And Then You're Dead

The Autopsy / Case File · Maternal & Infant

The Undetectable Ones

Congenital anomalies are the single leading cause of infant death in America. A meaningful share of them are never found until the child is already in trouble.

One in 33 babies born in the United States each year arrives with a birth defect. Most of those children live. The ones who don't account for a fifth of all infant deaths nationwide — more than any other single cause, including prematurity and SIDS. Prenatal screening has narrowed the gap between a defect existing and a defect being found before it's fatal, but the gap hasn't closed. It's still wide enough, for the most common category of birth defect, that a third to two-thirds of cases go undetected until birth.

And Then You're Dead · September 2026 · 8 min read

Join the discussion →
A newborn infant in a hospital incubator, wrapped in a blanket with a nasal cannula and monitoring leads attached
An infant in a hospital incubator. Congenital anomalies — birth defects — are the leading cause of infant death in the United States, ahead of prematurity, SIDS, or any other single cause.

Birth defects are common enough that most people know a family touched by one, and rare enough that they still register as unlucky, unusual, something that happens to someone else's pregnancy. The actual numbers sit closer to routine than either instinct suggests: about 1 in every 33 babies born in the U.S. each year — roughly 120,000 infants — arrives with a structural or functional anomaly present from birth. The overwhelming majority survive, many after surgery, some without ever needing intervention at all. The minority who don't survive make congenital anomalies the leading cause of infant death in the country, responsible for about 1 in 5 of all deaths in the first year of life.

What Actually Kills, By Category

Birth defects aren't one thing; they're a category covering everything from a cleft palate that requires a single corrective surgery to a heart with chambers that never fully separated in utero. The distribution of which defects occur most often doesn't track cleanly with which defects kill most often, but congenital heart defects sit at the top of both lists — they're the single most common category of major birth defect, accounting for roughly 27% of the total, and they're disproportionately represented among the fatal cases.

0 in 5
of all U.S. infant deaths attributable to a congenital anomaly — the single leading cause of infant death nationally
0%
of U.S. live births affected by a congenital heart defect, the most common major birth defect category, of which roughly a quarter are classified as critical

Behind congenital heart defects, musculoskeletal anomalies account for about 18% of major birth defects, genitourinary anomalies about 15%, orofacial clefts about 5%, and neural tube defects — including spina bifida and anencephaly, among the most severe and often fatal category — about 2%. That last figure is small in share but outsized in consequence: neural tube defects carry among the highest case-fatality rates of any major birth-defect category, a legacy of a condition modern prenatal folic acid fortification has reduced in incidence but not eliminated.

The Detection Gap, In One Number

Prenatal screening exists precisely to catch these conditions before birth, giving families and clinicians time to plan delivery at a hospital equipped for immediate surgical intervention, or in some cases to treat a condition in utero. For congenital heart defects specifically — the largest single category — that screening system catches meaningfully less than half of cases in routine practice. A large U.S. cross-sectional study estimated the overall prenatal detection rate for congenital heart defects at just 34%; a separate international cohort study put detection of critical congenital heart defects specifically closer to 50%.

0%
estimated overall U.S. prenatal detection rate for congenital heart defects, per a large cross-sectional study — meaning roughly two-thirds go unfound before birth
0%
sensitivity of newborn pulse oximetry screening for detecting critical congenital heart defects after birth, the safety-net test that catches many cases prenatal ultrasound misses
A baby with an undetected critical heart defect can look entirely healthy at birth, pass an initial newborn exam, and go home — because the defect that will eventually kill them doesn't announce itself until the specific physiological transition it depends on goes wrong, sometimes days later.

That's the clinical logic behind pulse oximetry screening, a low-cost, noninvasive test now standard in virtually every U.S. birthing hospital: a sensor placed on a newborn's skin measures blood oxygen saturation, flagging levels low enough to suggest a heart defect the prenatal ultrasound missed. On its own, pulse oximetry catches about 76% of critical congenital heart defects, with a specificity above 99.9% — few false alarms, but also, by definition, roughly a quarter of critical cases it doesn't catch either. Layered together — prenatal ultrasound, newborn pulse oximetry, and a standard physical exam — the combined detection rate for critical congenital heart defects climbs as high as 96%. It's the layering that does the work; no single screening step in the sequence comes close to that number alone.

The Cases That Fall Through Every Layer

The remaining gap is where the fatal cases concentrate disproportionately. A defect missed by prenatal ultrasound, not flagged by pulse oximetry because its physiological signature doesn't affect oxygen saturation in a way the test detects, and not obvious on a standard physical exam, sends a family home with an infant who appears healthy and isn't. Some of these infants are brought back to an emergency department in acute distress once the defect's effects become physiologically unavoidable — a scenario clinicians specifically design screening protocols to prevent, and one that still occurs often enough to be a recognized pattern in pediatric cardiology, not an outlier.

The consequence isn't abstract. A congenital heart defect caught prenatally can be managed with a delivery plan built around it: a hospital with a cardiac surgical unit on-site, a care team present in the delivery room instead of racing to respond to a decompensating newborn hours or days later. A defect caught only after a baby has already gone into crisis loses that planning window entirely, and outcomes for emergency, unplanned intervention are measurably worse than for the same defect managed on a planned timeline.

A Disparity That Widened, Not Closed

CDC's most recent long-run analysis of infant mortality attributable to birth defects, covering 2003 through 2017, found real progress and a persistent racial gap sitting inside it. The infant mortality rate attributable to birth defects fell 10% overall across those fifteen years — but the decline wasn't evenly distributed, and the underlying disparity between Black and white infants didn't close. It widened, slightly, even as both groups' rates fell.

0
infant deaths attributable to birth defects per 10,000 live births among infants of Black mothers in 2017, versus 9.9 among infants of white mothers
0%
how much higher the birth-defect infant mortality rate was for Black infants than white infants in 2017 — up slightly from a 32% gap in 2003, despite both rates declining

The rates by group tell the more precise version of that story: infant mortality attributable to birth defects fell 12% among infants of white mothers and 11% among infants of Black mothers between 2003 and 2017 — genuinely close declines in absolute terms — while infants of Hispanic mothers saw only a 4% decline over the same period. Progress happened broadly. The starting gap was large enough, and the declines similar enough in scale rather than concentrated among the groups furthest behind, that the relative disparity between Black and white infant mortality from birth defects was, if anything, slightly wider at the end of the study period than at the beginning. Screening access, prenatal care quality, and the timing of diagnosis all plausibly contribute to that pattern, though CDC's own reporting is explicit that the analysis wasn't designed to isolate which factor drives the gap most.

And Then You're Dead

The technology to find most fatal birth defects before delivery already exists, and it's improving — 96% detection when every screening layer works together is a real number, not an aspiration. But "every layer working together" is a description of an ideal case, not the median one, and the defects that survive all three layers unnoticed are disproportionately the ones a family finds out about only once a newborn who looked perfectly fine stops looking that way.

That's it. That's the whole thing.

Comments

Loading…

Comments are anonymous and moderated.

Sources

Photo: "Mcmaster NICU infant 6978.jpg" by Peter K Burian, via Wikimedia Commons (CC BY-SA 4.0), cropped from the original.