Few public-health numbers carry as much emotional weight as the U.S. maternal mortality rate. It is not just a statistic on a government spreadsheet; it represents mothers, babies, families, doctors, nurses, hospital systems, public policy, and the deeply uncomfortable question of whether pregnancy in America is more dangerous than it should be. So when a new report suggested that maternal deaths may have been overestimated, the reaction was immediate: some people breathed a sigh of relief, some raised an eyebrow, and the CDC essentially said, “Not so fast.”
The debate centers on how the United States counts maternal deaths. More specifically, it focuses on a small but powerful item on death certificates: the pregnancy checkbox. That little box was designed to help identify deaths that might otherwise be missed. But critics argue it may also pull in deaths that are not truly caused or aggravated by pregnancy. In other words, the checkbox may be both a helpful flashlight and, occasionally, a disco ball throwing light everywhere.
The new analysis, published in the American Journal of Obstetrics & Gynecology, argued that official maternal mortality numbers from the National Vital Statistics System may have been inflated because some deaths were counted based mainly on a checked pregnancy box rather than clear pregnancy-related causes listed on the death certificate. The CDC disagreed, saying the alternative method could miss real maternal deaths that the checkbox was created to catch in the first place.
So, are maternal deaths being overestimated? The careful answer is: possibly in some ways, but that does not mean the U.S. maternal health crisis is imaginary. The better question is what kind of counting system produces the most accurate, useful, and humane picture of risk.
What Counts as a Maternal Death?
A maternal death is generally defined as the death of a woman while pregnant or within 42 days after the end of pregnancy, from a cause related to or aggravated by pregnancy or its management. That definition excludes deaths from accidental or incidental causes. For example, a death caused by an unrelated accident would not normally count as a maternal death, even if the person happened to be pregnant.
That sounds simple until real life walks into the room wearing muddy boots. Pregnancy can worsen existing heart disease. It can delay cancer treatment. It can complicate infection, blood pressure, bleeding risk, mental health, and access to care. A cause may not look “obstetric” in a narrow sense, yet still be connected to pregnancy. This is why maternal mortality measurement is more complex than counting births, deaths, and calling it a day.
In the United States, the CDC’s National Center for Health Statistics uses death certificate data to produce official maternal mortality rates. Those certificates include cause-of-death information completed by medical certifiers, and they also include pregnancy-status information. The pregnancy checkbox asks whether the person was pregnant or had recently been pregnant at the time of death.
Why the Pregnancy Checkbox Was Added
The pregnancy checkbox was not invented to create scary headlines. It was added because maternal deaths were historically undercounted. Without a prompt asking about pregnancy status, some deaths related to pregnancy could be missed entirely, especially when the death occurred outside a delivery room or weeks after birth.
Imagine a postpartum patient who dies from a cardiovascular complication several weeks after giving birth. If the certifier does not know or record the recent pregnancy, the death may disappear into a general heart-related category. To public-health officials, that is a huge problem. You cannot prevent what you cannot see. The checkbox was meant to wave a flag and say, “Please look closer.”
States adopted the checkbox gradually, and all states had implemented it by 2017. That staggered rollout matters because national trends before and after adoption are difficult to compare. When a surveillance system suddenly gets better at finding cases, the numbers may rise even if the true risk has not changed as dramatically. It is the public-health equivalent of cleaning your glasses and suddenly noticing dust on every shelf.
What the New Report Claimed
The new report challenged the idea that U.S. maternal mortality had dramatically increased over the past two decades. Researchers argued that official numbers rose partly because the pregnancy checkbox identified deaths that did not clearly include pregnancy-related causes on the death certificate.
Using an alternative method, the researchers counted maternal deaths only when pregnancy was mentioned among the multiple causes of death. Under that approach, they found a much lower and more stable maternal mortality rate: around 10 deaths per 100,000 live births in both 1999–2002 and 2018–2021. That is very different from the official trend, which showed a sharp increase over time and reached 32.9 deaths per 100,000 live births in 2021.
The study’s central argument is not that no women are dying from pregnancy-related causes. Rather, it argues that official surveillance may have blurred the line between deaths that occurred during or near pregnancy and deaths caused or worsened by pregnancy. That distinction is crucial. A death that happens during pregnancy is not automatically a maternal death under the formal definition.
Why the CDC Disagrees
The CDC’s response is essentially this: removing deaths identified only by the pregnancy checkbox may solve one problem while creating another. Yes, checkbox errors can happen. But the checkbox also catches real maternal deaths that might not otherwise be identified through cause-of-death wording alone.
