Sources, Assumptions & Data Gaps
Transparency register · site-wide
Sources, assumptions and the evidence we plan to generate next.
Medical Cooling separates sourced facts from calculations and planning choices. A source link means that the linked source supports the cited external fact; it does not validate AIRCHILL. Derived values show arithmetic from sourced/local inputs. Transferred values are explicit cross-market comparison proxies and are kept separate from national epidemiology. Assumptions are editable planning inputs. Missing values are deliberately left open rather than filled with false precision.
Linked to the underlying publication, registry, authority or public dataset wherever an online source is available.
The formula or derivation must remain visible and is not presented as an independently measured value.
A local denominator multiplied by a foreign, European or other non-local rate. It may support sensitivity analysis but is not national epidemiology and is not auto-loaded into the impact calculator.
Used to test a scenario. It is not a clinical effect, authority decision, quotation or forecast.
The model leaves the epidemiologic default open rather than manufacturing a local number. A visitor may enter a scenario value, which is then labelled USER INPUT.
Current assumption and data-gap register
This table is the site-wide control layer for the values most likely to be misunderstood as facts. It should be updated whenever a new registry count, authority decision, quotation, product design freeze or clinical result replaces an assumption.
| Area | Value / question | Status | Current handling | What would replace it |
|---|---|---|---|---|
| Cardiac arrest · Germany · national denominators | 2024: ≈136,000 OHCA overall; ≈54,000–67,000 EMS resuscitation attempts | SOURCED + VERSIONED | 2022 was reported at about 60,000 EMS resuscitation attempts. For 2023, the first annual-report communication used about 55,000; a later broader extrapolation was about 71,000. For 2024, the annual report yields about 54,000–67,000 depending on total versus reference-group extrapolation, while the later nationwide communication uses about 67,000. The ≈136,000 figure is the broader all-OHCA event denominator, not the EMS-resuscitation denominator. | Next audited German Resuscitation Register report using the same denominator definition and transparent extrapolation method. |
| Temperature management · Germany | 2024 DGRR reference group: 17.3% of admitted OHCA patients; Cardiac Arrest Centers: ≈35% of admitted OHCA patients | SOURCED PRACTICE RATES | These are treatment-prevalence measures from different care cohorts, not efficacy estimates or target temperatures. ERC–ESICM 2025 emphasizes temperature control and fever prevention after ROSC; routine induced deep hypothermia is not a universal treatment requirement. | Future registry data using a harmonized temperature-control definition plus protocol-specific target, duration and patient phenotype. |
| Cardiac arrest · Germany | 7,500 potentially narrower early-cooling patients/year | ASSUMPTION | Operational modelling cohort; not registry-derived eligibility. | Prospective inclusion/exclusion study using final intended use and EMS workflow. |
| Stroke · Germany | 21,609 thrombectomies in 2024 | SOURCED SYSTEM FRAME | Updated nationwide hospital/procedure frame. AIRCHILL eligibility remains unknown and must be applied as a separate addressability parameter. | Final airway/timing eligibility study in EVT population plus state-level procedure extracts. |
| Neonatal HIE · generic fallback | 1.5 per 1,000 live births | TRANSFERRED PLANNING RATE | Used only where a comparable local HIE source has not yet replaced it; not device eligibility. | Market-specific clinically validated HIE incidence plus current therapeutic-hypothermia eligibility. |
| HIE · USA | ~1.7 per 1,000 live births; ~6,131/year when applied to 2025 births | LOCAL RATE + DERIVED COUNT | Population-based clinically reviewed US incidence is used instead of the generic 1.5/1,000 fallback. | Current national clinically confirmed moderate/severe HIE and therapeutic-hypothermia eligibility counts. |
| HIE · China | Literature range ~3–6 per 1,000; model midpoint 4.5/1,000 | RANGE / ASSUMPTION | The atlas exposes the range; 4.5/1,000 is only the midpoint used for an editable model conversion, not a measured national incidence. | Current nationally representative clinically validated HIE incidence and treatment eligibility. |
