Somewhere in Louisiana, a household sat down on the worst financial day of their life and filled out a federal form. Their income was $1,800 a month. Their debt to a single hospital was $386,758 — and 98% of everything they owed in the world was medical. They signed it under penalty of perjury.

That hospital buys its outpatient medications at a deep federal discount — 25%, 50%, sometimes as much as 70% off — through a program called 340B. And the entire reason that program exists, the reason Congress created it and requires drug manufacturers to participate, is that the savings are supposed to help low-income and vulnerable patients. Patients, in other words, exactly like that family.

So did the savings reach them? Here's the uncomfortable truth at the center of the newest episode of the State of the Patient Podcast: nobody can say. Not the family. Not the public. Not even the federal agency that runs the program.

A program almost nobody can see into

340B has been around since 1992. The design is elegant: a manufacturer sells a covered safety-net provider a drug at a deep discount, the provider bills insurance at the normal rate and keeps the spread, and those savings are supposed to flow into patient benefit — charity care, lower copays, patient assistance, expanded services in communities that don't have other options.

Manufacturer gives the discount. Provider captures the savings. Patient sees the benefit. When it works, it works.

The problem is the last step. In 2024, covered entities purchased more than $81 billion in outpatient drugs under 340B — a 23% jump in a single year, and one of the fastest-growing pieces of the entire U.S. drug spend. Yet hospitals are not required to publicly report how much they save or how much of it reaches patient care. The nonpartisan Government Accountability Office and the HHS Office of Inspector General have both flagged, for years, that the public data are simply insufficient to show the program is meeting its safety-net mission.

And it's invisible to you personally, too. As Terry Wilcox puts it in the episode: "There's no box that says 'I'm a 340B patient.'" A medication could save your hospital $5,000, and you'd pay the same copay and walk out knowing nothing.

Why we went to bankruptcy court

Health policy has argued for twenty years about how often medical bills cause bankruptcy — mostly with dueling surveys that never move anyone. So Patients Rising asked a narrower question, one that can be answered from primary documents: when Americans are pushed into bankruptcy owing hospital money, which hospitals appear as the creditors, and how often?

Bankruptcy court is the right place to ask, because federal law requires filers to list every creditor and every amount owed — under penalty of perjury, in public records. It is one of the very few places the relationship between a specific hospital and a specific struggling patient becomes visible to anyone outside the billing office.

Over the past year, we reviewed 903 consumer bankruptcy filings across six states — Virginia, Wisconsin, Washington, Colorado, Louisiana, and Maine. For every filing, we recorded each medical creditor and amount, then checked each hospital against the federal 340B registry. Every judgment call was made conservatively: when we couldn't confirm a hospital in the registry, we excluded it. Which means the numbers understate the footprint — not overstate it. Every personal identifier was stripped out.

What the filings showed

Of the 903 filings, 406 — 45% — listed a debt to a hospital we could verify as a 340B participant, totaling $6,622,650. Not projections. Line items, sworn to by households listing everything they owed.

And the consistency is the finding. Verified 340B debt appeared in 53% of Virginia filings, 52% in Maine, 37% in Colorado, 34% in Louisiana, 32% in Washington, and roughly two-thirds in Wisconsin. Six states — different geography, different hospital markets, different politics — same pattern. Within each state, the debt clustered in a handful of large nonprofit systems: Aurora in Wisconsin, Northern Light in Maine, Providence in Washington, CHRISTUS and Ochsner in Louisiana, Bon Secours Mercy Health in Virginia, UCHealth in Colorado.

We want to be very clear about what that does and does not mean. We are not saying any of these systems caused any bankruptcy. We're saying their names appear on these schedules, in their own communities, over and over — and no one can see whether the discounts they receive ever reached the people whose names are on the other side of the page.

We'll also tell you the part that cuts against us, because if we didn't, someone else would. Across the verified filings, medical debt was often a substantial minority of a household's total unsecured debt — a median of 15% in Maine to 33% in Washington. This isn't a data set cherry-picked for maximum medical debt. And that's exactly why it's striking that the same nonprofit 340B systems keep recurring anyway. In the most severe filings, though, medical debt dominates — like the Virginia household that owed $82,206 to a single hospital, 88% of everything they owed.

What the record says about itself

Two findings aren't about dollars at all. In Colorado, 40 filers could name only a parent corporation — Centura, CommonSpirit, Intermountain — not the hospital that treated them. Sit with that: a person could not identify which hospital billed them. And if you can't identify the hospital, you can't find its charity-care policy, can't apply, can't appeal. Patients cannot exercise rights they cannot locate.

The second is harder to hear. Under hospital accounting conventions, a bad-debt write-off — including a balance discharged at the end of a bankruptcy — can be counted toward the "community benefit" a nonprofit hospital reports to justify its tax exemption. So a hospital can pursue a low-income patient, send them to collections, push them all the way through federal bankruptcy court, write the balance off at discharge — and report that write-off as evidence of its generosity. If the debt was always going to be written off, why did that family have to pass through bankruptcy court to get there?

Disclosure, not demolition

None of this happens because someone is breaking the law. It happens because of how the program is built — key terms like "340B patient" and "charity care" have no enforceable federal definition, there's no requirement to pass the discount to the patient, and community benefit is self-reported. So the episode lands on three reforms that follow directly from the findings: define "340B patient" in law; require standardized, audited public reporting that separates charity care actually delivered from bad debt written off; and disclose 340B at the point of care — so patients are told before the bill, not after the discharge. Plus one principle most people accept the moment they hear it: no hospital receiving 340B discounts should pursue collections against a patient it hasn't first screened for financial assistance. The screening should come before the debt.

This isn't a call to cut, cap, or dismantle 340B. If the program is doing what it's supposed to — and in many places it is — reporting will prove it. Good programs survive scrutiny. Where the savings aren't reaching patients, we ought to be able to have that conversation honestly, with numbers, instead of shouting past each other for another decade.

Four hundred and six households, in six states, listed a 340B hospital among the creditors they could not pay. We know their incomes and their balances. We know everything about them except the one thing the program was built to guarantee: whether any part of the discount ever reached them. They deserve an answer. So do you.

Read more and get involved

Read the full research: Medical Bankruptcy in America — 340B Hospitals as Creditors in Consumer Bankruptcy (Patients Rising, July 2026) → [WHITE PAPER LINK].

Explore our full bankruptcy series and the Patients' Right to Know campaign → [LINK].

And if you have a story about a hospital bill you never saw coming, or help you were never told existed →  [SHARE YOUR STORY]