Doula and midwifery care statistics: birth outcomes, workforce gaps, and Medicaid coverage

She asked for a midwife at her first prenatal visit and was told the practice did not have one. She asked about a doula and was told Medicaid would not cover it. She delivered with a resident she had never met, an epidural she had not planned for, and a primary cesarean that no one fully explained. When I reviewed her chart two years later, the operative note listed "failure to progress." What it did not list was the absence of continuous support in a crowded labor unit where the nurse was covering three rooms.

Doulas and midwives are not interchangeable, and neither replaces obstetric care when complications arise. They do, however, sit at the center of a live policy story: state Medicaid programs are expanding doula reimbursement, midwife-attended births have climbed, and researchers keep finding lower intervention rates under midwife-led and continuously supported models. Journalists covering those bills need clean national numbers. This article gathers the ones that hold up under primary-source scrutiny.

What follows is the latest reachable data on midwife-attended births in the United States, outcome differences in midwife-led care, doula-supported birth outcomes and cost estimates, state Medicaid doula coverage, workforce size, and how U.S. midwifery density compares with peer countries.

How many U.S. births are midwife-attended

The cleanest national figure comes from the Government Accountability Office. In a 2023 report, GAO analyzed CDC birth data and found that midwives attended 12% of all U.S. births in 2021: more than 430,000 births, or about 11.9% when reported to one decimal place. That share is higher than two decades ago, but it remains a minority model of care in a system still built around physician-led hospital delivery.

Most midwife-attended hospital births are attended by certified nurse-midwives (CNMs). Certified midwives and certified professional midwives account for a smaller share and practice under state laws that vary sharply. Community birth (planned home birth and freestanding birth centers) remains about 2% of U.S. births.

Context matters when you quote the 12% figure. Midwives disproportionately care for low-risk pregnancies, and selection into midwifery care is not random. The attendance rate is still the right national denominator for journalists: of every 100 U.S. births, roughly 12 list a midwife as the attendant of record. Medicaid was the primary payer for 40.2% of U.S. births in 2024, according to CDC/NCHS data. Federal Medicaid law has long required state programs to cover CNM services, but network adequacy, hospital credentialing, and low reimbursement still limit real access.

Midwife-led care and birth outcomes

Outcome data on midwife-led care are strongest for continuity models, in which the same midwife or small team follows a woman from pregnancy through birth into the early postpartum period, rather than a one-time midwife presence at delivery. The landmark synthesis is the Cochrane review by Sandall and colleagues.

In the 2016 Cochrane review of 15 trials with 17,674 women, midwife-led continuity models were associated with lower rates of regional analgesia (average RR 0.85), instrumental vaginal birth (average RR 0.90), and preterm birth before 37 weeks (average RR 0.76). Women were more likely to have a spontaneous vaginal birth (average RR 1.05). The review found no clear difference in cesarean rates in that version of the analysis, and no increase in adverse maternal or neonatal outcomes for the measured endpoints. Breastfeeding initiation did not differ significantly between groups in the pooled trials.

A 2024 Cochrane update by Sandall and colleagues, covering 17 studies and 18,533 women, refined the picture. Women in midwife continuity models were less likely to experience a cesarean birth or instrumental birth and more likely to have a spontaneous vaginal birth. They also reported more positive experiences of care across pregnancy and the postpartum period, with cost savings noted in the antenatal and intrapartum periods. On preterm birth, the updated review concluded midwife continuity models may make little or no difference. That is a softer finding than the 2016 RR of 0.76. The shift is worth quoting carefully: the preterm signal is weaker in the newer synthesis than many secondary sources still claim.

State-level observational work points in a similar direction. Vedam and colleagues, writing in PLOS ONE in 2018, built a Midwifery Integration Scoring System for all 50 states. Higher integration scores tracked with more spontaneous vaginal birth and breastfeeding and with lower rates of cesarean birth, preterm birth, low birth weight, and neonatal death. The authors estimated that midwifery integration explained nearly 12% of state variation in neonatal mortality, separate from the large contribution of race. Correlation is not causation, and states that integrate midwives may also fund broader maternal health infrastructure. Still, the gradient is hard to ignore.

Against a national cesarean rate of 32.4% in CDC final data for 2024, any care model that reliably lowers operative delivery without raising harm attracts policy interest. For hospital-level variation, see cesarean section rates statistics.

Doula support: outcomes and cost savings

Doulas provide continuous nonclinical support: physical comfort and emotional presence, plus advocacy, across the perinatal period. They do not manage labor medically. The strongest experimental evidence on continuous labor support comes from the 2017 Cochrane review by Bohren and colleagues, which pooled 26 trials with more than 15,000 women across 17 countries.