| Heat stroke · Germany | Official hospital system frame available; severe/device-eligible subset remains unknown | SOURCED SYSTEM FRAME ELIGIBILITY GAP | Destatis reports an average of more than 1,400 inpatient treatments per year in 2003–2023 for damage due to heat and sunlight; more than 800 were recorded in 2023. This direct-coded hospital frame is broader than clinically defined severe heat stroke. Separately, RKI heat-attributable mortality estimates were about 4,500 in 2022, 3,200 in 2023, 3,000 in 2024 and 2,500 in 2025. Those statistical excess-mortality estimates are not a treatable heat-stroke cohort. | Prospective or registry-based severe-heat-stroke denominator with ventilation/intubation status and explicit clinical definition. |
| Severe TBI · Germany | 2013–2017: ≈8,220 observed moderate/severe TBI per year; corrected incidence estimate 13.6/100,000; ≈11,400/year only after historical rate scaling | SOURCED HISTORICAL DERIVED CURRENT-SCALE FRAME | The TraumaRegister DGU analysis observed 41,101 cases over five years (≈8,220/year) and estimated 13.6/100,000 after correcting for under-coverage and prehospital deaths. Scaling that historical rate to the current population gives ≈11,400, but this is not a current registry count. The site’s 7,500 operational cohort remains an ASSUMPTION. In the historical cohort, 43.6% were intubated prehospital. | Current national severe/intubated TBI denominator using explicit GCS, ventilation and timing criteria plus prospective AIRCHILL eligibility screening. |
| Concussion · Germany | ~44,000 medically diagnosed sports-related concussions/year · source vintage 2016 | SOURCED HISTORICAL | DOSB published this order-of-magnitude estimate in 2016 and already noted likely under-diagnosis. It remains useful as historical context for adjacent non-invasive research, but it is not current national surveillance and is not a current AIRCHILL indication. | Current national community/sports surveillance using a comparable concussion definition. |
| International severe TBI / concussion | No automatic local default where a comparable severity/community denominator is missing | DATA GAP | Deep Research removed the previous automatic German-rate population scaling from country defaults. Broad TBI hospitalization and hospital-only concussion counts are not treated as equivalent local epidemiology. | Local severity-specific trauma registry / surveillance data using comparable definitions. |
| USA OHCA | 263,711 EMS-treated non-traumatic OHCA (CARES 2024) | SOURCED | Hospital admission / ROSC are no longer used as automatic eligibility filters because they are downstream outcomes. AIRCHILL addressability and reach are separate editable parameters. | Final trial-eligible phenotype, timing window and actual adoption. |
| China OHCA | Population × BASIC-OHCA incidence; attempted-resuscitation filter | DERIVED | Population-scaled planning estimate, not audited annual national volume. | Current nationwide annual EMS-resuscitation count. |
| China thrombectomy | 81,865 procedures in reporting stroke centres (2023) | SOURCED NETWORK COUNT | Replaces the older 104,704 arithmetic estimate as the default national stroke-centre procedure frame. It is still not the same as an audited all-China population procedure census. | Current nationwide audited procedure count with province-level coverage. |
| China · 31 province-level regions | 2024 population + crude birth rates; provincial OHCA/MT direct overrides where available | MIXED SOURCE HIERARCHY | NBS supplies local denominators. Provinces without clinical direct data use visible national-rate derivations. Zhejiang OHCA and Guangdong MT override the national scaling with local evidence. | Province-specific EMS, stroke-centre, HIE, heatstroke, severe-TBI and concussion registries. |
| Germany · 16 Bundesländer | Exact 31 Dec 2025 populations; national OHCA/EMS-resuscitation and MT totals may be allocated by population share only when state-level clinical counts are unavailable | DERIVED OHCA DERIVED MT | For current German comparisons, the state layer should derive from the sourced national ≈67,000 EMS-resuscitation-attempt frame rather than the old transferred European 55/100,000 proxy. MT may likewise allocate the sourced 21,609 German 2024 total by population share. Neither allocation is a measured Bundesland clinical volume, and the broader ≈136,000 all-OHCA event denominator is kept separate. | State-/network-level Reanimationsregister and thrombectomy procedure extracts using compatible denominator definitions. |
| Taiwan OHCA | Historical incidence scaled to 2025 population | HISTORICAL / DERIVED | Not presented as a current national annual count. | Current national OHCA registry volume. |