Cause-of-death information is not perfect. Certifiers may not always know a patient’s pregnancy history. Medical records may be incomplete. A postpartum complication may be described in broad terms without explicitly tying it to pregnancy. In those cases, requiring pregnancy to be written into the cause-of-death section could undercount deaths that truly belong in maternal mortality statistics.
That is the heart of the disagreement. The new report worries about false positives: deaths counted as maternal when they should not be. The CDC worries about false negatives: deaths excluded even though pregnancy played a role. Both concerns are legitimate. Unfortunately, maternal mortality data are not a math quiz where the answer is hiding neatly in the back of the book.
What the Latest CDC Data Show
According to final 2024 data from the National Center for Health Statistics, 649 women died of maternal causes in the United States in 2024, compared with 669 in 2023. The 2024 maternal mortality rate was 17.9 deaths per 100,000 live births, not significantly lower than the 2023 rate of 18.6. That marks a major decline from the pandemic peak in 2021, when the rate reached 32.9.
The racial disparities remain striking. In 2024, the maternal mortality rate for Black women was 44.8 deaths per 100,000 live births, far higher than the rates for White women, Hispanic women, and Asian women. Age also matters: women age 40 and older had a rate of 62.3 deaths per 100,000 live births, about five times higher than the rate for women younger than 25.
These numbers complicate the “overestimated or not?” debate. Even if some official counts are inflated by reporting issues, the burden is not evenly distributed. Black women and older mothers still face substantially higher risk. A cleaner dataset would not magically erase those disparities. It might sharpen them, explain them better, and help target prevention more effectively.
Maternal Mortality vs. Pregnancy-Related Deaths
One reason this topic gets confusing is that different systems count different things. “Maternal mortality” usually refers to deaths during pregnancy or within 42 days after pregnancy from causes related to or aggravated by pregnancy. “Pregnancy-related deaths,” often reviewed by Maternal Mortality Review Committees, can include deaths up to one year after pregnancy if the death is determined to be related to pregnancy.
That longer one-year window captures important postpartum risks. Many serious problems do not politely occur within six weeks and then stop because the calendar says so. Cardiovascular conditions, infection, blood clots, mental health crises, substance-use complications, and chronic disease can all appear or worsen after delivery.
Maternal Mortality Review Committees are especially valuable because they review medical records, social factors, health care access, and other context that death certificates alone cannot provide. These committees often ask not only “What happened?” but also “Could this have been prevented?” That second question is where public health becomes practical.
Why Better Data Matters
Accurate maternal mortality data are not just for researchers in sensible shoes. They influence funding, hospital protocols, public awareness campaigns, insurance decisions, postpartum care standards, and state-level health policy. If deaths are overcounted, resources may be misdirected. If deaths are undercounted, preventable harm may remain hidden. Neither outcome is acceptable.
Think of maternal mortality data like a smoke alarm. A smoke alarm that goes off when toast burns is annoying. A smoke alarm that stays silent during a real fire is dangerous. The goal is not to throw away the alarm; the goal is to make it smarter.
The pregnancy checkbox may need better verification, not elimination. The CDC has already acknowledged data-quality concerns and continues to evaluate the checkbox’s impact. State vital records offices also now provide additional verification for many death certificates with a positive pregnancy checkbox. That is progress, though not a perfect fix.
Examples That Show Why the Debate Is So Tricky
Example 1: A Clear Obstetric Death
A patient dies shortly after delivery from severe hemorrhage directly related to childbirth. This is clearly a maternal death. Both the CDC method and the stricter alternative method would likely count it.
Example 2: A Possible Indirect Maternal Death
A pregnant patient with heart disease dies from a cardiac event. Pregnancy may have placed additional strain on the cardiovascular system. If the death certificate does not clearly mention pregnancy as a contributing factor, a stricter method might miss it, while the checkbox could trigger further identification.
Example 3: A Checkbox Error
A death certificate mistakenly marks a person as recently pregnant when there was no pregnancy. That could lead to an overcount if not caught. These errors are especially concerning because maternal deaths are relatively rare, so even a small number of mistakes can noticeably change the rate.
Example 4: Cancer During Pregnancy
A pregnant patient with cancer dies after treatment was delayed or altered because of pregnancy. Is that maternal? It may be, if pregnancy affected treatment options or the disease course. This is exactly where rigid rules can fail and careful review becomes essential.
The Real Takeaway: Measurement Is Messy, but the Crisis Is Real
The debate should not be reduced to two cartoonish camps: “The numbers are fake” versus “Never question the numbers.” Good science lives in the uncomfortable middle. The U.S. needs to improve maternal mortality measurement while also reducing preventable deaths.
