The $196 Million Question
In 2022, federal auditors reviewed a sample of New York's non-emergency medical transportation (NEMT) claims and found that 72% failed to meet documentation requirements, projecting roughly $196 million in improper payments across the program. Not all of that was fraud. Much of it was something more mundane and more preventable: trip logs with a missing signature, a mileage figure that didn't match the GPS record, or a completed-trip code billed for a ride that never actually happened.
If you run an NEMT company — even a two-van operation shuttling dialysis patients three days a week — this is the risk that sits underneath your business model. You're not just driving people to appointments. You're generating a claim for every trip, and that claim has to survive an audit that can happen months or years after the ride ended. The business that wins in this industry isn't necessarily the one with the newest vans. It's the one whose books can answer "prove it" for any trip on any given day.
Why NEMT Bookkeeping Is Different From Other Transportation Businesses
A rideshare driver or a courier service invoices a customer and moves on. An NEMT provider is paid — almost always — by a third party that wasn't in the vehicle: a state Medicaid agency, or more commonly, a transportation broker the state has contracted to manage the benefit. That third-party payment structure changes everything about how you need to track money.
You're not billing for a ride. You're billing for a documented, verifiable event. Every trip requires supporting evidence: patient name and Medicaid ID, date of service, actual pickup and drop-off times, origin and destination addresses, driver name and signature, a passenger or witness signature, odometer readings at both ends of the trip, the vehicle's license plate or fleet ID, an authorization or trip ID number, and the correct HCPCS billing code. Most states also require GPS records that confirm the billed mileage follows the most direct route — not a longer route that pads reimbursement.
Your books, in other words, have to reconcile two parallel ledgers: the financial one (what you were paid, when, by which broker) and the operational one (what actually happened on the road, trip by trip). When those two don't match, that's not just an accounting discrepancy — it's the exact pattern an auditor is trained to flag.
Understanding the Broker Reimbursement Model
Most states no longer pay NEMT providers directly. Instead, they contract with a transportation broker — a private company that manages the entire NEMT benefit for the state. The broker verifies rider eligibility, assigns trips to providers in its network, sets negotiated rates, and processes payment. Your business relationship, day to day, is with the broker, not with the state Medicaid office.
This structure matters for your bookkeeping in a few concrete ways:
- Reimbursement timing is broker-controlled, not statutory. Payment cycles vary by broker — some pay weekly, some biweekly, some monthly, and some hold back a percentage pending trip verification. You need a receivables aging schedule that tracks amounts owed by broker, not just in aggregate, because a payment delay from one broker shouldn't be confused with a shortfall from another.
- Rates are set per trip type, not negotiated per ride. Ambulatory trips typically reimburse in the $20–$50 range, wheelchair transport $40–$100, and stretcher service $150–$300, often plus a per-loaded-mile add-on of $2–$5. Your chart of accounts should separate revenue by service tier so you can see which trip types are actually profitable — wheelchair van margins commonly run 15–20% net, ambulatory closer to 10–15%, and stretcher service 20–30%, because the higher rate outpaces the incremental equipment and staffing cost.
- Denials and clawbacks need their own tracking category. A broker that finds a documentation gap won't just deny the claim — it may recoup a payment already made, sometimes weeks after the fact. If your books record revenue at time of trip completion without a mechanism to true up for later denials, your P&L will look better than your bank account.
The Real Cost Structure Behind Each Trip
Because reimbursement rates are fixed by the broker, your profitability is determined almost entirely by cost control — and NEMT costs are more granular than most small business owners expect. A realistic per-trip cost breakdown for a wheelchair van looks something like:
- Driver labor: $9–$13.50
- Fuel: $2.70–$3.60
- Vehicle depreciation: $1.50–$2.70
- Insurance: $1.20–$2.25
- Dispatch and scheduling software: $1.50–$2.50
- Administrative overhead: $2.00–$3.50
That's roughly $19–$30 in cost per trip before you know what the reimbursement will be. Driver wages and payroll taxes alone typically run $35,000–$45,000 per year per vehicle, and with industry turnover commonly cited between 50% and 100% annually, competitive pay (often $18–$25/hour) isn't optional — it's the only lever that keeps a van staffed. Commercial auto insurance for NEMT vehicles adds another $7,000–$16,000 per vehicle per year, higher for new operators without a claims history.
