๐Ÿš‘ EMS / Patient Billing

Congress Banned Surprise Bills in 2022. It Left Out the Ambulance.

About 1.5 million privately insured patients a year take an emergency ground ambulance ride out of network and are still at risk of the bill. The No Surprises Act covers the emergency room, the anesthesiologist you never met, and the helicopter. It does not cover the ground ambulance that got you there, and the federal committee that was supposed to fix that sent its report to Congress in August 2024 and has heard nothing since.

A white and red ambulance with emergency lights flashing outside a hospital emergency entrance at night, a paramedic wheeling a stretcher toward the doors

The Problem

On January 1, 2022, the No Surprises Act took effect and ended one of the most hated practices in American health care: the bill from a provider you never chose. Emergency room care came under the ban. So did out-of-network clinicians working at in-network hospitals, and air ambulances. Ground ambulances did not. The Centers for Medicare and Medicaid Services says so plainly on its own consumer guidance page: ground ambulance services are not covered by the billing protections in the No Surprises Act.

Congress did not leave ground ambulances out by accident. One reason it held back, cited in the Health Affairs ownership study and in contemporary reporting, was local governments' fear that federal rate-setting would cut the revenue their fire departments earn from ambulance service. Another was simple disagreement over what insurers should pay if patients were taken out of the equation. So Congress created a federal advisory committee to study the problem instead. The Ground Ambulance and Patient Billing Advisory Committee met through 2023, voted on recommendations in November 2023, and sent its report to Congress on August 28, 2024; the report itself (dated March 29, 2024; transmitted August 28, 2024) is published by CMS. Its primary recommendation was a ban on balance billing for 911 and interfacility emergency transport, plus a cap on patient cost-sharing at the lesser of $100, adjusted for inflation, or 10 percent of the payment rate the committee set out in Recommendation 12. As of late 2026, Congress has passed no law acting on it. The report is two years old. The law it called for is still unwritten.

And the bills keep coming. Four studies, five estimates, and five different answers, because they measure different populations, years, and definitions of the problem:

StudyPopulationYearsWhat it measuredRate
Chhabra et al., Health AffairsOne national insurer2013-17All ambulance rides (ground + air) with a potential surprise bill71%
Adler et al., Health AffairsCommercial claims2014-17Emergency ground transports with a potential surprise bill28%
Gong et al., JAMA Network OpenEmployer plans, 2.03 million services2015-20Ground ambulance services billed out of network54.8%
FAIR Health brief (emergency lines)Private claims2018-22Emergency ground claim lines out of network, 202262.0%
FAIR Health brief (all lines)Private claims2018-22All ground claim lines out of network, 202259.4%

Those studies disagree on the exact share because they measure different populations and years. They agree on the direction: roughly half of privately insured emergency ground rides are out of network. A Health Affairs study of a large national insurer put the median potential surprise bill at $450 for ground transport, with an aggregate impact of $129 million a year from that one insurer alone, whose market share the study does not disclose, so the national total is unknown but certainly larger. A JAMA Network Open study put the mean total financial burden at $434.70 per out-of-network service against $132.21 in network, an out-of-network ride costing the patient 3.3 times as much.

You do not choose your ambulance. The ambulance chooses your bill.

Share this line on X

Pricing is a black box. Nobody publishes it: not the agency that sets the charge, not the insurer that pays part of it, and not the county that, in some places, sets the rate by ordinance.

FAIR Health puts the average 2020 base charge for emergency transport at $940 for basic life support and $1,277 for advanced life support, before mileage, while the average allowed amount was $522 for basic life support and $758 for advanced life support. But averages hide the spread, because the Health Affairs ownership study found that private-sector ambulances charge substantially more than public-sector ones, and private-equity or publicly traded owners charge more still. Global Medical Response is the largest private operator: parent of American Medical Response, it operates across about 1,400 US counties and completed a $5.4 billion refinancing in 2025. KKR combined its Air Medical business with American Medical Response, which it had bought from Envision Healthcare for $2.4 billion in a deal that closed in 2018, and took the combined company public in May 2026 under the ticker GMRS, raising $478.7 million at a roughly $3.35 billion valuation. A patient who gets AMR in one county and a fire-department ambulance in the next can be billed two very different amounts for the same ride. Neither price is published anywhere the patient can see it before the stretcher arrives.

