Global Impact Atlas · Deep Research snapshot 10 Aug 2026

Local data first. Assumptions only where the data stop.

The Global Impact model now separates epidemiology, clinical addressability, real-world reach and assumed treatment effect. The atlas covers the USA, China, Taiwan, Qatar, the EU member states, Germany by Bundesland, China by province, Switzerland, the United Kingdom, the United Arab Emirates (with Dubai-specific context where available), Saudi Arabia, India, Brazil, New Zealand, Australia, Japan, South Korea, Canada, Mexico and Argentina. Every default retains its provenance: SOURCED, DERIVED, TRANSFERRED, ASSUMPTION, HISTORICAL or DATA GAP. Transferred values are comparison proxies, not national epidemiology.

85+country and subnational research regions in the current data architecture
6indication models: OHCA, thrombectomy stroke, neonatal HIE, severe heat stroke, severe TBI and concussion
No false zeroswhen a comparable local denominator is unknown, the model displays DATA GAP rather than silently scaling Germany
0 AIRCHILL patientsall impact outputs remain scenario arithmetic, not clinical evidence
Method correction from Deep Research. ROSC, hospital admission and “survived event” are outcomes or downstream states, not generic eligibility filters when cooling could be initiated earlier. OHCA is therefore stored as separate stages where available: assessed → resuscitation attempted → ROSC → hospital admission → discharge survival → favourable neurological outcome. The impact calculator applies user-editable addressability and reach only after the epidemiologic/procedural denominator.

Clickable evidence map

Click a country. Drill into Germany or China.

Map fill intensity represents the selected metric when a numeric default exists. The source-quality badge remains separate so a large derived estimate is never visually mistaken for a measured registry count. Switch between absolute annual counts and rates where the denominator is meaningful.

SOURCED direct/localDERIVED formula from local/source inputsTRANSFERRED external comparison proxyASSUMPTION planning inputHISTORICAL older/local proxyDATA GAP no comparable default
World · selected research markets

What changed from the previous model

Better local defaults — and more visible gaps.

RegionOHCA defaultThrombectomy defaultHIE defaultHeat strokeSevere TBIConcussion
No transferred TBI/concussion default outside its evidence context. Broad TBI hospitalizations, trauma-registry cohorts and community concussion estimates measure different populations. The previous population-scaled German defaults are no longer used as automatic country values. A transferred sensitivity is never activated automatically; the visitor must explicitly choose it, and it remains labelled TRANSFERRED. A different manually entered value is labelled USER INPUT.

Interactive societal impact & break-even lab

Source default → addressability → reach → effect → societal value.

Select a region from the map or dropdown. Sourced and locally derived denominators can load as defaults. TRANSFERRED comparison proxies are deliberately not activated automatically; the visitor must explicitly choose “Use transferred proxy”. If no compatible denominator exists, enter a scenario value and it will be marked USER INPUT while the original data gap remains visible.

Active annual denominator
Patients treated / year
Additional favourable outcomes / year
Cumulative favourable outcomes
Annual gross societal model value
Annual intervention cost
Annual net societal model value
Break-even variable cost / treated patient
Discounted cumulative net monetary value
Source status


Interpretation boundary. +1/+3/+5 pp are scenario sensitivities, never AIRCHILL efficacy estimates. “Years” means repeated new annual cohorts. Addressability, reach, price and societal value are separate editable assumptions.

Subnational methodology

Germany and China: direct overrides beat national scaling.

Germany · 16 Bundesländer

2025 state populations are sourced from Destatis. Until state-specific clinical counts are loaded, OHCA is shown only as a TRANSFERRED European comparison proxy (55/100,000), while thrombectomy remains a DERIVED population-share allocation of the sourced German 2024 total of 21,609. Neither is presented as a measured state volume. Reimbursement uses a different geography: procurement and fee planning should ultimately drill down to Landkreise, kreisfreie Städte and other responsible rescue-service bodies rather than treating the Bundesland epidemiology layer as the payment layer. Open the reimbursement atlas →

Destatis state populations · German 2024 stroke/MT analysis

China · 31 province-level regions

2024 population and crude birth rates come from the NBS Statistical Yearbook 2025. Where no provincial clinical registry value is available, OHCA uses the national BASIC-OHCA rate/attempt fraction and MT uses the national stroke-centre procedure rate as explicit DERIVED placeholders. Zhejiang OHCA and Guangdong MT override those placeholders with direct regional evidence.

