Maternal Deaths Fall, But 104 Countries Miss SDG Target
Every two minutes, somewhere in the world, a woman dies because she tried to bring a child into it. That number has been improving. It is still obscene.
The latest global accounting, published in March 2026 in The Lancet Obstetrics, Gynaecology & Women's Health as part of the Global Burden of Disease 2023 study, puts the death toll at 240,000 women in 2023 alone. That figure represents 5.5 percent of all deaths among women and girls aged 10 to 54 worldwide. It also represents a decline of more than a third from 1990, when the ratio stood at 321 deaths per 100,000 live births. Today the number sits at 190.5. Real progress. Real families kept intact. Real women who are alive today who would not have been thirty years ago.
And yet 104 of the 204 countries and territories in the analysis, led by researchers at the Institute for Health Metrics and Evaluation at the University of Washington, have missed the Sustainable Development Goal target of fewer than 70 maternal deaths per 100,000 live births. Seventy-three of those countries sit above 140 per 100,000. Sixteen hover between 100 and 140. Fifteen sit between 70 and 100, just shy of the line. The target was meant to be reached by 2030. Less than five years remain. The math is no longer flattering.
Worse is the direction of travel. Between 2000 and 2015, during the Millennium Development Goals era, maternal mortality fell by nearly 3 percent per year on average. Since 2015, when the SDGs replaced those goals, global declines have averaged only about 0.5 percent annually. In 120 of the 204 countries studied, the rate of change has slowed or reversed since 2015. Researchers describe this as "plateauing." Plateauing is a generous word for what is really happening. In some countries, women are dying in childbirth at rising rates.
The geography is familiar to anyone who follows this field. Sub-Saharan Africa carries roughly 70 percent of the global burden. In 2023, the countries with the highest absolute numbers of maternal deaths were Nigeria (about 32,900), India (about 24,700), the Democratic Republic of the Congo, Ethiopia (about 11,900), and Pakistan (about 10,300). Five countries, Nigeria, India, Ethiopia, Pakistan, and Cameroon, account for roughly 36 percent of all maternal deaths worldwide. One in three.
The countries with the highest ratios of deaths per 100,000 live births tell a different story. Liberia, the Central African Republic, Haiti, Eritrea, and Sierra Leone top that list. Sierra Leone, which had the world's worst MMR for most of the last two decades at above 1,100 per 100,000, has pulled its rate down dramatically since the civil war and the Ebola years, though it still ranks near the top. The country has only about 3 physicians per 100,000 population, and roughly 10,000 auxiliary nurses work in the public health system on an unsalaried volunteer basis because the government cannot afford to pay them. Liberia's ratio has fallen from 728 in 2012 to 628 in 2023 according to the World Bank, still higher than the sub-Saharan African regional average of 448. These are not abstractions. They are the ratios you end up with when fewer than two doctors per 10,000 people exist in the country you happen to be born in.
How women actually die
The causes are depressingly predictable. Obstetric hemorrhage and hypertensive disorders of pregnancy, including preeclampsia and eclampsia, together account for more than 40 percent of maternal deaths globally. A 2025 WHO systematic analysis covering 2009 to 2020 data, drawing on 139,381 maternal deaths across 129 countries, put hemorrhage at 27 percent and hypertensive disorders at 16 percent of all maternal mortality. Indirect causes, meaning existing health conditions that complicate pregnancy such as HIV, malaria, anemia, and diabetes, accounted for another 23 percent. Pregnancy-related sepsis, complications from abortion, and embolism made up most of the rest.
Hemorrhage deaths are falling as a share of the total. The GBD 2023 analysis found that hemorrhage's contribution dropped from about 36 percent in 1990 to 21.7 percent in 2023. That fall reflects wider availability of uterotonic drugs like oxytocin and misoprostol, uterine balloon tamponade for managing postpartum bleeding, and blood transfusion services, even if unevenly.