| Taiwan thrombectomy | ~2,000/year | HISTORICAL DIRECT | NHIA figure from before reimbursement-window expansion; current activity may be higher. | Current NHIA annual procedure count. |
| Qatar thrombectomy | ~38/year | HISTORICAL AVERAGE | Derived from registry period and reported MT rate. | Current annual national stroke-registry procedure count. |
| Qatar HIE | Current official live-birth datasets now available | SOURCED BIRTH DENOMINATOR HIE RATE TRANSFERRED | Qatar Open Data/National Planning Council now exposes annual, quarterly and monthly registered live-birth datasets, replacing the stale 2021 denominator. Any HIE count still requires an explicit HIE incidence assumption or Qatar-specific clinical rate. | Qatar-specific clinically validated HIE incidence and therapeutic-hypothermia eligibility. |
| EU OHCA comparison proxy | 55 OHCA/100,000 × national population where no compatible local value is loaded | TRANSFERRED PROXY | The proxy is retained for comparison only. The previously used 89/100,000 value is not treated as interchangeable because EuReCa/Utstein denominator definitions and registry coverage can differ. Transferred values are not auto-loaded into the calculator. | Denominator-compatible national EMS/OHCA registry extraction for each member state. |
| EU thrombectomy comparison proxy | ~66.4k/year across the current 2025 population snapshot | TRANSFERRED PROXY | Population × 249 strokes/100k × 80% ischaemic share × 7.4% MT sensitivity. Germany’s sourced 21,609 MT in 2024 overrides this proxy. | Country-level audited EVT/MT procedure counts from national claims, hospital or stroke-registry systems. |
| EU economic calculator · currency | Country-specific ISO 4217 currency | CORRECTED SYSTEM FIELD | EU membership no longer implies EUR. Czechia = CZK, Denmark = DKK, Hungary = HUF, Poland = PLN, Romania = RON and Sweden = SEK; Bulgaria is treated as EUR in the 2026 research frame. | Automated currency metadata plus dated FX source for cross-currency comparisons. |
| Primary value vs proxy | National/direct denominator and transferred sensitivity are separate concepts | GOVERNANCE RULE | SOURCED/DERIVED local values may load as defaults. TRANSFERRED values require explicit user activation. DATA GAP is never converted to zero, and USER INPUT does not overwrite the source record. | Versioned machine-readable register with primary_value, proxy_value, evidence_status and vintage_status. |
| United Arab Emirates / Dubai | UAE population ~11.29m (2024); 18,685 inpatient beds (~1.66/1,000). Dubai resident population ~4.471m (Q3 2025). | CLINICAL DATA GAP / SOURCED SYSTEM FRAME | Dubai is retained as a distinct emirate/subregional research target. UAE/Dubai system capacity and population are sourced, but they do not substitute for OHCA, EVT, HIE, severe heatstroke, severe TBI or concussion denominators. | DHA / Dubai Corporation for Ambulance Services / MOHAP: EMS-resuscitation OHCA, EVT procedures, T67.0 + hospital/ICU status and severe S06.x + ventilation/GCS filters. |
| Saudi Arabia | Population ~35.3m (2024); SOHAR recorded 3,671 OHCA cases in selected regions during 2019–2022 | NATIONAL DATA GAP / REGIONAL SOURCE | SOHAR is retained as regional evidence and is not scaled into a national Saudi OHCA count. No generic international rate is substituted. | SRCA/MOH national EMS denominator and catchment coverage; claims-based EVT; national T67.0 heatstroke and severity-specific trauma extracts. |
| Brazil | Population 213,421,037 (1 Jul 2025); OHCA ~100k derived frame; SUS MT ~759 in 2024; TBI hospitalisations ~131,015/year historical mean | MIXED: DERIVED / SOURCED | OHCA is a literature-derived planning magnitude, MT is a public-system procedure signal, and TBI hospitalisations are broader than severe/intubated TBI. | National OHCA registry-quality denominator; all-payer MT count; DATASUS severity/ICU filters; T67.0 and S06.0 extracts. |
| New Zealand | Population ~5.325m (Jun 2025); 2,466 treated OHCA in Jul 2024–Jun 2025; mild-TBI population evidence available | SOURCED / HISTORICAL CONTEXT | OHCA can be used as a national treated denominator. Mild-TBI evidence is not relabelled as total concussion or severe TBI. | National EVT count, HIE denominator, severe heatstroke and severity-specific TBI extracts. |
| Australia | Population ~27.5m research frame; ~26,000 OHCA/year; ~710 extreme-heat hospitalisations/year; 10,700 concussion hospitalisations in 2021–22 | SOURCED SYSTEM FRAMES | Heat hospitalisations are broader than clinically defined severe heatstroke, and concussion hospitalisations understate community burden. | National EVT procedure count, clinically defined severe heatstroke subset and severe/intubated TBI denominator. |
