What Our First Tennessee's Medicaid Doula Pilot Taught Us About What Birth Equity Actually Requires
For fifteen months, the Tennessee Doulas Association ran a state-funded pilot that did something the traditional doula model in this state had never really done: put trained birth support in front of Medicaid-eligible and uninsured women, with a deliberate focus on Black and refugee moms who face the highest risk in Tennessee's maternal health outcomes.
Our goal was to serve 35 families. We came out of it having served 101 women — more than our goal, and well past the number we thought was realistic once a late start ate into our timeline. We came out of it with data that looks good on paper: a majority-vaginal birth rate, almost no NICU or ICU stays, and one save that still gives our doulas chills to talk about. A mom sent home twice by a hospital in two days was pushed by her doula, over a holiday weekend, to go back a third time. She had a placental abruption and an emergency C-section. She survived a hemorrhage that could have killed her, but her baby did not survive. Two doulas walked with her through that grief, and through her recovery, because that is what continuous support means — it doesn't end when the outcome isn't the one you hoped for. They say to be grateful we saved the mom, but we feel that we could have saved the mom and baby in a different system.
That story is the whole case for doula care in one paragraph. It's also, if we're honest, not the story this pilot is really about. The story this pilot is about is everything that had to go wrong, or almost go wrong, or quietly fail, before that mom ever got a doula who would listen to her.
The system was not waiting for us
We expected partnership. We had lined up 14 local organizations before the grant even started. Three of them actually referred us a client. The rest were "too busy," didn't return calls, or — and this is the part worth sitting with — turned out to be actively skeptical of doulas. We spent real time in this pilot re-explaining, to clinical staff, what a doula does and doesn't do, because the bias against community-based birth support is still alive in the rooms where referrals get made. Only one clinic, Vanderbilt's Birth Center, actually opened its doors to us — let us into staff meetings, helped us make a video for their patients. One, out of 14 provider relationship we tried to build.
Most of our clients didn't come from a warm handoff at all. They came from Google. A mom searching on her own, at whatever hour she had a free minute, finding our website and filling out a form. In a program explicitly designed to reach women who are hardest to reach, self-selection through search-engine discovery was our biggest referral channel. That's not a pipeline. That's a gap where a pipeline should be.
And that gap had a cost we can measure. We wanted clients enrolled by 20 weeks, when doula education and trust-building actually has room to work. Instead, because word traveled slowly and the grant itself started late, most women joined us well past that point — often not until the third trimester, the same pattern we see with clients who pay full price and simply procrastinate, except these were women with far less slack in their lives to spare.
If this work is going to reach people early enough to matter, a program needs 10 months to build referral relationships before it ever opens intake, and 18 months to actually run — not the compressed runway we were given.
"Free" doula care ran into a system that wasn't built for the women who need it most
Some of what we learned doesn't fit neatly into a grant report line item, but it's the most important part of what this pilot surfaced.
Clients don't always want a doula that looks like them. We assumed refugee moms would want a doula from their own community. In Nashville, that assumption was often backwards. Women who spoke English frequently asked for a Caucasian, American doula — not because they didn't trust their community, but because their community was too tight-knit, and they didn't want someone from it knowing the most intimate details of their pregnancy and birth. Cultural matching, the thing funders and programs default to as the obvious right answer, is not automatically what safety looks like to the person living it. In our African American clients, they wanted a caucasian doula because they felt like a doula like that would get more respect from a system that is primarily caucasian - and they were not wrong.
Exploiting is happening. We ran into a clinic, Clinica Hispana, billing TennCare patients out-of-pocket for visits that were covered by TennCare, telling women it was illegal to switch providers, and telling them a vaginal birth "wasn't possible because they were in America." Our doulas educated those moms on their actual rights and helped them find new care — and the clinic was not happy about it. That's not a hypothetical bias. That's a documented instance of a Medicaid patient being misinformed and financially exploited by the exact system that was supposed to be caring for her, caught only because a doula was in the room asking questions.
Clients lack resources. We met clients living in an attic with no bed, navigating homelessness, being put in jail. and clients with no car, at all. Our doulas were not prepared for the heartache and the feeling that they were unprepared to serve. We were shocked to find creative client solutions like many clients calling EMS to transport them when in labor. This is crazy to use emergency services to transport for a non emergency situation. There has to be a better way.
They did not want our education. We also offered a free virtual parenting class as part of this grant. Zero pregnant moms attended, despite many saying they wanted to. Our read: when you are fighting to keep a job and hold your life together, a class is a luxury, not a priority — no matter how free it is or how much you meant it when you signed up.
And then there was ghosting — clients who would engage, accept support, and then simply disappear. Our doulas had never dealt with this before, because paying clients who've invested money in their care don't tend to vanish. It rattled our team badly enough that we watched our typical three-year doula burnout timeline collapse to three months. We started this grant expecting to run it with four to six doulas. We finished having cycled through 26, because the emotional cost of pouring yourself into clients who weren't always ready to be poured into, on top of new documentation burdens doulas weren't trained for and in some cases couldn't read English well enough to complete, was real and it was heavy.
None of this is a doula-training problem. The doulas were well trained to serve clients with money, resources, and education. It's what happens when a support model designed around a clientele with insurance, income, and a car gets pointed at the population that actually needs it most.
Why this belongs in front of other advocates, not just funders
It would be easy to write this pilot up as a success story: exceeded our enrollment goal, strong birth outcomes, one dramatic save. All of that is true and it matters. But if the only story that leaves this pilot is "doulas improve outcomes," we will have wasted the hardest-won part of what 26 doulas and 101 families actually taught us.
What this pilot proved is that access is not just a funding problem — it's a trust problem, a timing problem, and a bias problem, stacked on top of each other. Clinics that won't refer to doulas because of institutional skepticism. Providers willing to misinform Medicaid patients about their own rights. A referral network of 14 partners that couldn't figure out how to refer.
A model built for cash-pay clients that had to be substantially rebuilt, in real time, to serve women without cars, without stable housing, without English fluency, without the freedom to plan a birth 30 weeks out. That is a lot to navigate for a big hearted doula with a traditional training.
None of that shows up in a bar chart. All of it is the actual work of birth equity.
If Tennessee wants programs like this to succeed — and the data says they should exist — the next iteration needs a longer runway before intake even opens, deliberate investment in the clinical partnerships that didn't materialize this time, a documentation system built for doulas rather than retrofitted onto them, and provider-facing education aimed squarely at the bias we ran into over and over. On a positive note, we did discover the CHANT program and other new resources we didn't know existed when we started. We also found a CRM that finally made data collection workable, eight months in. We learned that cultural matching isn't the safety net we assumed it was. That's what a real pilot is supposed to produce — and now that we know it, the work is making sure it doesn't have to be relearned by the next program that tries this.
Our biggest take away is that your can't just throw traditionally trained doulas into serving a new demographic and expect it to go well. It is kind of like painting a broken down car a new color and hoping it can run well. TennCare doulas need specialized training and support to process their doula experience. They need to learn about all the amazing resources that TN has to offer families. Instead of living on islands, we need to learn to live in community with space for doulas. Don't expect the doulas to change to the system, but let doulas be the change for a broken system. Welcome their input and give them support to walk through the journey with their clients.
The Tennessee Doulas Association's Doula Pilot ran March 2024 through June 2025, funded by a state grant, serving 101 Medicaid-eligible and uninsured pregnant women across 13 Tennessee counties with prenatal, birth, and postpartum doula support.
Comments