This story was originally published by ProPublica, a Pulitzer Prize-winning investigative newsroom. Sign up for The Big Story newsletter to receive stories like this one in your inbox.
Since the Supreme Court overturned Roe v. Wade last year, doctors have warned that limiting abortion care will make pregnancy more dangerous in a country that already has the highest maternal mortality rate among industrialized nations.
The case of Mylissa Farmer, a Missouri woman, is one example. Last August, her water broke less than 18 weeks into her pregnancy, when her fetus was not viable. She was at risk for developing a life-threatening infection if she continued the pregnancy. Yet during three separate visits to emergency rooms, she was denied abortion care because her fetus still had a heartbeat. Doctors specifically cited the state’s new abortion law in her medical records and said they could not intervene until her condition worsened. She eventually traveled to Illinois for care.
Even for people who don’t develop sudden life-threatening complications, doctors note that carrying a pregnancy to term is inherently risky because rapid physical and hormonal changes can exacerbate chronic health conditions and trigger new complications. If more people are forced to continue unwanted pregnancies, there are bound to be more pregnancy-related deaths: A study by the University of Colorado estimates a 24 percent increase in maternal deaths if the United States bans abortion federally. They predicted the increase would be even higher for Black patients, at 39 percent. Currently, 14 states have total abortion bans.
The University of Colorado estimates a 24 percent increase in maternal deaths if the United States bans abortion federally. The increase would be even higher for Black patients.
Additionally, when abortion is illegal, it makes the procedure more dangerous for those who still try to terminate their pregnancies. The World Health Organization found that unsafe or illegal abortions account for up to 10 percent of maternal deaths worldwide.
As the United States enters its second post-Roe year, advocates say it’s important to gather data on the impact abortion bans are having on the health of pregnant people to help both policy makers and voters understand the life-or-death consequences of the restrictions. Without such accounting, they say, the public may remain ignorant of the toll. Maternal mortality rates would be a crucial gauge of impact.
Despite the stakes, experts say, at least in the short term, it may be difficult or impossible to track the number of lives lost due to limits on abortion access.
ProPublica spoke to four members of state maternal mortality review committees. Here are some of the challenges they see to drawing any clear conclusions from maternal mortality data in the near future.
The data can be inconsistent
Each state has its own system for compiling the data maternal mortality researchers work with. The quality of the data varies vastly by state. It can involve comparing birth and fetal death records, scanning through obituaries, and sometimes begging coroner’s offices to send death records. Many states are still working toward a complete system.
“It really depends on the rigor of the contributing entities,” said Dr. Michelle Owens, a maternal-fetal medicine specialist and the clinical chair for Mississippi’s maternal mortality review committee. “We rely so heavily upon the information we glean from these sources, and if that information is not as reliable … it will definitely have a negative impact on our work and understanding of what the contributing things may have been and what the gaps are.”
All the maternal mortality experts that ProPublica spoke with noted issues with the “pregnancy check box” used in death certificates to denote whether a patient was pregnant at the time of death or within the previous year. In Florida, Dr. Karen Harris, an OB-GYN and a member of Florida’s maternal mortality review committee, has observed the check box “overselect some patients who were never pregnant, or not pregnant in the last year, and it underselects patients who were pregnant.”
Sometimes the check box is wrong because of clerical errors, the researchers said. Other times, it’s simply not filled out because no autopsy was performed to verify whether the person was pregnant. That information could be important in measuring deaths that happen early in pregnancy — including murders. Homicide is a leading cause of death for pregnant or recently pregnant Americans, and researchers also would like to measure how abortion bans, which could force people in abusive relationships to carry unwanted pregnancies, affect those numbers.
Studying pregnancy-associated deaths within a year of pregnancy helps researchers account for any additional factors like substance abuse, unstable housing, suicide or mental health problems. These could be important in identifying deaths connected to continuing an undesired pregnancy.
The data can also be slow — some states, like Florida, provide data to the committee for the past year right away. But others are years behind. Currently, many states have only released data through 2019.
Records may not address abortion access
One of the thorniest questions facing maternal mortality experts: How can they determine if abortion access was a factor?
Dr. Lynlee Wolfe, an assistant professor at the University of Tennessee Medical Center and a member of the state’s maternal mortality review committee, wishes maternal mortality review reports could include a check box for the question, “Did inability to get an abortion play a role?”
“But you often can’t dig that out of notes,” she said. “I think what we’re asking is kind of an untrackable number.”
The experts said they could look into causes of death that may be linked to a patient’s inability to get an abortion when they’re having an emergency pregnancy complication: Sepsis, hemorrhage and heart issues, for example, are all worth studying to see if medical records might indicate if doctors delayed ending the pregnancy because the fetus still had a heartbeat.
“I think what we’re asking is kind of an untrackable number.”
But beyond that, when the pregnancy was unwanted or exacerbated broader health concerns, it could prove very difficult to determine if abortion access was a factor in the patient’s decision-making.