| Japan | National OHCA denominator available: 140,575 transported cardiopulmonary-arrest patients in 2023; 90,550 cardiogenic cases | SOURCED NATIONAL OHCA OTHER INDICATIONS PARTIAL | Japan FDMA reports 140,575 transported cardiopulmonary-arrest patients in 2023, including 90,550 cardiogenic cases; 28,354 cardiogenic arrests were bystander-witnessed. These are national system counts, not AIRCHILL eligibility. EVT, HIE, severe heatstroke, severe TBI and concussion still require separate indication-specific extracts. | MHLW NDB Open Data for EVT and national heat/trauma/claims datasets for the remaining indications. |
| South Korea | 33,586 OHCA cases transported by 119 EMS in 2023; population frame ~51.685m in 2025 | SOURCED NATIONAL OHCA | The national 119-EMS surveillance value replaces the previous OHCA data gap. It is a system denominator, not AIRCHILL eligibility. EVT, HIE, severe heatstroke, severe TBI and community concussion remain separate gaps. | HIRA/NHIS procedure-code extraction for EVT; national heat/trauma/claims datasets for the remaining models. |
| Future post-operative cooling model | No universal postoperative denominator is defined | DATA GAP / DEFINITION REQUIRED | No country is assigned a postoperative cooling population. A future model must first define target procedures, then apply ICU/ventilation and clinical-eligibility filters. | Frozen target-procedure list plus country-specific procedure counts and prospective eligibility definition. |
| International economics | Value per favourable outcome | DATA GAP | Global Impact calculator defaults to zero; user must enter a local value. | Country-specific societal-cost / QALY / payer analysis for the exact endpoint. |
| All impact models | +1 / +3 / +5 percentage-point effects | ASSUMPTION | Common sensitivity grid only; not AIRCHILL efficacy estimates. | Prospective randomized AIRCHILL clinical evidence. |
| Cold-air clinical operating envelope | Provisional lower first-generation setpoint: −20 °C; human escalation concept −10 → −15 → −20 °C | ASSUMPTION | Chosen because prolonged product-specific endotracheal porcine evidence exists around −20 °C. Human studies have briefly delivered approximately −35 to −40 °C, but these did not establish tissue/cilia safety; longer/high-ventilation exposure at warmer temperatures can produce airway inflammation/epithelial stress. Temperature alone is not a dose. | Bench dose characterization + route-specific prolonged preclinical confirmation + staged human safety/performance data agreed with regulators/ethics/DSMB. |
| Cold-air cilia threshold | Claim that −32 °C immediately destroys cilia | NOT ESTABLISHED / DATA GAP | Human ciliated-cell evidence supports temperature-dependent reduction in ciliary beat frequency; sport/cold-air literature supports mucociliary impairment with repeated cold/dry exposure, but no validated −32 °C destruction threshold was identified. | Direct human route-specific ciliary/mucociliary function and recovery study across defined temperature, humidity, flow and duration. |
| Cold-air ongoing trial status | CLARINET NCT07220928 completed; CLARINET2 planned but separate public registry not verified as of 23 Aug 2026 | SOURCED REGISTRY STATUS | CLARINET enrolled 60 adults and used −15 °C eucapnic cold-air challenge with airway mechanics, symptom, inflammatory and epithelial-integrity endpoints. Results are not posted in the reviewed ClinicalTrials.gov record. CLARINET2 is described as a planned follow-on, not counted as an ongoing registered trial. | Posted CLARINET results and/or verified CLARINET2 registry/publication. |
| Study planner | Control rates, effect, sites, recruitment, attrition and costs | ASSUMPTION | Editable planning defaults; binary power approximation only. | Protocol-specific current event rates, CRO/site quotes and regulator/statistician agreement. |
| Hardware scope / make-buy | Monitoring, suction, ECG/defibrillation, telemedicine and other mature peripheral functions | MANAGEMENT DESIGN PRINCIPLE | These functions are planned primarily as purchased components, OEM modules or standard interfaces rather than new proprietary subsystem R&D. This can support a broader first certifiable hardware platform, but it does not remove manufacturer responsibility for supplier qualification, interfaces, electrical/software compatibility, risk management and system-level V&V. | Frozen architecture, named suppliers/OEM agreements, component evidence, interface specifications and completed integrated-system verification. |