The upshot: an ambulatory van generally needs 6–8 trips a day to clear its costs, a wheelchair van 4–6, and a stretcher vehicle 2–4. If you're not tracking cost per trip by vehicle and by service tier, you can't tell whether a slow day is a scheduling problem or whether that particular vehicle type simply isn't pulling its weight. This is exactly the kind of job-costing insight that generic invoicing software won't surface — it requires treating each vehicle like its own mini profit center in your books.
The Compliance Risk Nobody Budgets For
Here's the part that makes NEMT bookkeeping fundamentally different from most small business accounting: a documentation error isn't just a lost deduction. It's a False Claims Act exposure.
Billing a completed-trip code — and the associated loaded mileage — for a ride where the patient was a no-show is a textbook example of what auditors are trained to catch, and it's explicitly called out by CMS's Medicaid Integrity program as a common NEMT fraud scheme. It doesn't have to be intentional to be a problem. If your trip logs are sloppy enough that a no-show and a completed trip look the same on paper, you're carrying the same audit risk as if you'd falsified the record on purpose. Penalties for a willful pattern of false claims start at roughly $14,000 per claim, plus treble damages — numbers that turn a single bad month of recordkeeping into an existential threat for a small operator.
Real cases bear this out: transportation and ambulance providers have paid multi-million-dollar settlements — $12.7 million and $9 million in two well-documented examples — for claims that didn't match what actually happened on the road. You don't need to be a large regional operator to be in that crosshairs; brokers and state Medicaid Fraud Control Units run pattern analysis across providers of every size, and a small business with concentrated, repeatable errors can actually stand out more, not less.
The practical takeaway: treat every trip record as a legal document, not a scheduling note. That means:
- Reconcile trip logs to invoices before you bill, not after. Every claim submitted to a broker should trace back to a specific trip record with all 14 required fields intact — pickup/drop-off times, both signatures, both odometer readings, and a GPS trace that matches the billed mileage.
- Flag no-shows the moment they happen, and route them to a separate accounting treatment (a no-show fee, if your broker contract allows one) rather than letting them get miscoded as completed trips downstream.
- Keep the financial and operational record in the same system, or at minimum reconciled on a fixed schedule. A trip-log gap discovered during an audit two years later is much harder to explain than one caught during this week's invoice run.
Setting Up a Chart of Accounts That Actually Tells You Something
A generic "transportation revenue" line item hides more than it reveals in an NEMT business. Because reimbursement rates, cost structures, and margins all vary by trip type, your chart of accounts should be built around the same dimensions a broker uses to pay you:
- Revenue split by service tier — ambulatory, wheelchair, and stretcher trips as separate income accounts, not one combined "transportation income" bucket. This is the only way to see that your stretcher vehicle, despite fewer daily trips, may be your most profitable asset.
- Revenue split by broker — if you contract with more than one broker (common in states with regional broker territories), track receivables and revenue separately per broker. A payment delay from one broker is a cash-flow issue; a shortfall from a second broker on the same week can look identical on a combined ledger and mask which relationship actually needs attention.
- A dedicated clawback/denial account — rather than netting denied or recouped claims against gross revenue, record them explicitly. This preserves a clean audit trail of what was billed, what was paid, and what was later reversed — the exact record a compliance review will ask for first.
- Per-vehicle cost centers — fuel, maintenance, insurance, and depreciation tagged to a specific vehicle ID. Since profitability differences between trip types are driven almost entirely by cost, not price, this is what lets you decide whether to retire an aging van or add a second stretcher vehicle instead of a third ambulatory one.
Multi-vehicle operators in particular benefit from being able to filter and query this level of detail quickly — pulling a per-vehicle, per-broker profitability report shouldn't require exporting three spreadsheets and reconciling them by hand. Beancount.io's documentation walks through setting up exactly this kind of tagged, multi-dimensional chart of accounts using plain-text entries you can query like a database.
Why Plain-Text, Version-Controlled Books Fit This Industry Well
NEMT bookkeeping is really a reconciliation problem: broker payments on one side, trip-level documentation on the other, split across multiple revenue tiers and multiple counterparties, all of it subject to audit long after the money has changed hands. That's a case where being able to see exactly what changed, when, and why — down to the individual ledger entry — isn't a nice-to-have, it's the thing that lets you answer an auditor's question in minutes instead of days.
Beancount.io provides plain-text accounting that gives you that kind of transparency and control over your financial data — every entry is human-readable, every change is tracked, and there's no black box between "what we billed" and "what we can prove." Get started for free and see why finance-conscious business owners are moving to plain-text accounting for exactly the kind of audit-ready recordkeeping that industries like NEMT demand.