What exists for the patient today? Almost nothing. The ambulance industry's software is built to help agencies bill and collect, not to help patients dispute: Zoll RescueNet, Traumasoft, and AIM sell billing tools to the agency that sends the bill, and ESO is the clinical record system underneath them. Generalist medical-bill negotiators will take an ambulance bill, but they work without ambulance-specific benchmarks. They have no Medicare fee schedule comparison for the ZIP code and service level, and no locally set rate where the county or city publishes one. They have no state-law matrix that says whether the provider can bill more than in-network cost-sharing for that trip, plan type, and provider type, and no dataset of what other patients paid for the same ride from the same agency. They negotiate blind in exactly the market where a benchmark would matter most. Insurers pay the allowed amount and stop, leaving the balance as the patient's problem in a system where 27 states have no ground ambulance surprise-billing protection at all. Georgetown researchers writing for the Commonwealth Fund counted 22 states with some protection in February 2026; Alabama's law took effect on October 1, 2026, bringing the count to 23, with Georgia's law making it 24 on January 1, 2027. Protections vary: some cover emergency trips only, some apply to HMO plans only, and several sunset within a few years. Oregon and Washington let self-funded plans opt in, a reminder that state protection reaches only state-regulated plans unless a state builds the opt-in.

Market Size

Original TAM calculation: start with the patient side. Peterson-KFF estimates ambulances transport about 3 million privately insured patients a year, and about half of the time those ambulances are out of network. That is 1.5 million exposed rides a year. The Health Affairs study gives a median potential surprise bill of $450 per ride. Multiply: 1.5 million times $450 is $675 million a year in patient exposure. That is not revenue. It is the pool the platform works against.

ScenarioExposed rides/yearMedian potential billTotal patient exposurePlatform revenue at 25% contingency, 40% recovery, 30% case resolution
Low (28% potential-surprise-bill rate, Adler et al.)840,000$450$378M$11.3M
Base (50% OON, Peterson-KFF)1,500,000$450$675M$20.3M
High (54.8% OON, JAMA 2024)1,644,000$450$740M$22.2M

Note: the scenarios mix definitions: Adler's 28 percent is a potential-surprise-bill rate, the base 50 percent is an out-of-network share, and the high 54.8 percent is a share of services billed out of network. They are not the same measure, which is why the table shows a range rather than a point estimate. The base-case total is $37.23 million, stated as about $37 million.

Provider-side software is the second revenue layer. NEMSIS's 2025 public dataset covers 63.6 million EMS activations reported by 14,801 agencies. Most of them are small operations, fire departments, volunteer squads, and county services, with no billing department of their own. Most activations do not end in a transport, which is why that activation count sits next to roughly 3 million privately insured transports a year. If 5,000 small and mid-size agencies pay $250 a month for a transparency portal that publishes their rates, benchmarks them against the Medicare fee schedule, and automates claim submission, that is $15 million a year in recurring revenue. Five thousand agencies is about a third of the 14,801 in the dataset. Data is the third layer: insurers, employers, and researchers buying the national price index the platform builds from every bill patients opt in to share. Price that at $2 million a year by year three, an estimate that assumes a dozen insurer and employer subscribers at roughly $165,000 a year each. Total TAM is about $37 million on the base case: $20.3 million patient contingency, $15 million provider software, $2 million data.

SAM narrows to the 27 states without ground ambulance balance-billing protection as of October 2026, where the patient-side product has the most to recover and the provider-side product has the most compliance work to do. Those states hold roughly half the US population, so the serviceable market is about $19 million, using population share as a rough proxy for the state count.

LayerSizeShare of prior layerBasis
TAM (base case)~$37M1.5M exposed rides, 40% recovery, 30% resolution, 25% contingency, plus provider SaaS and data
SAM~$19M~50% of TAM27 unprotected states, roughly half the US population
SOM, year three~$4.4M~23% of SAM60,000 resolved cases (about 200,000 acquired), 400 agencies, data subscriptions

Small. Real. Reachable. SOM in year three: resolve 60,000 patient cases at an average bill of $450, recovering 40 percent, take 25 percent, and earn $2.7 million. Sign 400 provider agencies at $250 a month for $1.2 million. Sell $500,000 of data subscriptions. Total year-three revenue is about $4.4 million. A small business by venture standards. This article says so plainly in the Limitations section. Resolving 60,000 cases means acquiring about 200,000, or 13 percent of the 1.5 million exposed rides, since only 30 percent of acquired cases resolve and pay. Four hundred agencies is 8 percent of the 5,000-agency target. Building it makes sense because the customers arrive pre-qualified, the distribution channel is every employer benefits portal and hospital discharge desk in the country, and the dataset gets more valuable with every bill patients opt in to share.