Women allocated to continuous support were less likely to have a cesarean birth (average RR 0.75) and more likely to have a spontaneous vaginal birth (average RR 1.08). Labors were shorter by about 0.69 hours on average. Use of any intrapartum analgesia was lower (average RR 0.90), as was instrumental vaginal birth (RR 0.90). Babies were less likely to have a low five-minute Apgar score (RR 0.62). The review found no evidence of harm. Subgroup analyses suggested the cesarean reduction was strongest when the support person filled a doula-like role rather than hospital staff or a family member alone, though those subgroup findings should be treated as exploratory.

U.S. Medicaid-specific numbers come largely from Katy Kozhimannil's research group. In a 2016 Birth study of 1,935 doula-supported Medicaid births compared with 65,147 regional Medicaid hospital births, preterm rates were 4.7% with doula support versus 6.3% regionally, and cesarean rates were 20.4% versus 34.2%. After adjustment for age, race-ethnicity, and chronic conditions such as hypertension or diabetes, doula care was associated with 22% lower odds of preterm birth (AOR 0.77, 95% CI 0.61-0.96). Among full-term births, adjusted odds of cesarean were substantially lower (AOR 0.44).

Cost modeling in the same paper found that reimbursing doula services at an average of about $986 (range $929-$1,047 across states) would often be cost-saving or cost-effective for Medicaid programs, driven by fewer preterm and cesarean births. In one regional simulation, doula-supported Medicaid deliveries were projected to save $58.4 million and avert 3,288 preterm births annually. Those are model outputs, not audited budget savings, but they are the figures most frequently cited in state fiscal notes.

Breastfeeding is another measurable endpoint. In a 2013 Journal of Midwifery & Women's Health analysis, Kozhimannil and colleagues reported near-universal breastfeeding initiation among doula-supported births: 97.9%, compared with 80.8% in the general Medicaid population studied. Duration of exclusive breastfeeding is less consistently measured. For national continuation rates after discharge, see breastfeeding and lactation support statistics. Survey data cited in this literature put doula use at roughly 6% of U.S. births in the early 2010s; continuous support is still the exception on most labor units.

Medicaid doula coverage expansion

This is the policy story reporters cover state by state. As of March 2026, the National Academy for State Health Policy reported that 26 states and Washington, D.C., reimburse doula services in their Medicaid programs. That is an increase of 14 states since NASHP's April 2024 analysis. More than half of states now have a benefit that barely existed a decade ago.

Implementation details diverge quickly. Per NASHP, all states with a statewide Medicaid doula benefit cover services under the preventive services category through a state plan amendment and allow independent billing with a Type 1 National Provider Identifier. Twenty-two states and D.C. also allow group billing. Seventeen states reimburse doula services through 12 months postpartum. Eight have issued a statewide standing recommendation so beneficiaries do not need an individual clinician referral. Labor and delivery reimbursement ranges from about $459 to $1,500 across states as of March 2026.

Coverage on paper is not a functioning workforce. Low rates, costly training mandates, and slow credentialing often mean few claims filed. The journalists who get this story right distinguish between "state X passed a doula bill" and "state X has active Medicaid-enrolled doulas billing for births." Maternal mortality remains higher in the United States than in peer countries; continuous support and midwifery integration are among the few scalable interventions with evidence of fewer cesareans and better experience. See maternal mortality statistics.

The U.S. midwifery workforce

Workforce numbers depend on which credential you count. The Bureau of Labor Statistics Occupational Employment and Wage Statistics program counted 6,960 employed nurse midwives in May 2023, with a mean annual wage of $131,570 and a median of $129,650. That figure excludes self-employed midwives and most non-nurse midwifery credentials, so it understates the full midwifery workforce.

The American College of Nurse-Midwives, drawing on certification and public data, estimates roughly 14,000 midwives in the United States when those not in clinical practice are included, and about 4 midwives employed per 1,000 live births. With more than 3.7 million U.S. births a year, reaching a World Health Organization minimum benchmark of 6 midwives per 1,000 live births would require at least 22,000 midwives by ACNM's estimate, a gap of about 8,200. Even at that minimum, the United States would still staff midwifery more thinly than many high-income countries with better maternal outcomes.

Density is uneven. BLS data show the highest employment levels in California and New York. Scope-of-practice laws, hospital privileges, and Medicaid parity all shape whether a licensed midwife can actually attend births. Restricted-practice states require physician supervisory contracts that many rural communities cannot supply. That gap compounds maternity care deserts. GAO also flagged clinical placement shortages and education costs as pipeline barriers.

