Clinical trial cost breakdown: a $1m Phase 2a example
A clinical trial budget has to pay for study oversight, patient care, vendors, data systems and recruitment. The shares depend on the protocol. A small remote crossover study cannot price a large multicountry trial.
On 4 November 2026, Lindus Therapeutics reported a Phase 2a ME/CFS trial costing just shy of $1 million. Its published cost shares support a useful budget example. Its comparison with seven modeled trials does not establish that AI alone caused a 13-fold cost reduction.
Download the illustrative clinical-trial cost worksheet (CSV). It applies Lindus's reported percentages to a rounded hypothetical $1 million, then changes one category at a time. The dollar amounts are calculations, not Lindus's exact expenditure or supplier quotes.
The reported trial and its cost boundary
Lindus describes ReMEdi as a randomized, double-blind, placebo-controlled Phase 2a trial of an oral investigational drug in adults with moderate myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). It reports:
| Study feature | Company-reported scope |
|---|---|
| Participants | 25 randomized |
| Design | Crossover: two weeks of active drug and two weeks of placebo, separated by a washout |
| Participant schedule | Nine weeks in total |
| Sites | One virtual site, operated by Lindus |
| Remote work | Video assessments, home blood draws and ECGs, drug delivery and electronic outcomes |
| Cost boundary | Protocol drafting through closeout, analysis and clinical study report writeup |
The source says its end-to-end budget includes Lindus and site personnel time, medical oversight and statistical oversight. It attributes the trial to registration ISRCTN74088209. The registration page could not be retrieved during this article's source check; the operational details above remain attributed to the company account. No independent budget audit or participant-level efficacy analysis is supplied here.
The nine-week participant schedule is not a nine-week protocol-to-report timeline. Startup, recruitment and final analysis need their own dates. The clinical study report is part of the stated cost boundary, so a proposal ending at last patient visit would cover less work.
Clinical-trial budget categories and the $1m worksheet
Lindus publishes these shares in its budget table. The descriptions below condense that table. Every dollar amount uses an illustrative $1,000,000 total.
| Budget category | Share reported by Lindus | Illustrative USD | Work included in the company's category |
|---|---|---|---|
| Trial oversight and monitoring | 44% | $440,000 | Project management, medical and clinical monitoring, coding, data management and vendor oversight |
| Pass-through costs | 30% | $300,000 | Devices, home tests, drug logistics, randomization, trial master file, ethics, investigator fees, statistical analysis and participant payments |
| Trial site | 12% | $120,000 | Investigators, nurses and coordinators |
| Citrus software and token spend | 11% | $110,000 | Electronic data capture, clinical database, trial management and electronic patient-reported outcomes |
| Other: central recruitment campaign | 3% | $30,000 | Central patient recruitment campaign |
| Total | 100% | $1,000,000 | Illustrative total, not an exact actual-cost disclosure |
These are the company's accounting categories. Investigator fees appear under pass-through costs, while investigator personnel also appear under site costs. The public table does not reconcile those boundaries. A sponsor building a budget should assign each invoice or labor charge once and ask the supplier to resolve any overlap.
The 3% recruitment line describes the central campaign. It does not prove that all screening labor and medical review cost 3%. Those activities can fall under site work or oversight. Likewise, 11% software and token spend is the reported combined line, not an isolated bill for model tokens.
Reproduce or change the example
The analysis script uses decimal arithmetic and checks that the shares sum to 100%. Run it from the repository root:
python3 scripts/analyze-clinical-trial-cost.py --check
python3 scripts/analyze-clinical-trial-cost.py --total 2000000
The second command prints the same assumed shares at a different illustrative total. It does not estimate the cost of enrolling more patients. The CSV repeats each scenario's total on its category rows; sum the category amounts once, not the repeated total column.
A cost reduction in one category is smaller at trial level
Holding every other category fixed gives the following sensitivity analysis. Each change is an assumption, not a forecast of AI performance or a vendor commitment.
| Illustrative change | Calculation | New total | Change against $1m baseline |
|---|---|---|---|
| Oversight and monitoring cost falls 25% | $440,000 × 25% = $110,000 saved | $890,000 | 11% lower |
| Pass-through costs double | $300,000 added | $1,300,000 | 30% higher |
| Central recruitment campaign cost halves | $30,000 × 50% = $15,000 saved | $985,000 | 1.5% lower |
A 25% reduction in staff hours would not necessarily mean a 25% reduction in the oversight bill. Staffing commitments, review work and contract pricing can prevent the sponsor from receiving the full savings. The example assumes a reduction in the cost itself. It includes no implementation expense or displaced work in another category.