For scale, the industry behind these numbers is large and consolidating. Grand View Research puts the US ambulance services market at $18.6 billion in 2023, with ground ambulance as the largest segment at about 71 percent of revenue. Medicare alone spent about $3.95 billion on ambulance services in 2021, which sits outside this startup's TAM, since Medicare patients are not the commercially insured population the platform serves. Medicare beneficiaries account for about 40 percent of EMS patients. The surprise-billing problem sits on top of a ground transport market that is consolidating under national private operators while the billing rules have not moved since the No Surprises Act took effect in 2022. Estimates of the market's size disagree: Grand View puts the 2023 US market at $18.6 billion with ground at about 71 percent, while Emergen Research puts the 2025 market at $14.5 billion with ground at about 89 percent. Two years of growth explain part of the gap, not just vendor disagreement. Both imply a ground market of roughly $13 billion, so the $13 billion figure survives the disagreement.

The Product

A two-sided bill defense platform: patient-facing on one side, provider-facing on the other, with a shared pricing dataset in the middle, built as six modules around one dataset, plus an explicit answer to the data conflict the index creates.

  • Bill scanner and benchmark engine: The patient photographs the bill. OCR extracts the HCPCS codes, A0426 through A0429, A0433, and A0434 for ground transport levels, plus mileage and supplies. The engine benchmarks the charge three ways: against the Medicare Ambulance Fee Schedule for that ZIP code and service level, against FAIR Health allowed amounts, and against the local ordinance rate where the county or city sets one. A bill at twice the Medicare rate gets flagged in red with the dollar gap shown. Nobody else publishes the benchmark.
  • State-law shield: A 50-state matrix of ground ambulance balance-billing rules, kept current by a legal researcher, covering the 23 states that had some protection as of October 2026, where the rules differ by plan type because fully insured plans regulated by the state are covered while self-funded employer plans usually are not. The shield tells the patient in plain language whether the provider appears barred from collecting more than in-network cost-sharing for that trip. If it may not, the shield generates the right complaint: to the state insurance department for a dispute with the insurer, and, depending on the state, to the insurance department, the attorney general, the state health department, or the state EMS authority for a dispute with the provider. In many states the insurance department is the enforcer against providers too. Self-funded employer plans generally have no insurance-department remedy, because the Employee Retirement Income Security Act (ERISA), 29 U.S.C. ยง 1144, preempts state insurance regulation of those plans, and the shield says so rather than sending the patient to a regulator with no jurisdiction. The shield also points the patient to the federal No Surprises Help Desk at 1-800-985-3059, which triages all ground-ambulance billing concerns, and, for a self-funded plan, to the Department of Labor's Employee Benefits Security Administration. When state law bars the bill, the state regulator is the enforcer, not the Help Desk. Many patients pay bills that their state's law, for state-regulated plans, would not have required them to pay, because nobody told them the protection existed.
  • Appeal and negotiation autopilot: It files the insurer appeal, negotiates with the provider, and structures a payment plan, on contingency: 25 percent of the amount saved, capped at $150 per case, free if nothing is saved, and free outright for bills the shield flags as barred under state law. While a case is being negotiated, the platform requires a written collections hold from the provider for the duration of the negotiation before it takes a case, and tracks the hold's expiry, so a bill does not go to collections mid-negotiation. The payment plan itself carries no fee; the contingency applies only to the amount saved. Most ambulance bills shrink when challenged, because a provider that settles for part of the balance today avoids the cost and delay of collections, where recovery on medical debt is typically a small fraction of face value. Negotiation here is not adversarial theater. It is arithmetic the provider already accepts in its own bad-debt accounting, though the exact recovery rates vary widely by agency and the platform's own case data would replace these planning assumptions within the first year.
  • Uninsured patients and access: The platform also handles uninsured patients, whose bills have no insurer appeal to file, by benchmarking the charge, screening for charity care under the hospital's 501(r) financial assistance policy where the biller is a hospital, and negotiating directly. It tracks appeal and collections deadlines on every case, and offers the whole flow in Spanish as well as English.
  • Provider transparency portal: Small EMS agencies publish their rates, benchmark them against the Medicare fee schedule, and automate claim submission and state compliance reporting. A fire chief hears a pitch that is not altruism: a published rate card with a Medicare benchmark is the cheapest defense against the next news story about a $2,000 bill, and the GAPB report already recommends that states collect and publish locally set rates. Agencies that publish first set the reference price. Agencies that wait get benchmarked by someone else's dataset.
  • National Ground Ambulance Price Index: Every bill a patient opts in to share, anonymized, feeds a public index of ground ambulance charges by service level, ownership type, and geography. The dataset is also the policy product: the one Congress asked for in 2021 and still does not have. Insurers, employers, and researchers pay for the full feed. Patients and journalists get the headline numbers free.
  • Data conflict, managed openly: The data product carries a conflict the platform has to manage openly. Insurers buying the index are the same insurers the platform negotiates against on the patient's behalf. Its structural answer is threefold: the feed sold to insurers is aggregated only, with no patient-level rows and no sub-county geography; no insurer that is an active counterparty in a pending negotiation buys a feed covering that negotiation; and an independent auditor reviews the data pipeline annually. One tension has no clean answer: aggregated benchmarks are exactly what insurers use to set allowed amounts, and a lower allowed amount means a larger balance bill, so the index can help the other side even while it is aggregated. Employers who buy the feed sponsor the self-funded plans the platform negotiates against, which is the same conflict on the other side of the table. Bills enter the index only when patients opt in to share them, de-identified under the HIPAA Expert Determination method, 45 CFR ยง 164.514(b)(1), with a qualified expert certifying that the risk of re-identification is very small, and with patients able to withdraw consent. The data opt-in is a separate consent from the contingency signup: a patient can use the bill defense without sharing their bill with the index, and the 48-hour cooling-off period covers the data opt-in too, so the patient can use the shield and the benchmark during the wait without their bill entering the dataset. An independent trustee keeps the consent ledger, encrypted at rest, with separation of duties, never sold or shared, and with breach-notification terms in the trustee agreement, so a patient can check whether their bill is in the dataset while the platform itself holds only one-way hashes with no identity mapping. Withdrawal is prospective only: once a bill is de-identified, aggregated, and sold, it cannot be pulled back from a feed already delivered, and the trustee's ledger is the one place a patient's identity touches the dataset, with a retention and destruction policy the auditor reviews. That no-sub-county-geography promise applies to the public feed as well as the insurer feed: no public cell smaller than a county and a service level is published, with cells of fewer than five bills suppressed. The five-bill minimum applies to every subdivision the feed publishes, including ownership type, so a single-operator rural county never appears as its own cell. The platform signs business associate agreements wherever it touches protected health information on an agency's behalf, and it discloses whether insurers could use the feed to set lower allowed amounts. A feed that helps insurers pay less while patients pay a contingency fee would be a second bill, not a defense.