| EU regulation | MDR class IIb | REGULATORY HYPOTHESIS | Conservative planning case for the current minimum certifiable configuration (MCC): a complete proprietary ventilation/cooling core plus as much mature purchased/OEM platform hardware as can be integrated without disproportionate classification or evidence burden. Scope is reduced primarily through intended purpose and claims, not through a disposable hardware-limited MVP. | Final intended purpose/design plus Notified-Body/competent-authority determination. |
| USA regulation | 510(k), De Novo fallback | REGULATORY HYPOTHESIS | Pre-Sub is the first formal gate. | FDA feedback on intended use, product code, predicate and evidence package. |
| China regulation | Imported Class III | REGULATORY HYPOTHESIS | Formal NMPA classification comes first. | NMPA classification / attribute determination. |
| Oxygen / LOX boundary | Device administration vs integral drug-device / medicinal-product route | UNRESOLVED ARCHITECTURE-DEPENDENT | External separately supplied medical oxygen is the lowest-complexity base case; integrated LOX is treated as a separate decision. | Frozen gas architecture plus NB/medicine-authority, FDA OCP and NMPA determination as applicable. |
| Standards | Exact applicable editions / national adoptions | TO BE FROZEN | Current matrix is an applicability map, not a declaration that every standard applies. | Final intended use, architecture and submission-date harmonized/recognized standards list. |
| Standards / component compliance budget | Subsystem-specific testing, certification, OEM integration and market-entry cost/time ranges | MANAGEMENT ENVELOPE | The detailed standards, cost & time matrix separates sourced legal/fee facts from planning estimates. Testing, laboratory, certification and engineering ranges are not authority fees or supplier quotations and are not additive calendar durations. | Named OEM, test-lab, Notified-Body, FCC/SRRC, CRO and regulatory-consultant quotations plus the frozen architecture and submission sequence. |
| Regulatory timeline / cost | EU, FDA, NMPA programme ranges | MANAGEMENT ENVELOPE | Planning sensitivity, not authority quotation or approval promise. | Notified-Body/FDA/NMPA feedback, lab/CRO quotes and final evidence plan. |
| Funding | €7.5m current round; €10–15m follow-on; €12/17/24m low/base/high | MANAGEMENT MODEL | Not committed financing and not quotations. | Signed financing, supplier/CRO quotes and updated development plan. |
| Founder capacity | Weekly hours, compensation, transition date, employment/IP terms | DATA GAP | Not modelled as full-time availability until documented. | Executed employment/secondary-employment and IP arrangements. |
| Patent control | Assignment / exclusive licence to operating company | OPEN CORPORATE ACTION | Public registers name the inventor/applicant; company ownership is not claimed. | Executed assignment or exclusive worldwide sublicensable licence plus register updates. |
| Germany EMS fleet | Absolute ambulance fleet count | DATA GAP | No official national count used; any fleet size must be derived from published deployment density. | Audited national vehicle inventory or validated district-level aggregation. |
| Commercial model | Final device price, consumable price, gross margin, adoption | UNVALIDATED COMMERCIAL ASSUMPTION | Peer and dealer prices are anchors only. | Frozen BOM, manufacturing quotes, tenders, customer interviews and signed procurement data. |
Register architecture. Evidence status and data vintage are separate concepts: a directly measured older registry value can be SOURCED + HISTORICAL. USER INPUT belongs to the active scenario and never overwrites the source record. The production target is a versioned machine-readable register with explicit primary values, proxies, source URLs, denominator definitions and review dates.
Linked primary and public sources used across the site
This directory prioritizes original publications, official registries, authorities and public datasets. Some project-internal documents, including the BMBF final report and investor model, are not public online and are therefore identified as internal/project sources rather than given a fabricated link.
AIRCHILL mechanism: Sedlacik et al. · high-flow cold-air MRI thermometry. The BMBF Go-Bio 031A530 final report and ISMRM 2016 poster are project sources and are available on request rather than linked to a non-existent public file.
