Methodology

This is an overview of how data on this site is sourced, defined, and time-bounded. Every figure traces back to CDC, AHRQ, March of Dimes, SAMHSA, WHO/UN, or peer-reviewed cohort studies — no advocacy estimates. Where multiple methodologies are defensible, the choice is documented below.

Map — maternal mortality by state

The map shows U.S. maternal mortality at the state level, with each tooltip surfacing the rate, the comparison to the national average, and a recognizable country whose rate matches.

Definition

Maternal death per the WHO/CDC standard: death of a woman during pregnancy or within 42 days of the end of pregnancy, from a cause related to or aggravated by pregnancy (ICD-10 maternal-cause codes). This is the same definition WHO and the UN use for international rankings, and what the U.S. reports for SDG 3.1.

Time window

Pooled 2019–2023 — a 5-year aggregate of CDC NVSS data. Pooling smooths small-state year-to-year volatility and the pandemic spike (2021) plus post-pandemic dip (2023); individual years would make Vermont, Maine, the Dakotas, etc. unreadable.

State values

From America's Health Rankings 2025 edition, which republishes CDC NVSS state-level pooled rates. Three small-population states (Maine, Vermont, Wyoming) had counts the source suppresses for stability — those show as “rate not published” on the map. Delaware also suppresses in AHR; we substituted the Delaware MMRC's own published rate (2020–2024, 17.3 per 100,000) and footnoted the swap in the tooltip.

U.S. average

23.5 per 100,000 live births, computed directly from CDC NVSS counts: 4,306 maternal deaths / 18,284,565 live births × 100,000 over 2019–2023. The state map and the headline stat both use this figure for methodological consistency.

Country comparison

From the WHO/UN MMEIG (Maternal Mortality Estimation Inter-Agency Group: WHO + UNICEF + UNFPA + World Bank + UNDESA), 2023 point estimates (released April 2025). Same 42-day-window definition. The country we pair with each U.S. state is hand-picked for recognizability, not just numerical closeness — when a U.S. state at MMR 17.3 matches Latvia and Egypt about equally, we pick Egypt because it is more recognizable to an American audience.

Sources

Back to the map ↗

Iceberg — the deeper toll

The pyramid shows that maternal deaths sit at the top of a much larger cascade of harms tied to pregnancy and birth. Six layers, area-proportional to U.S. annual counts, all 5-year-pool annual averages (2019–2023) where the data supports it.

Time window

5-year pool 2019–2023, matching the state map.

Layer-by-layer

1. Mothers die — 870/year

Pregnancy-related deathsper CDC PMSS (Pregnancy Mortality Surveillance System): deaths during pregnancy or within 1 year postpartum, from a cause related to or aggravated by pregnancy, determined by CDC expert review. This is a broader window than the WHO 42-day standard used on the map — the wider window catches late postpartum cardiovascular, mental-health, and overdose deaths that occur 6–12 months out and which the 42-day cutoff misses. For the iceberg's “deaths tied to being pregnant” framing, PMSS is the cleanest fit.

2. Babies die — 41,000/year

Stillbirths (fetal deaths at ≥20 weeks gestation) plus infant deaths (deaths before age 1). Both directly published annually in CDC NVSS reports. 2023: 20,005 stillbirths + 20,162 infant deaths = 40,167; the 5-year pool annual average is essentially the same (~41,000) because both metrics held remarkably steady through the pandemic.

3. Mothers nearly die — 60,000/year

Severe maternal morbidity (SMM).CDC's public-facing summary cites “as many as 60,000 women each year.” The annually-published delivery-stay surveillance count (AHRQ HCUP, 20-indicator standard) is approximately 33,000; CDC's broader 60,000 figure includes near-fatal events that occur in the postpartum window beyond the delivery hospitalization, which extended-window peer-reviewed studies confirm pushes the count into the 50,000–60,000 range.

4. Babies need intensive care — 430,000/year

Composite of (a) NICU admissions at birth and (b) significant non-NICU hospitalizations during the first year of life, de-duplicated for overlap (NICU graduates have higher year-1 readmission rates). NICU admissions alone: ~352K (NCHS Data Brief 525, 2023). Adding non-NICU year-1 hospitalizations from AHRQ HCUP's Kids' Inpatient Database, with overlap correction (~20% of NICU graduates rehospitalized in year 1), yields the ~430,000 figure. Lower bound: ~400K (~1 in 9 babies); central: ~430K (~1 in 8 babies).

5. Mothers in mental-health crisis — 720,000/year

Mothers experiencing a perinatal mental-health condition — postpartum depression (CDC PRAMS, ~12.5% of births), postpartum anxiety (peer-reviewed meta-analyses, ~6–8%), or substance-use disorder during pregnancy (SAMHSA NSDUH, ~5–7%) — composite, de-duplicated for overlap. The resulting “1 in 5 mothers” figure is the standard public-health framing in the academic and policy literature on perinatal mental health.

6. Mothers carry lasting physical effects — 1,200,000/year

Mothers with a persistent physical complication of pregnancy or birth — urinary or fecal incontinence, pelvic floor dysfunction, chronic perineal or cesarean pain, persistent hypertension after preeclampsia, gestational diabetes progressing to type 2 diabetes, etc. Composite from peer-reviewed cohort studies, de-duplicated using published co-occurrence rates. The conservative 1-in-3 floor (~1.2M) draws on the best-attested pelvic-floor cluster from the NIH-funded MOAD cohort plus modest add-ons; the Maternal Health Study's direct measurement of “persistent physical health problem at 12 months” lands closer to 1-in-2 (~1.8M). We use the conservative number.

Why these six layers

Each layer captures a distinct, measurable harm tied to pregnancy or birth in the U.S. The cascade is built so each layer is bigger than the one above — visualizing how rare the deaths are relative to the much larger population of women and infants whose lives are reshaped by the same underlying gaps in care.

Back to the iceberg ↗

Causes of pregnancy-related death

The donut breaks pregnancy-related deaths down by underlying cause and surfaces the headline preventability figure.

Cause breakdown

From CDC's Maternal Mortality Review Committees (MMRCs), aggregated by the federal ERASE-MM program across 36 states, 2017–2019 data (Trost et al.). The most recent fully-published CDC breakdown is 2017–2019; 2022 and 2023 data have not been released yet.

Why not the 2019–2023 window of the other charts? CDC publishes MMRC data year-by-year as case reviews complete, and 2022 and 2023 are not yet released. The 2021 release (August 2025) reports headline numbers but doesn't publish full sub-category percentages, and 2021 is heavily distorted by COVID — infection temporarily ranked #1 at ~33% that year. Across the years that are fully published (2017–2019, 2020), the structural picture is stable: mental health conditions, hemorrhage, cardiovascular events, embolism, and cardiomyopathy each account for ~9–23% of deaths, with no single dominant cause.

Preventability

87% of pregnancy-related deaths are preventable per the CDC Maternal Mortality Review Information Application (MMRIA), 2021 data, 46 states, published August 2025. This determination is made by state MMRC clinical reviewers based on detailed case review.

Sources

Back to the causes chart ↗