Unit Economics

MetricUnitValue
Median potential bill per case (planning proxy)per case$450
Assumed recovery fractionshare40%
Average patient recovery per resolved caseper resolved case$180
Contingency take rateshare25%
Revenue per resolved caseper resolved case$45.00
Case resolution rate (paying)share30%
Free carve-out cases (appear barred under state law)share2%
Revenue per acquired caseper acquired case$13.50
Cost to serve per caseper acquired case$9.00
Patient-side CACper acquired case$35.00
Patient-side contribution per acquired caseper acquired case-$30.50
Expected-revenue:CAC (patient side)ratio0.39:1
Patient CAC payback—Never (negative contribution)
Provider SaaS priceper agency-month$250
Provider SaaS gross marginpercent85%
Provider CACper agency$1,800
Provider LTV (3 years)per agency$7,650
LTV:CAC ratio (provider)ratio4.25:1
Startup costcompany$1.18M
Break-evenmonthMonth 24 (estimate)

Methodology note: the patient-side numbers are deliberately ugly. With a $35 acquisition cost, $9 to serve, and $13.50 in expected revenue per acquired case, the patient product loses $30.50 per case on its own. It is the honest math, and it is why the patient product is a dataset acquisition channel, not the business. Every case a patient opts in to share adds a priced bill to the index and a provider relationship to the funnel. Start with the $450 figure: it is the median potential bill from the Health Affairs study, used as a planning proxy for the amount at stake, not the amount recovered. The 40 percent recovery fraction is a planning assumption, and it is the number the whole patient side turns on. Raise recovery to 60 percent and revenue per acquired case climbs to $20.25. The contribution improves to negative $23.75, still a loss. Cut recovery to 20 percent and revenue falls to $6.75, the contribution to negative $37.25. A 20 percent resolution rate is worse: $9.00 in revenue, negative $35.00 in contribution. A $50 acquisition cost pushes the contribution to negative $45.50. A $15 service cost pushes it to negative $36.50. Resolution here means paying resolution: 30 percent of acquired cases produce a negotiated reduction or a successful appeal that pays the contingency. About 2 percent more are free disputes on bills that appear barred under state law, so 32 percent of acquired cases resolve in the patient's favor; the rest produce a dataset entry, if the patient opted in, and a referral, which is worth something, but not revenue. On the take side, 25 percent is capped at $150 per case, which does not bind at the base-case $180 recovery but protects patients on large bills. Free carve-out cases, bills that appear barred under state law, are assumed to cover about 2 percent of acquired cases, on top of the 30 percent paying resolution rate. Those cases cost the same $9 to serve and produce no revenue, and that cost is already inside the $9.00 average service cost, which is a blended average across all acquired cases.