Hypertensive disorders are moving the other way. Preeclampsia's global prevalence now hovers around 4.4 percent of pregnancies. Its share of maternal deaths rose from 16.6 percent in 1990 to 20.1 percent in 2023. In India, a middle-income country grappling with both tradition and a rapidly medicalizing obstetric sector, hypertension-related maternal deaths rose from 10 to 12.1 percent of the total. Some of this is better detection. Some are demographic. Older mothers, higher obesity rates, and more pre-existing chronic hypertension all drive preeclampsia risk. All of it is preventable with early antenatal blood pressure monitoring and urine protein screening and treatable with magnesium sulfate and antihypertensives. Both drugs cost pennies.
One trend is getting less press than it deserves. Late maternal deaths, defined as those between 43 days and one year after delivery, have more than doubled globally as a share of the total, from 1.3 percent to 3.2 percent. That points at something uncomfortable. Much of the maternal health system still ends at the six-week postpartum visit. The world of obstetric medicine has quietly improved at keeping women alive through delivery. It has not figured out how to keep them alive afterward. Cardiovascular complications, severe mental illness, opioid overdose, and delayed progression of hypertensive disease are stalking women during a period when the clinical establishment has stopped watching them. In the United States, postpartum suicide and unintentional overdose now account for a larger share of pregnancy-associated deaths than hemorrhage. That shift demands a different kind of care, one organized around a whole year after birth rather than six weeks.
The co-lead author of the new study, Ira Martopullo of the University of Washington, captured the issue honestly: "Maternal mortality is both a health system challenge and a reflection of broader inequities affecting women's health." The sentence reads like a diplomatic cliché until you translate it. Women die in childbirth because the societies they live in are willing for them to.
The pandemic, the retreat, and the cost of walking away
COVID-19 did something grim to this picture. Using an excess mortality approach, the GBD researchers estimate roughly 22,900 additional maternal deaths occurred during 2020 and 2021 combined. Places with already-strained systems, including parts of Latin America and the Caribbean, Central and Eastern Europe, Central Asia, and North America, saw maternal deaths climb. In the United States, the number of pregnancy-related deaths rose from about 660 pre-pandemic to 1,222 in 2021 before returning to 676 in 2023. Some of that reflected direct COVID infection in pregnant women, who are more vulnerable to severe disease and who, in the first year of vaccine rollout, were widely advised against the shot by physicians uncertain of the data. Some reflected deferred care and health system overload. Most reflected both.
The United States deserves its own section here, because its position is unique among wealthy countries and the contrast is instructive. In 2023, the US maternal mortality rate fell to 18.6 deaths per 100,000 live births, down from 22.3 in 2022. That is the lowest it has been in years, and the decline among White and Hispanic mothers was statistically significant. The US rate is still higher than nearly every other high-income country and several times that of Norway or Sweden. The state-level picture is grimmer. Louisiana sat at 41.9 per 100,000 in 2023, a figure that would place it outside the top two-thirds of world rankings. Mississippi, Alabama, and Tennessee report similar pregnancy-related death ratios. California and Massachusetts sit below 25. This is not a country with a single maternal mortality rate. It is a country with Nordic outcomes for some women and outcomes resembling Morocco for others.
The racial gap is the sharpest indictment. Black women in the US died at 50.3 per 100,000 live births in 2023, a rate 4.5 times that of Asian women (10.7), four times that of Hispanic women (12.4), and 3.5 times that of White women (14.5). The gap has not meaningfully narrowed over a decade. American Indian and Alaska Native women faced a rate of 40.7; Native Hawaiian and Pacific Islander women, 37.0. These differences persist across income levels, education levels, and zip codes. Serena Williams nearly died from a postpartum pulmonary embolism after her daughter's birth, despite being a world-famous athlete with access to the country's best medical care, because nurses initially dismissed her symptoms. If that happens to her, consider what happens to women who are not her.
Age matters too. The US maternal mortality rate for women age 40 and older was 59.8 per 100,000 live births in 2023, nearly five times the rate for women under 25. The average age at first birth in the US has been creeping upward for forty years, pushed by education, economics, and the reality that having children before 30 has become financially ruinous for much of the professional class. Older pregnancies carry higher risk. A health system that has not adapted to older mothers produces more dead ones.