How the U.S. compares internationally

International comparisons are imperfect (definitions of "midwife" and counting methods differ), but the direction of the gap is consistent. A 2020 Commonwealth Fund analysis of OECD data found that the United States and Canada had the lowest overall supply of midwives and obstetrician-gynecologists combined among 11 high-income countries: about 12 to 15 providers per 1,000 live births, versus two to six times that supply elsewhere. In the United States, obstetricians outnumber midwives. In most peer countries, the ratio is reversed.

The Commonwealth Fund and ACNM both cite roughly four midwives per 1,000 live births in the United States. Countries such as Sweden and the United Kingdom staff midwifery at several times that density, as does Australia, and midwives there provide a large share of prenatal care and delivery for healthy pregnancies. In the Netherlands, midwives also attend a substantial share of home births (about 13% of Dutch births in Commonwealth Fund reporting).

High-income countries with strong midwifery systems train midwives at scale, pay them, grant hospital privileges, and build transfer protocols for complications. The U.S. pattern is the inverse: high intervention rates, physician-dominant staffing, and midwifery treated as optional rather than core infrastructure.

What the evidence does not settle

A few cautions belong next to every quotation-ready statistic above.

First, selection bias is real. Women who seek midwifery or doula care often differ from those who do not, even inside Medicaid. Randomized trials reduce that bias for midwife continuity and continuous support; observational Medicaid doula studies cannot eliminate it completely. Kozhimannil's team is explicit about this limitation.

Second, "midwife-led care" is not one intervention. Caseload continuity, team midwifery, CNM hospitalists, and community CPMs are different models with different risk panels. Pooling them muddies effect sizes.

Third, the 2024 Cochrane update softened the preterm-birth conclusion for midwife continuity compared with 2016. Anyone still writing that midwife-led care "cuts preterm birth by 24%" without noting the updated review is overstating the current evidence.

Fourth, doula Medicaid expansion is young. We do not yet have multi-state claims analyses showing that every new benefit produced the cesarean and preterm reductions predicted by earlier models. Reimbursement levels, training rules, and managed-care contracting will determine whether the policy matches the research.

None of that voids the core finding. Continuous support and midwife continuity improve experience, reduce some interventions, and look favorable on cost when cesarean and preterm events are priced honestly. The open question is not whether the model works in trials. It is whether U.S. payment and hospital systems will let it operate at scale. Birth attendant and support choices are worth raising at the first prenatal visit; our ovulation calculator and period calculator can help with cycle timing while those care decisions are made.

Frequently asked questions

What percentage of U.S. births are attended by midwives?

About 12% of U.S. births were attended by midwives in 2021, according to a 2023 GAO analysis of CDC data (more than 430,000 births). The share has risen over two decades but remains far below many high-income countries where midwives attend a majority of low-risk births. Most U.S. midwife-attended births occur in hospitals with certified nurse-midwives.

Does midwife-led care lower cesarean rates?

The 2024 Cochrane review of midwife continuity models found women were less likely to have a cesarean or instrumental birth and more likely to have a spontaneous vaginal birth, with more positive care experiences. Earlier 2016 pooled results showed clearer reductions in preterm birth and instrumental delivery than in cesarean rates specifically. Effects depend on the model of care and transfer systems.

How does doula support affect birth outcomes?

A 2017 Cochrane review of continuous labor support found lower cesarean risk (average RR 0.75), higher spontaneous vaginal birth, shorter labor, and no identified harms. In U.S. Medicaid research by Kozhimannil et al., doula-supported births had cesarean rates of 20.4% versus 34.2% regionally and preterm rates of 4.7% versus 6.3%, with 22% lower adjusted odds of preterm birth.

How many states cover doulas through Medicaid?

As of March 2026, 26 states and Washington, D.C., reimburse doula services in Medicaid, according to the National Academy for State Health Policy. That is 14 more states than in April 2024. Coverage scope, postpartum duration, reimbursement rates, and training rules vary widely, so enrollment and claims volume differ even among states with a benefit on the books.

How many midwives practice in the United States?

BLS counted 6,960 employed nurse midwives in May 2023. ACNM estimates about 14,000 midwives total when nonclinical and other credentials are included, or roughly 4 midwives per 1,000 live births. Reaching a WHO minimum of 6 per 1,000 births would require at least 22,000 midwives, a gap of about 8,200.

How does the U.S. midwifery workforce compare internationally?

The United States has about 4 midwives per 1,000 live births and a physician-dominant maternity workforce. A 2020 Commonwealth Fund comparison found the U.S. and Canada had the lowest combined supply of midwives and obstetricians among 11 high-income countries. In most peer countries, midwives outnumber obstetricians and provide most care for healthy pregnancies.