Margin lives in the provider SaaS: $250 a month at 85 percent gross margin, $7,650 lifetime value over three years against $1,800 in acquisition cost. Break-even at month 24 is an estimate, not a derivation: it assumes the provider side reaches 200 agencies by month 24 and carries the patient side. If provider adoption runs six months behind plan, break-even slips to about month 30 and the company needs roughly $250,000 more capital: six months of $35,000 monthly burn is $210,000, plus a $40,000 hiring buffer.

Startup Costs

CategoryCostNotes
Benchmark engine + bill OCR$150KHCPCS parsing, Medicare fee schedule integration by ZIP, FAIR Health data licensing. 5 months.
State-law matrix + compliance research$90K50-state balance-billing rules by plan type, updated quarterly. One legal researcher, part time.
Appeal + negotiation workflow$120KLetter generation, insurer appeal tracking, provider negotiation queue, payment-plan tooling. 4 months.
Provider transparency portal$110KRate publication, Medicare benchmarking, claim submission, state reporting. 4 months.
Price index + public dataset$60KAnonymization pipeline, index computation, public dashboard, API for subscribers.
Pilot program (3 agencies, 2 employer groups)$50KRun 500 patient cases at cost in exchange for case-study rights and references.
Sales + industry presence, year 1$80KEmployer benefits conferences, hospital discharge partnerships, state EMS association events.
Legal and regulatory$100KHIPAA compliance, contingency-fee rules by state, insurance appeal regulations, data licensing, unauthorized-practice-of-law and debt-adjusting licensing review, attorney supervision.
Build subtotal$760K
Operating runway (12 months at $35,000/month)$420KFive-person team, cloud hosting, support. Covers the build period and the pilot phase.
Total startup cost$1.18M

Go-to-Market

Phase 1 (months 1-6): Build the benchmark engine and publish the price index free. A public Ground Ambulance Price Index, with median charges by service level and state, is the cheapest possible inbound engine, because every journalist writing about ambulance bills, every employer benefits manager, and every state legislator working on the issue needs that dataset. Sign two employer groups as design partners, offering free bill defense for their employees in exchange for distribution to their benefits portal.

Phase 2 (months 7-12): Run 500 patient cases end to end through the pilot employers, where each case has to prove two numbers rather than one. First, that the platform recovers money often enough to beat its $35 acquisition cost plus $9 service cost. Second, that the benchmark survives contact with a real provider, meaning a bill flagged as twice the Medicare rate actually negotiates down. If either number fails, the patient product is a content site, not a business, and the provider SaaS has to carry the company alone.

Phase 3 (months 13-24): Open the provider portal in the 27 unprotected states, where agencies have the most to gain from publishing a defensible rate card, because a published rate with a Medicare benchmark is the easiest way to show a county board that its prices are reasonable. Add hospital discharge-desk distribution. The bill defense offer reaches the patient in the same week as the bill, when the patient is most likely to act. A cooling-off period applies to every channel, not just discharge desks: no contingency agreement is signed until 48 hours after the patient first sees it, and the patient can cancel within seven days at no cost. Hospital discharge desks are often owned by or affiliated with the same system that sends the bill, so the platform pays nothing for referrals, since a payment for a patient referral can raise exposure under the federal Anti-Kickback Statute, 42 U.S.C. ยง 1320a-7b(b), and any hospital relationship gets counsel review before launch. Publish a quarterly Ambulance Billing Index showing out-of-network rates and median patient burden by state, alongside the share of bills that exceeded what state law allowed the provider to collect. A state with a balance-billing ban is not ranked against one without one, and patients can see where the gap actually is.

Competitive Landscape

CompanyWhat It DoesGap
Zoll RescueNetWidely used EMS billing and records software for agenciesBuilt to help agencies collect. No patient-facing benchmark, no dispute tool, no public pricing
ESOEMS electronic health records and data platform, widely deployed in fire and EMSClinical and operational data, not billing defense. No price index, no patient product
Traumasoft / AIMEMS dispatch, ePCR, and billing suites for mid-size agenciesSame gap as Zoll: the biller's tool, not the patient's. No state-law matrix
Global Medical Response / AMRLargest private ambulance operator, about 1,400 US counties, public since May 2026 at a $3.35 billion valuationIs the biller. Has no incentive to benchmark its own prices or help patients dispute them
Goodbill / Dollar For / ResolveGeneralist medical-bill negotiation and charity-care screeningNo ambulance-specific benchmarks. No Medicare fee schedule comparison, no ordinance rates, no state-law matrix for ground transport
StateIQEMS intelligence platform: agency directory, CMS fee schedules, facility dataIntelligence for agencies and government. No patient dispute resolution, no contingency product
Health insurersPay the allowed amount, run appeals for their own membersStop at the allowed amount. The balance bill is the patient's problem, and the insurer has no incentive to fight it
This startupPatient-side benchmark, appeal, and negotiation plus provider transparency and a public price indexMust earn trust from patients who are in crisis and one-time users, and from providers who profit from opacity

Every existing player sits on one side of the bill. Zoll, ESO, Traumasoft, and AIM sell to the agency that sends the bill, while Goodbill and Dollar For sit on the patient's side but work without ambulance-specific data, and GMR simply is the bill. Nobody owns the benchmark: the number that says what this ride should have cost, in this ZIP code, at this service level, under this state's law. Whoever publishes it first becomes the reference point for negotiations, news stories, and statehouse hearings.