CDC-supported reviews have found that a large share of pregnancy-related deaths are preventable. That means better emergency response, stronger postpartum follow-up, improved management of chronic conditions, safer hospital protocols, more respectful care, and better access to maternity services can save lives. Better counting should serve prevention, not distract from it.
The new report is useful because it forces a serious conversation about data quality. The CDC’s disagreement is useful because it reminds us that a stricter counting method may look cleaner while missing deaths that matter. Public health rarely improves by choosing the simplest answer; it improves by choosing the most accurate one and then doing the hard work.
What Should Change Next?
First, death certificates should require clearer documentation when the pregnancy checkbox is marked. If pregnancy status is checked, certifiers should be encouraged to explain how pregnancy did or did not contribute to death. A checkbox alone should be a starting point, not the final verdict.
Second, states should continue strengthening Maternal Mortality Review Committees. These committees can examine records in depth and identify prevention opportunities that national death certificate data may miss. Their work is slower than automated statistics, but it is richer and often more useful.
Third, public reporting should explain uncertainty more clearly. A maternal mortality rate should not be presented as if it dropped from the sky carved into stone tablets. Readers deserve to know how the number was calculated, what changed in the surveillance system, and where errors may occur.
Fourth, maternal care must improve regardless of the final statistical debate. Whether the “true” rate is closer to 10, 18, or 30 deaths per 100,000 live births, preventable maternal deaths are still unacceptable. A lower number is not a permission slip to shrug.
Experiences and Practical Lessons From the Maternal Mortality Debate
One experience that comes up again and again in maternal health discussions is the gap between what data systems record and what families live through. A spreadsheet may say “cardiac condition,” “infection,” or “hypertensive disorder,” but a family remembers the phone call, the waiting room, the unanswered concern, the appointment that was delayed, or the symptom that was brushed aside. This is why the debate about overestimation must be handled with care. Behind every counted case, and behind every missed case, there is a story that deserves accuracy and respect.
Clinicians often describe maternal risk as something that does not end at delivery. In everyday conversation, birth can sound like the finish line. In medical reality, it is often the beginning of a vulnerable recovery period. Blood pressure can spike after birth. Heart symptoms can be mistaken for exhaustion. Postpartum pain may be normalized too quickly. A patient may be sent home with instructions that are technically correct but practically confusing, especially when she is sleep-deprived, recovering, and caring for a newborn who has apparently mistaken 3 a.m. for party time.
Another practical lesson is that patients and families need plain-language warning signs. Severe headache, chest pain, trouble breathing, heavy bleeding, fainting, swelling, fever, or thoughts of self-harm should never be waved away as “just postpartum.” Health systems have improved educational materials, but the message must be repeated in clinics, hospitals, discharge papers, telehealth visits, and community programs. A warning sign is only useful if someone recognizes it in time.
There is also an experience familiar to many public-health workers: data improvement can look like the problem is getting worse. When better screening finds more cases, the numbers rise. That can be politically awkward, but it is not failure. It is visibility. The pregnancy checkbox may have revealed deaths that were previously hidden, even while introducing classification errors. The solution is not to become allergic to data; it is to improve verification so the numbers become both sensitive and specific.
Families, clinicians, researchers, and policymakers all benefit from the same principle: listen carefully before concluding. Listen to patients when they say something feels wrong. Listen to clinicians when they identify documentation problems. Listen to statisticians when they warn about misclassification. Listen to communities when they point out unequal treatment. Maternal mortality is not only a medical issue; it is a communication issue, a systems issue, and a trust issue.
The most useful experience-based takeaway is simple: better counting and better care must happen together. Accurate numbers help identify where the fire is. Compassionate, timely, high-quality care helps put it out.
Conclusion
So, are maternal deaths being overestimated? The best answer is nuanced: some official numbers may be affected by pregnancy checkbox misclassification, but stricter methods may also undercount genuine maternal deaths. The CDC disagrees with the new report because it believes the proposed approach could miss deaths that the checkbox was designed to detect.
The debate should push the United States toward better data, not smaller concern. Maternal mortality reporting needs more verification, clearer death certificate documentation, stronger review committees, and more transparent public communication. At the same time, the country must continue addressing the real risks that pregnant and postpartum patients face, especially Black women, older mothers, and people with limited access to quality care.
In the end, the goal is not to win an argument about whether the number is too high or too low. The goal is to make the number accurate, make the care better, and make preventable maternal deaths rarer. That is the kind of math every family can support.