The pass-through scenario also shows why cheap document drafting cannot determine the whole budget. Home procedures, logistics and participant payments remain payable even if a team drafts documents faster. None of these assumptions changes the required scientific design or safety work.
Read the $13.3m comparison as modeled cost
Lindus compares ReMEdi with seven completed, company-sponsored Phase 2 ME/CFS trials starting in 2015 or later. It says it extracted the GlobalData estimates in August 2025. All seven were randomized, double-blind and placebo-controlled.
| Measure | Seven-trial benchmark reported by Lindus | ReMEdi reported by Lindus |
|---|---|---|
| Total trial cost | $13.3m mean, modeled | About $1m, company-reported |
| Participants | 62 mean enrolled | 25 randomized |
| Cost per participant | $289,000 mean, modeled | About $40,000 |
The seven comparator costs are modeled estimates, not audited actual spending. Their aggregate publication does not establish matched site counts, visit schedules, follow-up or endpoints. ReMEdi's crossover also means each participant contributed active-drug and placebo data.
Using rounded totals, $13.3m divided by $1m is 13.3. That is a ratio between a modeled benchmark mean and a company-reported trial cost. A causal estimate of AI savings would need a comparable alternative delivery method and controls for changes in study design and scope.
The benchmark's reported mean cost per participant is a separate statistic. Dividing its $13.3m mean total by its 62 mean participants gives about $215,000, not $289,000. A mean of trial-level ratios can differ from a ratio of means. The underlying seven budgets are not published in the account, so the $289,000 figure remains attributed rather than reconstructed. The rounded $289,000/$40,000 ratio is about 7.2, with the same comparability limits.
The $21m Phase 2 figure is per program
The Lindus post also cites a $21m average Phase 2 cost. Its cited source, HHS/ERG's September 2024 Drug Development final report, Table 11, gives $21.0 million in mean modeled out-of-pocket Phase 2 spending per drug development program, in 2018 dollars across 13 therapeutic areas.
That model assumes 1.52 Phase 2 trials per program on average (Table 5), and an average Phase 2 trial size of 235 participants (Table 4). The $21m figure is not a quote for one trial. It also differs from the report's failure-adjusted and capitalized cost measures. No inflation adjustment is applied here, and it should not be read as a current bid for ReMEdi's protocol.
Build a comparable sponsor budget before choosing a CRO
Keep sponsor expenditure separate from CRO revenue and margin. A sponsor may pay a CRO for services, reimburse pass-through charges, and pay sites or laboratories directly. A CRO's headline contract value may omit those direct payments. Gross margin requires the provider's revenue and cost records; it cannot be inferred by subtracting the published 30% pass-through share from this case.
For a new protocol, replace the worksheet's percentages with quantities and scoped prices:
| Budget input | Evidence to request | Scope change that needs a fresh estimate |
|---|---|---|
| Enrollment and screening | Randomized target, screen-failure assumptions, paid procedures per screen | More screened candidates per randomized participant |
| Site work | Startup and closeout fees, per-visit work, site count | Extra sites, countries or procedures requiring attendance |
| Treatment and follow-up | Schedule of activities, number of contacts, safety procedures | Longer observation or more assessments |
| Data and oversight | Monitoring plan, transfers, query handling, safety review and database lock | More vendors or data sources to reconcile |
| Supply and reporting | Drug manufacture and packaging boundary, logistics, analysis outputs and CSR | Extra supply runs or a broader reporting package |
Ask each bidder to state exclusions, change-order rates, sponsor-retained work and taxes. Keep any contingency separate from the scoped base cost. The public Lindus account does not provide a detailed reconciliation for every possible sponsor expense, such as prior drug development or financing. Its stated protocol-to-CSR boundary is narrower than total asset-development cost.
A remote design can replace some travel and physical-site work while adding home logistics and device support. A crossover can reduce the number of distinct people needed for a particular design, but its suitability depends on the treatment and disease. Neither choice transfers automatically to another study. Compare bids against the same protocol and deliverables, then record which assumptions changed.
Sources and calculation limits
Source check: 16 November 2026. The budget shares and ReMEdi operational claims come from Lindus's 4 November account. HHS/ERG provides the broader program-cost model. The CSV is an Arca-authored illustrative calculation, not a dataset of actual trial expenses. It contains no patient data, clinical efficacy estimates or cost-effectiveness conclusions.