Why Now

The federal fix is documented and stalled. Congress created an advisory committee to solve this problem, and it sent its recommendations in August 2024: ban balance billing, cap copays, make rates public. Two years later, no federal law has acted on them. A documented, unclosed gap with a published federal framework is the best possible timing for a private solution, because the framework tells the startup exactly what the eventual law will look like while the delay leaves the window open to build inside it.

Meanwhile, the largest private operator just went public. GMR listed on the NYSE in May 2026 as GMRS, so its revenue, debt, and margins are now in SEC filings. That has never been true before in this industry.

The largest ambulance company in America is now public. Its prices are still private.

Share this line on X

Raw material for a price index is public for the first time. The Medicare Ground Ambulance Data Collection System now collects cost, revenue, and volume data from providers nationwide, and the same NEMSIS dataset behind the provider-market numbers above is public, so the work is assembling the index rather than discovering the data.

The state patchwork makes the product: 23 states have some protection and 27 do not, and the rules differ by plan type within each state, so compliance complexity that varies by geography is exactly what software is for. Every new state law is a feature, not a threat: it adds a rule to the matrix and a reason for an agency to buy the portal.

Demand is already proven. The No Surprises Act's independent dispute resolution system, which arbitrates provider-initiated payment disputes, received more than 1.46 million disputes in 2024 and about 2.6 million in 2025, far more than the system was built to handle, and providers filed nearly all of them. Ground ambulances cannot use that system at all. The law excludes them, and the advisory committee's proposal to extend a dispute process to ground ambulances was voted down. Ground ambulance patients have no mechanism, and that absence is the opening.

Original Contribution: The $500 Million Transfer

A calculation nobody has published in this form. Start with the exposure. Peterson-KFF puts the count at about 3 million privately insured patients a year, with about half out of network. That works out to 1.5 million exposed rides. A JAMA Network Open study puts the mean total financial burden per out-of-network ground ambulance service at $434.70, against $132.21 in network. The gap, $302.49, is the average excess burden an out-of-network ride imposes on the patient. That $434.70 splits into two parts. That split is the whole point: $160.00 in observed patient cost-sharing and $274.70 in potential balance billing.

Multiply it out. At 1.5 million exposed rides a year times $302.49, the excess patient burden is about $454 million a year. Now apply the advisory committee's own package: a ban on balance billing plus a cap on patient cost-sharing at $100 per ride. The two parts of the package move different money, so they are tabulated separately rather than averaged together.

ComponentPer-ride figureAnnual amount (1.5M rides)What moves it
Observed patient cost-sharing (Gong)$160.00$240MCap: ($160 - $100) x 1.5M = $90M
Potential balance bill (Gong)$274.70$412MBan: $274.70 x 1.5M = $412M
Total OON burden (Gong)$434.70$652MCap + ban combined: $502M
Median potential surprise bill (Chhabra)$450$675MBan only: $675M; a cost-sharing cap cannot touch a balance bill

The honest range is $502 million to $675 million a year. The split matters more than the range. On the JAMA mean, the committee's package moves about $502 million a year, but only about $90 million of that is the copay cap: the cap cuts observed patient cost-sharing from $160.00 to $100, $60 a ride. The other $412 million is the balance-billing ban, which wipes out the $274.70 balance bill outright. On the Chhabra median, the balance bill alone is $450 a ride, so the ban by itself moves about $675 million a year, and the cap adds nothing to that figure because a cost-sharing cap cannot reduce a balance bill. Either way, the package the committee recommended is a half-billion-dollar annual transfer, and it has sat in a report to Congress for two years. One more precision: the committee's actual recommendation is the lesser of $100, adjusted by the consumer price index, or 10 percent of the payment rate set out in Recommendation 12, not a flat $100. That rate is a hierarchy: the amount in a state balance-billing law where one exists, then a state or locally set rate, then a negotiated rate, then a multiple of Medicare that Congress never set. Where that rate is under $1,000, 10 percent of it is less than $100, so the cap bites lower and the cap-only transfer runs larger than $90 million; the $90 million figure is the floor for the cap, not the ceiling. The committee never published a rate distribution, so the lesser-of figure cannot be computed from public data, and this article does not try. The $454 million excess burden and the $502 million combined transfer differ by $48.3 million for a simple reason: the package does not just remove the excess over in-network care, it also cuts the in-network-level cost-sharing from $132.21 to the $100 cap, $32.21 a ride, which the excess-burden figure does not count. One scope note: the Gong figures pool emergency and non-emergency rides, while the emergency-only row of the same table implies about $519 million for the combined package, so the $502 million figure is conservative.

