very different approaches to maternal health, exemplified by how they confront pregnancy-related complications and deaths. The U.K. treats
every mother’s death like a public health disaster; the U.S. can barely
of its maternal mortality problem
Women’s health in general tends to be an afterthought for American researchers and policymakers, said Barbara Levy
, former vice president for health policy at the American College of Obstetricians and Gynecologists, now a consultant in the Washington, D.C., area. In a public health system that’s been underfunded for decades
, pregnancy-related issues get especially short shrift.
“If we’re not seeing a red-hot button to raise our concern, then we just aren’t particularly curious about what might be going on.” Yet the stakes of this kind of population-based research couldn’t be higher, Levy said. “This is the future of our world. This is the next generation.”
Patching Together a View
Starting with the first, very limited case reports from China, there’s been a deluge of studies seeking to fill the information void about COVID-19 and pregnancy. In the U.S., much of the research has come from hot spots like New York City
, where physicians in the throes of the pandemic have reported on their own patients and pushed out their findings with breakneck speed.
The general trend has been somewhat reassuring: COVID-19 can make pregnant women very sick, but many seem to remain asymptomatic or become only mildly ill. There have only been a few cases of suspected “vertical transmission” — mothers passing the virus to their babies in the womb — but those remain under investigation, and most infants born to COVID-19-positive women have done fine. Some red flags have also emerged: A couple of cases of heart-related
complications, indications of placental abnormalities
and reports of asymptomatic women who spiraled
after giving birth.
The underlying data in many of these case reports is overlapping and confusing, with little or no peer review to vet the findings. “In none of these series can we be certain of the underlying population denominators nor the degree to which they are affected by the biases” inherent in small-scale studies conducted in a single hospital or city, Knight wrote
in her BMJ commentary. “The results are thus almost impossible to interpret.”
Meanwhile, numerous efforts are underway to do larger scale, longer-term, more scientifically meticulous studies that could give a fuller picture of how the virus affects pregnant women and babies in different parts of the U.S. (which is likely to be different from how it affects women in the U.K. or Sweden
). In late May, the National Institute of Child Health and Human Development
announced a sweeping initiative, a series of studies at the 12 academic institutions in its Maternal-Fetal Medicine Units Network
, the preeminent obstetric research collaborative in the U.S. The network, which dates to the mid-1980s and stretches from the Rhode Island to Utah, is “a very well-oiled machine that can put protocols into place very quickly,” said Diana Bianchi
, a prenatal geneticist who heads the NICHD, which is one of the National Institutes of Health.
One big study will compare women who delivered at the various sites between March and December of this year with women who delivered in the same hospitals in the same period in 2019, Bianchi said. “They’re asking, first of all, is there a difference in the number of deaths? Is there a difference in the number of cesarean sections that have been performed? Is there a difference in terms of the maternal near misses?... If we see any increased incidence of complications, death, whatever, then the presumption is the difference is due to COVID.” Another study will follow 1,500 COVID-19-positive pregnant women to understand the course of the disease and its treatment.
A different approach, the PRIORITY
study, originates at the University of California, San Francisco, and UCLA. The idea is to build a large nationwide registry of women who contracted COVID-19 at any point in their pregnancy and to follow them for at least a year after they give birth, miscarry or have an abortion. A key goal is to make the registry more geographically, racially and economically diverse than is typical of many academic-based studies, “because we want the results to accurately reflect the true representation of pregnant people that are being impacted by COVID 19,” said Vanessa Jacoby
, an associate professor at UCSF who is one of the lead investigators. To that end, the researchers are working with birth advocates in communities of color around the country to get the word out to potential participants directly; they don’t need a doctor or hospital to take part. In the first few weeks, many of the women enrolled were pregnant health care workers.
Another UCSF-based study, the ASPIRE
registry, will focus on early pregnancy, when the risks are greatest for birth defects and miscarriages, and aims to eventually enroll 10,000 participants.
“This disease is so humbling because it’s new to humans, and we keep learning more and more about its manifestations in real time,” Bianchi said. “We have not gone through a nine-month cycle with pregnant women. And so I’m very concerned about the consequences of infection in the first trimester.” She points to Zika
, a mosquito-borne virus that was once thought to be harmless because the symptoms in men and nonpregnant women were often quite mild. Only after newborns exhibited devastating neurological disabilities did researchers realize how dangerous the virus really was.
Recent reports of COVID-19’s unexpected effects — like a life-threatening inflammatory syndrome in children — have intensified providers’ unease. On a Facebook group for OB-GYNs with 4,400 members, doctors have recently been trading stories
about an alarming, if highly anecdotal, increase in miscarriages and stillbirths in some of their practices. But the OB-GYNs didn’t know whether what they were seeing was a real spike caused by COVID-19, an increase caused by something else or a statistical fluke magnified by their own anxiety. And what, they worried, might the reports about clotting disorders
in younger, nonpregnant adults mean for expectant and new mothers, who are already more susceptible
to life-threatening blood clots?
“The gap in knowledge — it’s chilling,” said Jane van Dis, a Los Angeles-area OB-GYN who is the medical director at the Maven
women’s health telemedicine platform and helps administer the Facebook group. “It’s very distraught-provoking to be expected to be an expert on a topic that we’ve spent our whole lives dedicated to, but for which we don’t have the answers right now, and they’re really important answers to be able to have.”
The truth is, answers may be elusive for a long time. “This virus keeps surprising us,” said Sonja Rasmussen
, former director of the CDC’s Office of Infectious Diseases, now a professor of pediatrics and epidemiology at the University of Florida medical school. “That was true for Zika and Ebola, too.” One of the most important lessons of past outbreaks, she said, is that “you can’t be too confident early on. It just shows the importance of collecting data and not being too sure about what you think you're going to see before you collect those data. You’ve got to keep an open mind.”