AIRCHILL Pivotal Trial Detailed Cost Model & Benchmark Studies
Finance model · version 1.0 · 29 August 2026
Detailed AIRCHILL pivotal-trial cost model and benchmark studies.
A bottom-up planning model for the current ~2,300-patient cardiac-arrest pivotal concept. It shows where the money is expected to go, which assumptions dominate uncertainty, and why published grants from apparently similar cooling trials cannot be treated as complete trial costs. Every figure is labelled as sourced, derived or an AIRCHILL planning assumption.
Executive model
Three cost scenarios for the same 2,300-patient trial.
€36.0m subtotal plus 15% contingency. Approx. €18,020 per randomised patient. Requires strong academic infrastructure, low site payments, highly efficient device deployment and substantial in-kind support.
€68.4m subtotal plus 15% contingency. Approx. €34,190 per patient. This is the current P50-style management planning case.
€132.8m subtotal plus 20% contingency. Approx. €69,287 per patient. Represents expensive CRO/site pricing, larger device fleet, slower recruitment and more US-like commercial execution.
Bottom-up detail
Base case: €68.38m before contingency, €78.64m including 15% reserve.
| Cost block | Lean €m | Base €m | High €m | Base-case assumption / driver |
|---|---|---|---|---|
| Protocol, regulatory strategy & submissions | 1.50 | 2.50 | 4.00 | CIP/SAP/IB/ethics/competent-authority packages, amendments, country coordination |
| CRO / central project management | 2.50 | 4.50 | 8.00 | Programme office, vendors, sites, issue management across 24-month recruitment |
| EDC / database / data management | 1.50 | 2.50 | 4.00 | EDC build, validation, queries, device-data integration, database lock |
| Monitoring | 2.00 | 4.00 | 8.00 | Risk-based central + on-site monitoring across EMS/hospital networks |
| Statistics | 0.80 | 1.50 | 2.50 | Randomisation, interim, SAP, final analysis, regulatory outputs |
| DSMB / medical monitoring | 0.50 | 1.00 | 1.50 | Independent board, closed reports, safety review, adjudication |
| TMF / quality / audit | 0.80 | 1.50 | 2.50 | eTMF, sponsor QA, vendor qualification, audits, inspection readiness |
| Site / EMS activation | 2.00 | 4.50 | 10.50 | Base assumes ~60 networks at blended €75k activation/contract/setup cost |
| Training & simulation | 1.00 | 3.00 | 5.00 | EMS/ED/ICU training, recertification, simulation and device competency |
| Screening burden | 1.50 | 3.00 | 5.00 | Screen failures, screening logs, emergency consent/documentation work |
| Randomised-patient site payments | 11.50 | 17.25 | 34.50 | Base: 2,300 × €7,500 blended EMS/hospital research payment |
| 90-day follow-up | 0.69 | 1.38 | 2.30 | Base: 2,300 × ~€600 central/blinded neurological follow-up |
| AIRCHILL investigational device fleet | 5.25 | 11.00 | 22.50 | Lean ~700×€7.5k; base ~1,100×€10k; high ~1,200×€18.75k |
| Commissioning / installation / calibration | 0.50 | 1.32 | 3.00 | Vehicle integration, acceptance, initial calibration and deployment |
| Service / maintenance / replacement pool | 1.00 | 2.75 | 6.00 | 24-month field service, swaps, calibration and repairs |
| Study consumables | 0.70 | 1.38 | 2.50 | Patient-side circuits, gas-path disposables and treatment-arm consumables |
| Logistics / warehousing | 0.60 | 1.50 | 3.00 | Inventory, shipment, returns and cross-border logistics |
| Clinical-trial insurance | 0.70 | 1.50 | 3.00 | Jurisdiction-specific participant/sponsor liability |
| Local regulatory / translations / legal | 0.40 | 0.80 | 2.00 | Local submissions, translations and contracts |
| Health economics / PROs | 0.30 | 0.70 | 1.50 | EQ-5D, resource-use dataset and economic analysis |
| Close-out / archiving / publication | 0.30 | 0.80 | 1.50 | Site closeout, archive, CSR/publication package |
| Subtotal | 36.04 | 68.38 | 132.80 | Before contingency |
| Contingency | 15% | 15% | 20% | Recruitment, inflation, device replacement, amendments and vendor uncertainty |
| Total | 41.45 | 78.64 | 159.36 | Current management planning range |
What drives the budget
Four variables dominate AIRCHILL trial economics.
Fleet efficiency
Moving from near-fleet-wide coverage to targeted high-volume dispatch can remove hundreds of study devices and the associated training, service and logistics burden.
€5k vs €15k per randomised patient
Across 2,300 patients, each €1,000 change in the blended research payment moves total cash cost by €2.3m before contingency.
Two years versus four
Slow recruitment extends CRO, monitoring, service, insurance, training-refresh and project-management costs even if the statistical sample size does not change.
Academic network versus commercial CRO
Existing academic infrastructure and in-kind services can make the cash budget look dramatically lower without making the true economic resource use disappear.