Half a billion dollars a year. That is what the committee's ban-and-cap package would move from patients to insurers and providers.

Share this line on X

Make it concrete per ride. A patient in a state without protection, transported by a private-sector ambulance, faces a mean burden of $434.70, the all-rides average, not a private-equity-specific figure. In network, that ride costs $132.21. The $302 difference is not medical care. It is the price of having no choice about which ambulance arrives, and no benchmark to challenge the bill afterward. Under the platform's contingency model, it needs to resolve about 200,000 cases a year to reach $9 million in patient-side revenue. That $9 million is the revenue at 200,000 resolved cases, about three times the year-three patient-side target, not a breakeven point. At the 30 percent resolution rate in the unit economics table, resolving 200,000 cases means acquiring about 667,000, or 44 percent of the exposed rides, and that is before the patient product's negative contribution is counted. The unit economics demand that scale, and the sensitivity in the Unit Economics section shows what happens if resolution rates or recovery amounts run lower.

Limitations

This analysis has several weaknesses that should be stated plainly.

First, the out-of-network rate is genuinely uncertain. The studies cited here give 28 percent, 54.8 percent, 59.4 percent, 62 percent, and 71 percent, and Peterson-KFF puts the out-of-network share at about 50 percent. They are all measuring different things: emergency transports only, all ground services, claim lines, one insurer's book, a national claims sample. The TAM table shows the low, base, and high cases for that reason. If the true emergency-transport rate is 28 percent rather than 50 percent, patient-side revenue falls by 44 percent and the year-three target is unreachable without provider SaaS carrying more weight.

Second, that $450 median comes from one national insurer's data from 2013 to 2017, prices have risen since then, and that insurer's network mix may not represent the country as a whole. JAMA's mean of $434.70 comes from employer-based plans from 2015 to 2020, also dated. There is no current, national, all-payer dataset of ground ambulance patient burden. This startup's price index is itself the fix for that gap, which means the TAM is built on the numbers the product is meant to replace.

Third, patient-side unit economics are negative on a standalone basis, as the methodology note says plainly. The business only works if provider SaaS adoption arrives on schedule and if the dataset becomes valuable enough to sell. Both conditions are unproven, so a contingency model that loses $30.50 per acquired case needs its second revenue stream to arrive before the first one burns through the runway.

Fourth, the regulatory risk runs both ways. A federal balance-billing ban would shrink the patient-side market overnight, since the bills the platform disputes would become unlawful to collect. Provider compliance would survive and grow under a ban, but the patient contingency revenue would fall sharply. In part, the startup is a bet that Congress stays slow. That is a reasonable bet on the record so far, but it is a bet on legislative inaction, and inaction is the one thing that can change without warning.

Fifth, large billers have no reason to cooperate. GMR and other national operators can refuse to negotiate with the platform, refuse to publish rates, and route disputes through their own legal teams. Its standing is the patient's legal right to appeal and the publicity of the index, not any contractual relationship with the provider. If the big operators stonewall, early resolution rates will run below the 30 percent planning assumption.

Sixth, a 25 percent contingency fee deserves scrutiny: it takes a quarter of a distressed patient's recovery, and a platform paid only when it wins will cherry-pick the easiest cases, leaving the hardest bills, often the largest ones, undefended. The model answers that with a $150 cap per case and a free dispute for bills the shield flags as barred under state law, but the cherry-picking incentive remains, and the cap does not fix it. A take-all commitment covers the pilot employers: the platform takes every case its design partners refer, easy or hard, and publishes its resolution rate by bill size so cherry-picking would show up in its own numbers. A non-attorney platform that files appeals and negotiates bills on a patient's behalf also has to answer for unauthorized-practice-of-law and insurance-adjuster licensing rules, which vary by state, and for the authorized-representative forms insurers require before they will talk to anyone but the patient. The mitigation is to operate under attorney supervision, partner with a licensed adjuster where a state requires one, and map the debt-adjusting and credit-services-organization statutes state by state before launch, with the legal budget line in the Startup Costs section covering that review.