Cold-air inhalation / airway safety: Marain et al. · 2026 systematic review of 39 cold-air challenge studies · CLARINET · NCT07220928 · Jaeger et al. · −40 °C short-exposure physiology · Hartung et al. · −35 °C short-exposure physiology · Eklund et al. · −15 °C / 50-min epithelial-stress study · Clary-Meinesz et al. · temperature and human ciliary beat frequency · Hanstock et al. · sub-zero exercise/airway review · D’Amato et al. · cold and respiratory tract review. These sources support a multidimensional temperature × humidity × ventilation × duration safety model; they do not establish one universal minimum safe inspired temperature.
Cardiac arrest: TTM2 · PRINCESS · PRINCESS2 design paper · PRINCESS2 100-patient pilot · HYPERION · German Resuscitation Register annual reports · 2024 nationwide German denominator reconciliation · ERC–ESICM 2025 post-resuscitation care. German 2024 values are deliberately separated into ≈136,000 all-OHCA events, ≈54,000–67,000 EMS resuscitation attempts depending on register/reference extrapolation, and downstream admitted-patient temperature-management rates. These denominators are not interchangeable.
Stroke: COTTIS matched-pair analysis · COTTIS-2 / NCT06301412.
Neonatal HIE: Shankaran et al. whole-body hypothermia trial.
Heat stroke: RKI heat-attributable mortality · Destatis heat/sunlight hospital treatment statistics · 2003–2023 update.
Severe TBI: LTH-1 · POLAR · Eurotherm3235.
Concussion: selective head–neck cooling cohort.
Therapeutic gases: HYBRID II hydrogen trial · TAME mild-hypercapnia trial.
Preclinical timing and dose: Lyden et al. depth–delay–duration study · intra-arrest cooling in a murine cardiac-arrest model · selective brain cooling during CPR in a porcine prolonged-arrest model · combined intra- and post-arrest cooling in a rat asphyxial-arrest model · delayed mild hypothermia after focal ischaemia · depth and duration in focal cerebral ischaemia. These models support a testable timing/thermal-dose rationale; they do not establish an AIRCHILL human effect size.
EU regulation: Regulation (EU) 2017/745 (MDR).
Clinical investigations & human factors: ISO 14155:2026 · IEC 62366-1:2015 + Amd 1:2020 · FDA Applying Human Factors and Usability Engineering to Medical Devices.
FDA: 21 CFR 868.5895 · continuous ventilator · Q-Submission Program · IDE process · MDUFA fees · OCP RFD / Pre-RFD process.
Newly closed / narrowed data gaps (August 2026): Japan FDMA now provides a national OHCA system denominator for 2023 (140,575 transported cardiopulmonary-arrest patients; 90,550 cardiogenic; 28,354 cardiogenic and bystander-witnessed). Qatar Open Data provides current annual birth and mortality statistics, with additional quarterly/monthly live-birth datasets, replacing the stale 2021 birth denominator. Destatis heat/sunlight hospital statistics provide a German direct-coded system frame (more than 1,400 inpatient treatments/year on average in 2003–2023; more than 800 in 2023), but not a severe/intubated heat-stroke denominator. RKI/BMG heat-attributable mortality estimates are kept as a separate population-level series because they are not clinically diagnosed heat-stroke counts. Destatis 2024 hospital diagnosis data provide four-digit ICD-10 counts and are the preferred national source frame for S06.x head injury/concussion; severity, intubation and sports attribution remain separate eligibility gaps. New Zealand national OHCA 2024/25 reports 2,466 ambulance-treated cases. These system denominators are not AIRCHILL eligibility or efficacy estimates.
International impact inputs: The Global Impact Atlas & Models covers USA, China, Taiwan, Qatar, all EU member states, Germany by Bundesland, China by province, Switzerland, United Kingdom, UAE with Dubai-specific context, Saudi Arabia, India, Brazil, New Zealand, Australia, Japan, South Korea, Canada, Mexico and Argentina. The evidence vocabulary now distinguishes SOURCED, DERIVED, TRANSFERRED, ASSUMPTION and DATA GAP; transferred proxies require explicit activation in the calculator. Key system sources include Eurostat, UAE MOHAP, Dubai DHA, Saudi GASTAT, Japan FDMA, Japan NDB Open Data, South Korea OHCA surveillance, HIRA, Hato Hone St John, AIHW and CIHI.