Comparable cooling trials
Published budgets are not directly comparable — so the table shows what each number actually means.
| Study / programme | Target / actual N | Published money | What the number represents | Approx. ratio | Lesson for AIRCHILL |
|---|---|---|---|---|---|
| TTM2 | 1,900 randomised | AUD 3,247,977 + 463,957 + 154,652 + 154,652 = AUD 4.021m | Named public/research grants listed by ANZICS; not a complete economic trial-cost statement. Routine hospital infrastructure and in-kind resource use are not monetised in this figure. | ~AUD 2,116 / patient in listed grants | Large pragmatic academic RCTs can run with relatively low visible grant cash when existing ICU/research infrastructure absorbs major resource use. |
| PRINCESS | 677 randomised | Grant amounts not publicly itemised in the paper | Funded by Swedish Heart-Lung Foundation and Laerdal Foundation; cooling devices were supplied by BrainCool at no cost. | Not calculable | Free devices and manufacturer support materially reduce the sponsor cash requirement; AIRCHILL must explicitly value its RTW fleet instead of treating hardware as free. |
| PRINCESS2 | 1,022 target; first 100-patient pilot published | Named Swedish Heart-Lung Foundation, Region Stockholm and Laerdal grants; full trial budget not publicly disclosed | Publications disclose funders/grant numbers, not a complete trial budget. | Not calculable | Operational pilot-before-pivotal design is directly relevant; financing cannot be benchmarked from grant acknowledgements alone. |
| EuroHYP-1 | 1,500 planned; 98 enrolled | €11.741m total project cost; €8.082m EU contribution | CORDIS total project budget covering a broad European phase-III hypothermia programme and supporting project activities. | €7,827 per intended patient; €119,804 per actual enrollee if total project budget is divided by 98 — not a true per-patient cost | The key warning: insufficient recruitment can destroy the economics. EuroHYP-1 stopped because of slow recruitment and cessation of funding despite a substantial budget. |
| CUCUMBER / COTTIS-2 | COTTIS-2 target 400 | €3.759m total H2020 project cost; €2.936m EU contribution | EU project covers product development/market preparation and the COTTIS programme, not a pure trial-only invoice. | ~€9,398 per targeted COTTIS-2 patient if naively divided by 400 | Useful EU device-plus-trial benchmark, but the numerator contains non-trial work; it therefore cannot be used as a direct site-cost estimate. |
| US complex therapeutic-device benchmark | Modelled pivotal study mean n=565 | US$54,332 per pivotal patient in 2018 dollars | Published Medidata-based per-patient model for complex therapeutic device pivotal studies; study model estimated ~US$30.7m out-of-pocket for the pivotal phase. | US$54,332 / patient | AIRCHILL base case (~€34k/patient including device fleet and central overhead) sits below this US complex-device benchmark; the high case (~€69k/patient) is above it. |
From EuroHYP-1 to COTTIS-2
Evolution of the European cooling strategy.
There is no verified formal trial called “EuroHYP-2” in the sources used for this evidence register. The scientifically relevant development line is better described as EuroHYP-1 → COTTIS-2: EuroHYP-1 tested systemic hypothermia in a broad acute-ischaemic-stroke population, while COTTIS-2 moves the cooling intervention into a more selected large-vessel-occlusion population undergoing endovascular thrombectomy.
What EuroHYP-1 taught: the study planned 1,500 patients but enrolled only 98 and stopped after slow recruitment and cessation of funding. Its reports describe substantial organisational and logistical burden from delivering prolonged hypothermia in routine stroke care. For AIRCHILL, this means that treatment complexity, shivering management, staff workload, eligibility and enrolment speed are not secondary operational issues — they are determinants of whether a pivotal trial can finish at all.
What COTTIS-2 changes: the modern strategy concentrates on patients with large-vessel occlusion who are already undergoing thrombectomy, with cooling initiated very early around the reperfusion window. This creates a more mechanistically enriched population and avoids designing the first efficacy trial around prolonged cooling of a broad group of awake stroke patients. The CUCUMBER/COTTIS-2 programme therefore represents a more relevant contemporary European comparator for AIRCHILL-STROKE than EuroHYP-1 alone.
Why AIRCHILL-STROKE is closer to COTTIS-2: the current AIRCHILL protocol also targets selected LVO patients undergoing EVT who require general anaesthesia and invasive ventilation for clinical reasons. AIRCHILL is intended to start after airway control and as early as feasible before or during thrombectomy, with a hard requirement that study treatment must not delay arterial puncture or reperfusion. The route differs — respiratory cooling rather than transnasal plus surface cooling — but the trial-design logic is similar: earlier peri-reperfusion neuroprotection in a selected, already instrumented population.
Remaining lesson from COTTIS-2: even a focused EVT trial can face site-activation, contracting and protocol-adherence problems. AIRCHILL should therefore complete a multicentre run-in, measure eligible patients and randomisations per centre-month, and prove that cooling can be delivered without delaying reperfusion before locking the pivotal network and budget.
Source register
Primary sources used in this comparison.
TTM2: ANZICS study page lists sample size 1,900 and the four named funding amounts. Source →
PRINCESS: JAMA/PMC publication reports 677 randomised patients, independent grant funding and that BrainCool supplied study devices without charge. Source →
PRINCESS2: 2026 pilot publication reports the first 100 participants of the ongoing 1,022-patient trial and names funding sources. Source →
EuroHYP-1: European Commission CORDIS reports €11,740,832.37 total cost and €8,081,665.40 EU contribution. The final trial report and registry document 98 of 1,500 planned patients and termination after slow recruitment/funding cessation. CORDIS → · Registry →
CUCUMBER / COTTIS-2: CORDIS reports €3,759,322.50 total project cost and €2,936,494 EU contribution; COTTIS-2 is designed for 400 patients. This programme is treated here as the practical European successor strategy to EuroHYP-1 — not as a trial formally named “EuroHYP-2”. CORDIS → · COTTIS-2 →
Complex medical-device benchmark: Sertkaya et al., JAMA Network Open 2022, reported a pivotal-study cost parameter of US$54,332 per patient in 2018 dollars and a mean pivotal sample of 565. Source →
AIRCHILL figures: All site-payment, device-unit-cost, staffing, CRO and contingency values are management assumptions until replaced by quotations. RTW deployment model → · Cardiac-arrest CIP → · Clinical Investigation Package →