For example, if a patient had a heart condition that carried a 50 percent chance of death in pregnancy, researchers would like to see whether the patient was counseled about the risk and offered a termination.
But in a state that had criminalized abortion, “no one’s going to write that down,” said Harris, the Florida doctor. “So we won’t be able to know in the in-depth review if this was a patient choice — or if it was something that was forced upon her.”
Researchers might be able to learn more about the patient’s state of mind and whether the pregnancy was desired or not from interviews with family members and social service records, Owens, the Mississippi doctor, said. But there’s no guarantee they would have discussed their feelings about the pregnancy with family members either.
“With stigma and controversy surrounding conversations and considerations around abortions, people are hesitant to share those thoughts and feelings outside a very small circle of trust,” she said.
Risk of political interference
Maternal mortality review committees are funded by their states, and some are overseen by state legislatures.
The maternal mortality review members ProPublica spoke with said they did not anticipate interference with their report findings, even if they found examples where abortion access was a factor in a maternal death.
But some maternal care advocates worry such committees are vulnerable to political interference and manipulation. Last year, the Texas Department of State Health Services announced it was delaying its 2019 maternal mortality review report, originally scheduled for September 2022, until mid 2023.
Some saw the delay as a way to keep negative numbers out of the public eye during election season and postpone their release until after the 2023 legislative session had ended. A member of the review committee said she believed there was no legitimate need for the delay and that it was “dishonorably burying these women.” ProPublica reached out to the committee and the Texas health agency to ask about these concerns, but did not receive any response.
After pushback, the report was partially released in December 2022. It found persistent disparities affecting Black mothers and showed that the childbirth complication rate had risen 28 percent since 2018.
In July, Idaho disbanded its maternal mortality review committee, making it the only state without one. Lawmakers cited the costs of operating the committee — though members said operating costs were about $15,000 a year and covered by a federal grant. The decision came after a lobbying group argued that the committee was a “vehicle to promote more government intervention in health care” and opposed its recommendation to extend Medicaid coverage to mothers for 12 months postpartum.
The sample size is small
Maternal mortality rates in the U.S. are higher than in other wealthy countries and have been rising in recent years, so many resources are devoted to studying root causes of the trend and possible strategies for reversing it. But the actual number of deaths is statistically small: In 2021, the U.S. saw an estimated 32.9 deaths per 100,000 births, or 1,205 total pregnancy-related deaths, according to the Centers for Disease Control and Prevention.
This makes it difficult to draw conclusions that are rigorous by epidemiological standards, said Dr. Elliot Main, a Stanford professor and the former medical director for the California Maternal Quality Care Collaborative.
While researchers may learn of individual cases where it’s clear that abortion access was an issue in the patient’s outcome, it could take years to have a data set large enough to reveal a clear picture.
Main also pointed out that many other factors influence maternal mortality rates, which muddles the picture. “Maternal deaths are so rare and often complicated in their underlying causes,” he said. “If you see a trend over time, we have to break it down to see what’s really causing that.”
Before the Supreme Court’s decision in Dobbs v. Jackson Women’s Health Organization struck down federal protections for abortion rights, U.S. maternal mortality rates were already rising. Influences include COVID-19, the opioid crisis and people having children at older ages, when they are at higher risk for complications. The U.S. also has long-standing racial and socioeconomic health care disparities affecting quality prenatal care — more than half of Georgia’s counties have no OB-GYN, for example. That can mean more patients go into pregnancy with undiagnosed health conditions and may be at higher risk for life-threatening complications.
Main and other researchers suggested that studying data on childbirth complications may provide more avenues for understanding the effects of abortion bans, because those are more common and would provide a larger data set to study.
Bans don’t prevent all abortions
One reason the impact of Dobbs on maternal mortality rates could remain limited even in states that have banned abortion is that some people who want to terminate their pregnancy are still able to do so, either by traveling or by ordering abortion medication in the mail.
It’s impossible to know the full picture of how many are able to jump through the hoops and obtain abortions even when there are no legal options nearby. But WeCount, a research project led by the Society of Family Planning that has been collecting data from abortion providers, estimates that in the six months following Dobbs, about 35,000 people in abortion-ban states were able to get abortions in other states — just over half of the people estimated to have sought abortions in those states, based on numbers from the same time period the previous year. It’s unclear what happened to the other half. Some may have continued their pregnancies, others may have ordered abortion pills in the mail, which could be sent by organizations based in Europe and Mexico and not be recorded in any database.
Still, having to travel out of state to a limited number of abortion providers meant more patients were forced to wait until their second trimester, researchers said, when an abortion can be more complicated.
And while abortion pills are considered an exceedingly safe method of terminating a pregnancy through the first 10 weeks, according to the Food and Drug Administration and leading medical organizations, patients should still have the option to take them with the instruction and care of a medical provider, advocates say.