EMS reimbursement research: The Global EMS Reimbursement Atlas separates who funds the ambulance service, who procures equipment, the likely first access route and whether a separate device-payment mechanism is actually documented. Germany is treated as a local remuneration geography under §133 SGB V; England is linked to the NHS Payment Scheme; the United States to the Medicare Ambulance Fee Schedule; Australia to state/territory ambulance financing rather than Medicare. Markets without comparable payment-path research remain DATA GAP.
Health economics and epidemiology used in impact models: Soler-Font et al. · first-year stroke costs by 90-day mRS (mRS 0–2 €18,568; mRS 4–5 €52,859; societal perspective) · CENTER-TBI intramural healthcare costs (PMID 37212277; median €3,800 mild, €37,800 moderate, €60,400 severe) · Kurinczuk et al. · HIE epidemiology (PMID 20554402; 1.5/1,000 live births, transferred high-income-country anchor) · Destatis · 654,241 live births in Germany in 2025 · Kruse et al. · lifetime costs of cerebral palsy (PMID 19416329; historical Danish lifetime cost anchor ≈€0.80–0.86m) · Nova-Díaz et al. · 2026 Spanish paediatric CP societal-cost model (mean €102,135/year; severe motor impairment ≈1.96× mild) · German Federal Ministry of Health · 2026 ambulatory long-term-care benefits (€1,859/month grade 4; €2,299/month grade 5). These are source anchors, not AIRCHILL efficacy estimates. Cross-country applications remain TRANSFERRED or ASSUMPTION unless a local denominator/cost study exists.
Evidence rationale for the +1 / +3 / +5 percentage-point sensitivities: These are not clinical standards and not AIRCHILL effect estimates. For OHCA, PRINCESS reported CPC 1–2 at 90 days in 16.6% vs 13.5% (absolute point estimate +3.1 pp, not statistically significant), while HYPERION reported 10.2% vs 5.7% (+4.5 pp) in non-shockable cardiac arrest; thus +1 pp is a deliberately cautious lower sensitivity, +3 pp is close to the PRINCESS point estimate and +5 pp is approximately the HYPERION magnitude. For stroke, COTTIS reported mRS 0–2 in 68.2% vs 29.5% (+38.7 pp) in a small non-randomized matched analysis; because the authors themselves caution that the effect is surprisingly large, the atlas deliberately uses much smaller +1/+3/+5 pp stress tests rather than transferring the observed signal. For neonatal HIE, established whole-body hypothermia reduced death or moderate/severe disability from 62% to 44% (18 pp), but any new transport/early-cooling strategy would be incremental to established care, so only small incremental sensitivities are used. For severe TBI, POLAR found essentially no benefit (48.8% vs 49.1% favorable GOSE), so 0 pp is the evidence-aligned reference and +1/+3 pp are more defensible exploratory positive cases than +5 pp. For concussion, percentage points are not the preferred endpoint; symptom severity and return-to-activity/time-to-play should replace generic pp scenarios in future model versions. PRINCESS · HYPERION · COTTIS · NICHD HIE · POLAR.
Heat-stroke sensitivity rationale: The best-supported anchor is rapid cooling rather than a pre-specified survival effect. In 274 exertional heat-stroke cases treated with cold-water immersion, mean cooling was 0.22 °C/min with 100% survival. A 521-patient evidence synthesis found no deaths among patients treated with modalities providing adequate cooling rates above 0.15 °C/min, whereas 23 deaths occurred among cases treated with insufficient cooling; complications were also substantially more frequent with slower cooling. Therefore +1/+3/+5 pp may be retained only as a conservative outcome sensitivity around an observational mortality signal of roughly 4–5 pp, while a prospective AIRCHILL feasibility study should primarily test cooling rate, time to target temperature, workflow and safety. These data do not establish AIRCHILL efficacy and cold-water immersion remains a distinct, established cooling modality for exertional heat stroke. DeMartini et al. · Cooling-rate and survival review.
Patents: EP3509683B1 · US11395900B2 · WO2018046128A1.
Evidence · Research Register · Global Impact Models · Regulatory & Reimbursement · Legal Disclaimer