You can run a trial flawlessly — enroll on time, enter clean data, pass every monitoring visit — and still lose money on it. Not because the science failed, but because the budget you signed never covered your real costs, and the payments you earned are sitting in a sponsor's queue 90 days past due. For an independent research site, the difference between a profitable study and an expensive lesson is rarely the medicine. It is the bookkeeping.
Clinical trial sites get paid through a system unlike almost any other business: fixed per-unit fees set before the work starts, startup payments that may or may not cover activation, separate pass-through costs you must invoice line by line, and a holdback the sponsor keeps until closeout. This guide walks through how each piece works, where sites typically leak money, and the tracking habits that keep every dollar you earned actually arriving.
How Site Payments Actually Work
Before a study starts, the sponsor sends a payment schedule as part of the startup package. It becomes an exhibit (or a detailed section) of the Clinical Trial Agreement (CTA) — the contract that governs the whole financial relationship. Most CTAs pay on negotiated fixed per-unit fees, not cost reimbursement. That single fact shapes everything: you are agreeing to a price per visit, per procedure, and per milestone, and if your costs run higher, the sponsor does not automatically make up the difference.
Study teams use the final payment schedule to set up billing, track what has been earned, and reconcile what has been paid. Treat it as your price list and your collection blueprint, not as paperwork to file away. Every per-visit rate, screen-failure rate, startup fee, pass-through line, payment term, and holdback percentage in that document will matter to your books for the next two to five years.
Startup Fees: Price Activation Like It Costs Real Money
Opening a study costs money long before the first patient walks in: regulatory document preparation, ethics review submissions, contract and budget negotiation, pharmacy setup, staff training, and building the protocol into your systems. Sponsors typically pay a one-time startup fee to cover this — often in the range of a few thousand dollars per site, higher for early- or late-phase drug trials and lower for observational or device studies. Industry surveys consistently find that startup economics vary widely by trial type and site setting, so a flat "standard" startup offer deserves scrutiny, not automatic acceptance.
Three rules for startup fees:
- Make them nonrefundable. If the study never enrolls a single patient at your site — a competitive-enrollment trial that fills elsewhere, a sponsor that pauses the program — the activation work still happened. A startup fee that must be returned when enrollment is zero means you worked for free.
- Itemize what it covers. Regulatory submissions, contract review, coverage analysis, system build, initiation-visit preparation. When the list is explicit, mid-study "can you also handle..." requests become amendments with their own fees instead of free extras.
- Separate institutional overhead. Academic and health-system sites add indirect-cost rates (commonly around 27–35% of direct costs, varying by institution) plus administrative or management fees. Know your institution's required rate before you negotiate, because it comes off your side of the budget, not the sponsor's.
Per-Patient Visits: Where Enrollment Math Decides Profitability
The core of site revenue is per-patient, per-visit payments: a fixed amount for each completed visit and procedure, paid as patients move through the protocol schedule. Your profitability therefore depends on two numbers — the per-visit rate and how many patients complete how many visits.
Build the budget visit by visit from the protocol's schedule of events, costing staff time, procedures, labs, pharmacy handling, and data entry for each one. Then stress-test it: what does this study look like at 50% of target enrollment? At 70%, with two early dropouts? Because payment arrives per completed unit of work, under-enrollment does not just reduce revenue — it strands the fixed startup investment across fewer paying visits. Sites routinely find that adding one more completed patient is what flips a study from loss to profit, which is exactly why realistic enrollment forecasting belongs in the financial review, not just the feasibility questionnaire.
Screen Failures: Stop Doing Unpaid Work
Screening a patient who turns out to be ineligible still consumes staff time, labs, ECGs, imaging, and data entry. If your CTA pays nothing for screen failures, every failed screen is a direct loss — and screen-failure rates of 20–30% or higher are common in many therapeutic areas.
Negotiate a screen-failure rate up front, typically prorated from the budgeted visit costs: payment for the screening visit plus any protocol-required procedures actually performed, up to an agreed cap per study or per site. Get three details in writing: the rate, the cap (e.g., paid screen failures up to a set number or ratio to randomized patients), and what documentation triggers payment. Then track screens attempted versus screens paid every month — unclaimed screen-failure payments are one of the most common leaks in site bookkeeping, because the work happened months before anyone checks the invoice.
Pass-Through Costs: Invoice Them Separately, Every Time
Pass-through costs are real expenses you incur on the sponsor's behalf and bill back at cost (sometimes with an agreed administrative markup). Standard pass-throughs include ethics review fees, advertising and recruitment costs, serious-adverse-event reporting support, document storage and archiving, and the staff time spent on contract and protocol amendments.
The failure mode is burying these in the per-patient rate or forgetting to invoice them at all. Instead:
- List every pass-through category in the budget, even ones you expect to be zero. A listed-but-zero line can be invoiced later; an unlisted line becomes a negotiation.
- Require invoice-based payment terms, not "paid with the next patient payment." Pass-throughs should be payable on invoice within the CTA's payment terms (push for 30 days, not 60 or 90).
- Log amendment work immediately. Each protocol amendment means re-consent, retraining, systems updates, and resubmissions. If amendment fees are a pass-through, open the invoice the week the amendment arrives — not at closeout, when leverage is gone.