Seventh, the price index carries a coordination risk the product section does not discuss: a public benchmark could let providers converge on the highest published price rather than compete it down, with a dominant operator treating the index as a price signal. The public feed should publish medians and ranges by geography, not agency-level prices, with the agency-specific benchmark shown only to the patient holding that bill.

Finally, patients are the hardest customers in health care to reach. They are sick, stressed, and dealing with a bill weeks after the ride, usually alone, with no prior relationship to any brand. Acquisition depends on distribution through employers, hospitals, and search, and none of those channels is cheap or fast. The $35 CAC assumes search and benefits-portal distribution work. If they do not, CAC rises toward $80, the patient product loses $75.50 per case, and the runway math breaks.

Strongest Counterargument

Here is the case against this startup: the market exists only because of a regulatory failure, and regulatory failures get fixed, or they get worse in ways that help nobody.

Start with the fix. The committee recommended a balance-billing ban and a copay cap, and the only thing missing is a bill that makes it through committee and a floor vote. If that vote happens, the patient-side business loses its reason to exist, and what remains is a compliance SaaS and a dataset that Zoll or ESO could build as a feature in a quarter. A startup whose core product is a bet that Congress stays broken has a built-in expiration date.

Then consider the alternative: Congress does nothing, and the biggest billers simply refuse to play. GMR does not need to negotiate with a startup: it has a legal department, a collections operation, and national scale, and it can decline every appeal. A 30 percent resolution rate assumes providers behave like rational collectors who prefer settling now to sending a bill to collections. Large operators with market power do not face that tradeoff the same way, because their patients cannot choose another ambulance.

Its answer is that the provider SaaS and the price index survive either outcome. Small agencies adopt first and create the benchmark the big operators eventually have to answer, and state laws keep arriving, 23 states and counting, which keeps the compliance product growing even if the federal ban lands. Really, the company is three businesses sharing one dataset, and the bet is that at least one of them reaches scale before the runway runs out. Plausible but unproven. It depends on small agencies adopting fast enough to fund the patient product while it loses money on every case. If either assumption fails, this is a content site with a price index and a burn rate, not a platform.

The Bottom Line

Congress banned surprise medical bills for almost everything except the ride to the hospital. About 3 million privately insured patients a year are transported by ground ambulance, about half of them out of network, and the average excess burden is about $302 per ride. A federal committee studied the problem and recommended a ban and a copay cap in August 2024. No federal law has acted on it. That is the whole story in one sentence: the gap is documented, the fix is stalled, and the business is the bridge between the two. Those bills are not a mystery. They are a dataset problem, a legal-information problem, and a negotiation problem, and all three are solvable by software, in a market of about $37 million on the base case that grows as state laws multiply and as the largest operator's finances become public through its SEC filings. Too small for a unicorn. Exactly the right size for a focused vertical business that charges for the benchmark, the appeal, and the negotiation instead of for the ride itself.

What You Can Do

This is general information, not legal advice. If you got an ambulance bill:

  1. Do not pay the balance-billed amount before checking whether your state bans balance billing for ground ambulances and whether your plan is state-regulated or self-funded, because the answer changes what you owe.
  2. Ask the provider for an itemized bill with the HCPCS codes, then look up the Medicare Ambulance Fee Schedule rate for your ZIP code and service level.
  3. If the charge is more than double the Medicare rate, appeal to your insurer and ask the provider for a reduction in writing, citing the benchmark.
  4. Appeal promptly: most employer plans give you at least 180 days to appeal a denial under federal rules (29 CFR ยง 2560.503-1(h)(3)(i)), and state deadlines can be shorter.
  5. If the bill is already in collections, dispute it in writing and ask for the original insurer explanation of benefits. Do not assume ignoring it is free, since an unpaid balance can still reach collections even while you dispute the amount.
  6. If the bill looks barred under state law, call the federal No Surprises Help Desk at 1-800-985-3059, which triages ground-ambulance billing concerns, and file with your state insurance department or regulator, which enforces the ban.

If you run an EMS agency: publish your rates and benchmark them against the Medicare schedule before someone else benchmarks them for you. If you build this: start with the index, not the negotiation tool. Build the index first. A free price index that every journalist and legislator bookmarks is distribution no competitor has, and a negotiation tool without distribution is a form letter with a login page.

Related

๐Ÿ“ฐ Pharmacy PBM Reimbursement Audit: the same audit-as-product model, applied to drug pricing instead of ambulance bills

๐Ÿ“ฐ Roof Replacement Price Transparency: the structural twin: an opaque price, a benchmark nobody publishes, and a transparency platform as the business

๐Ÿ“ฐ Hospice SFP Risk Intelligence SaaS: the same regulatory-complexity play, applied to a different corner of health care