Then there is the question of what happens next, and this is where the article has to take a position. In January 2025, the incoming US administration issued an executive order suspending nearly all foreign aid programs. Within months, 83 to 86 percent of USAID awards were canceled, including a reported 86 percent of all maternal and child health projects. The FY 2026 budget request included no funding for bilateral maternal and child health efforts. The Global Gag Rule was reinstated. USAID itself was slated for dissolution, with its global health activities nominally absorbed by the State Department.
USAID was not a minor player in this story. It provided roughly 40 percent of donor funding for family planning worldwide and was one of the largest sources of support for emergency obstetric care in low-income settings. It was also, for many fragile states, the backbone. An internal USAID memo, reported by KFF, warned that the cessation of programming would eliminate services for 16.8 million pregnant women annually and postnatal care for 11.3 million newborns within their first two days of life. Modeling studies published in The Lancet in mid-2025 projected that sustained cuts could cause 14 million additional deaths by 2030, including 4.5 million children under five. Research in Health Policy and Planning estimated that in six West and Central African countries alone, the cuts would cause approximately 1,000 additional maternal deaths in 2025, with Niger facing a projected 90 percent rise in maternal mortality and Nigeria more than 300 additional deaths. A separate LiST model analysis put the likely global MMR increase at 18.4 percent from 2024 levels if cuts are sustained.
This is the part of the story that most coverage buries at the end. The decline in global maternal deaths over the past three decades was not an accident of prosperity. It was built with deliberate, unglamorous funding: midwife salaries, oxytocin stockpiles, ambulances to rural clinics, antenatal care campaigns that taught women to recognize the warning signs of preeclampsia, and free-at-point-of-use delivery services. Pulling that scaffolding away produces exactly the results you would expect. Pretending otherwise is either ignorance or lying.
Countries that have done well show what sustained investment looks like. Ethiopia, Bangladesh, Nepal, and Cambodia have each cut their maternal mortality rates substantially over the past two decades by expanding facility-based delivery and training skilled birth attendants, despite remaining poor. India reduced its ratio from 508 in 1990 to 116 in 2023, a drop of nearly 80 percent. Sierra Leone's Free Health Care Initiative, launched in 2010, removed user fees for pregnant women and pushed facility delivery up sharply. Chile cut maternal deaths in half since 2000 in part by expanding midwifery. Ireland, which guarantees postpartum home visits from a public health nurse for up to four years after birth, recorded zero maternal deaths in 2022. US states that protected abortion access saw maternal mortality fall 21 percent relative to states that restricted it. None of these outcomes are mysterious. They are the product of choices.
A maternal death is almost always the result of several failures in sequence. A woman does not make it to a clinic in time. The clinic does not have blood products or a trained attendant. The referral system sends her to a hospital too far away. No one checks on her in the weeks after birth. Each link in that chain can be fixed. What cannot be fixed from outside is the political will to fix it.
One more data point worth sitting with. Many of the countries with the highest maternal mortality ratios also have the weakest data systems for tracking it. Sierra Leone's death registration data is flagged by the WHO as "unavailable or unusable for quality reasons." Large parts of sub-Saharan Africa rely on verbal autopsies and household surveys rather than vital registration. The 240,000 figure for 2023 carries an uncertainty interval of 208,000 to 280,000, which is a polite way of saying the true number could be 17 percent higher than reported. Women who die in childbirth in rural Chad or South Sudan rarely appear in any global database. Their deaths have to be modeled, not counted.
The findings out of Seattle will not change behavior by themselves. Reports land, receive coverage for a week, and are absorbed into the background noise of global health data. What matters is whether governments and donor institutions respond as if 240,000 deaths a year represent a scandal rather than a statistic. At current trajectories, the 2030 SDG target will be missed by a wide margin. If aid cuts continue at their 2025 pace, the next Lancet paper on this subject will almost certainly report an increase, not a decrease, in global maternal mortality. That would be the first such reversal in thirty-five years.
Something is wrong with a world that can measure, with Bayesian precision, exactly how many women will die over the next five years because of funding decisions made in Washington, Geneva, and Abuja and then proceed to make those decisions anyway. The data is no longer the problem. The data has been clear for decades. Hemorrhage, preeclampsia, and infection kill women who would not die if someone had shown up with the right drug, the right hands, and the right hour.
Whether anyone shows up is the question.