Overhead and Institutional Fees: The Cut That Comes Off Your Side
If your site sits inside a university, hospital, or health system, part of every sponsor dollar goes to the institution before it reaches your study account. Industry clinical-trial indirect rates vary by institution — budget documents in the 27–35% range are typical — and they generally apply to the direct-cost base excluding certain categories like patient-care subcontracts. On top of that, many institutions assess clinical-trial office fees, management fees, or per-study activation charges.
Independent sites are not exempt from this logic either: your "overhead" is rent, insurance, CTMS licenses, regulatory subscriptions, and the coordinator time that no single study fully funds. Either way, build the budget on fully loaded costs. A per-visit rate that covers the coordinator's hourly wage but not the 30% institutional overhead — or the independent site's rent and systems — is a rate that guarantees a loss you will not notice until year-end.
Sponsor Receivables: Getting the Money You Already Earned
Earning revenue and collecting it are different jobs, and sponsors make the second one harder than it should be. Nearly a quarter of research sites report payment delays of 90 days or longer, and holdback provisions — where the sponsor retains a percentage of each payment until database lock or closeout — can keep the final 5–10% outstanding for years. Some agreements even include "pay when paid" clauses limiting the site's right to collect from a CRO until the sponsor pays the CRO, an arrangement over which the site has no visibility or control.
Run receivables like a business that intends to be paid:
- Invoice on earned milestones, not on reminders. Define who creates each invoice, what triggers it (visit entered in EDC, monitoring visit completed, milestone certified), and how many days after the trigger it goes out. Sites without a defined invoicing process discover earned-but-unbilled revenue months late.
- Track payment terms per study. Thirty-day terms in one CTA and 60-day terms in another mean a single "days outstanding" report misleads you. Age receivables against each agreement's own terms.
- Reconcile payments to visits monthly. Match every sponsor payment to specific patients, visits, and invoices in your CTMS. Short-pays, missed screen failures, and unpaid pass-throughs surface only when earned-versus-paid is reconciled line by line.
- Chase early and escalate in writing. A 45-day follow-up email, a 60-day formal notice referencing the CTA's payment clause, and a principal-investigator-to-sponsor escalation at 90 days will collect more than hope. Document every step — the collection log is also your evidence if a dispute reaches the contract level.
- Budget for the holdback. Treat retained amounts as long-dated receivables in your cash forecast, not as missing money. And negotiate the holdback percentage and release trigger (e.g., release on completion of closeout visit, not on final study publication) before signing.
The Billing Trap: Standard of Care vs. Research
One accounting error can dwarf all the others: billing the wrong payer. Trial sponsors pay only for protocol-required activities that are not standard of care; routine care still goes to the patient's insurance. Every study needs a coverage analysis — a visit-by-visit determination of what is billed to the sponsor versus billed to insurance — completed before enrollment starts. Billing insurance for research procedures risks False Claims exposure; billing the sponsor for standard care leaves insurance revenue uncollected and inflates study costs. Keep the coverage analysis with the study file, train coordinators on it at initiation, and re-check it every time a protocol amendment changes the schedule of events.
Common Mistakes That Quietly Erase Margin
- Signing the sponsor's first budget. Initial sponsor budgets systematically undercount coordinator time, screen failures, amendments, and closeout. Negotiate from your own costed budget, not theirs.
- Forgetting closeout costs. Records archiving, final data queries, closeout visits, and long-term document storage arrive after the last patient payment. Price a closeout fee or reserve explicitly.
- Recognizing revenue on enrollment instead of completion. Revenue is earned when the visit is completed and documented — often contingent on monitoring verification — not when the patient signs consent. Books that record income at enrollment overstate revenue and understate the collection problem.
- Letting amendments accumulate unbilled. Three "minor" amendments without fees can equal a coordinator-month of unpaid work. Each amendment reopens the budget conversation.
- Co-mingling studies in one account. When five studies share one bank balance and one spreadsheet, no study's profitability is knowable. Track revenue, receivables, and direct costs per protocol, per sponsor.
The Bookkeeping System That Makes It Work
None of the above works without a per-study ledger discipline: a separate tracking account for each protocol recording contracted rates, visits completed, invoices issued, payments received, pass-throughs billed, and holdback retained. Reconcile earned-versus-paid monthly, review AR aging against each CTA's terms, and forecast cash with realistic enrollment and payment-lag assumptions — including the cost of financing receivables when sponsors pay quarterly in arrears. Missed milestones, unpaid screen failures, and drifting payment terms are all visible in a monthly report; they are all invisible in an annual one.
Accurate per-study books do more than support collections. They give you the cost history that makes the next budget negotiation empirical instead of hopeful: your actual per-visit cost, your actual screen-failure rate, your actual days-to-payment by sponsor. Sites that know their numbers negotiate better rates, decline unprofitable studies earlier, and concentrate staff on the protocols that pay.
Simplify Your Financial Management
Running a research site means tracking dozens of payment streams — startup fees, per-visit revenue, screen failures, pass-throughs, and holdbacks — across sponsors with different terms and timelines. Beancount.io provides plain-text accounting that gives you complete transparency and control over your financial data, so every study's true profitability is knowable instead of guessed. Get started for free and bring version-controlled clarity to your site's books.