NBS Statistical Yearbook 2025 · Zhejiang OHCA · Guangdong thrombectomy

EU · all 27 member states

Country populations use the 2025 Eurostat population dataset. Outside national direct sources, the clickable member-state map uses EU-level OHCA and thrombectomy rates only as TRANSFERRED comparison proxies. The OHCA proxy is 55/100,000 and is not considered interchangeable with the previously used 89/100,000 figure because the underlying Utstein/registry denominator may differ. These proxies are not national registry epidemiology and are not auto-loaded into the calculator.

Eurostat population 2025 · EuReCa

HIE needs ranges, not a magic point

The USA uses a clinically reviewed population-based rate around 1.7/1,000 births. China is shown with a 3–6/1,000 literature range and a 4.5/1,000 midpoint only as an ASSUMPTION for model conversion. Other markets retain a visible gap or transferred planning rate until a comparable local source is loaded.

US HIE source · China HIE literature

Research-draft status

What is strong — and what still needs local registry extraction.

USA. CARES provides a strong 2024 OHCA denominator; the national MT estimate remains older (2021). HIE now uses a local clinically reviewed incidence rather than the previous generic 1.5/1,000 assumption. Severe heatstroke, severity-specific TBI and community concussion remain data gaps. CARES/AHA · HIE

China. National OHCA attempted is derived from BASIC-OHCA; the thrombectomy default is updated to 81,865 procedures in reporting stroke centres (2023), replacing the older 104,704 arithmetic estimate. BASIC-OHCA · China Stroke Report

Germany. The thrombectomy default is updated from 18,809 to 21,609 in 2024. The state-level table is currently a transparent population-share derivation until local procedure extracts replace it. Severe TBI remains a historical moderate-to-severe proxy, not an intubated severe-TBI count. German MT analysis · TBI source

Canada. About 60,000 OHCA/year is a sourced national magnitude. CCDSS reports 97,010 strokes in adults in 2023/24, but this is not an EVT count, so thrombectomy remains a data gap instead of being inferred from all stroke. Heart & Stroke · CCDSS

UAE & Dubai. The register now carries the official system context without converting it into clinical counts: UAE population ~11.29m in 2024 and 18,685 inpatient beds (~1.66/1,000); Dubai resident population ~4.471m in Q3 2025. OHCA, EVT, HIE, severe heatstroke, severe TBI and concussion remain denominator-specific data gaps until DHA/DCAS/MOHAP extracts are available. MOHAP Open Data · DHA Open Data

Saudi Arabia. Population is now framed at ~35.3m for 2024. The SOHAR registry is retained as regional evidence (3,671 OHCA cases across selected regions, 2019–2022), not a national incidence estimate; national SRCA/MOH and claims extracts remain the next step. SOHAR · GASTAT

Japan. Population is now ~123.05m (2025 research frame). Nationwide EMS infrastructure is documented, including ~7.64m ambulance dispatches in 2023 and 5,415 ambulance teams by Apr 2024, but dispatch activity is not an OHCA denominator. The atlas therefore keeps clinical values open while targeting FDMA Utstein OHCA and MHLW NDB EVT extraction. FDMA · MHLW NDB Open Data

South Korea. The previous generic gap has been replaced for OHCA: national surveillance reported 33,586 OHCA cases transported by 119 EMS in 2023. With the 2025 population frame, the atlas can calculate a system rate while keeping AIRCHILL addressability separate. EVT remains a targeted HIRA/NHIS claims extraction. National OHCA surveillance · HIRA

Brazil, New Zealand & Australia. Population denominators are now loaded so existing sourced/derived clinical frames can be rate-normalised: Brazil 213.421m (2025), New Zealand ~5.325m (Jun 2025), Australia ~27.5m research frame. Broad TBI, heat and hospital-only concussion measures keep their phenotype warnings and are not relabelled as device-eligible populations.

UK. OHCAO, SSNAP, HES and TARN are the preferred next extraction paths. A historical OHCA magnitude can be displayed as HISTORICAL/ASSUMPTION, but current MT, severe TBI and community concussion stay open until denominator-compatible extracts are loaded. OHCAO · SSNAP

India, Mexico, Argentina. Public national denominators are currently insufficient for several models. The atlas therefore shows gaps instead of scaling foreign incidence rates to very large populations. Local registry/authority outreach is the next step. India OHCA review · Mexico OHCA literature · Argentina stroke programme

Heatstroke. This remains one of the largest denominator gaps. The productive hierarchy is: local severe heatstroke count → T67.0 hospitalizations → ICU/organ-failure subset → only then a standardized per-100 sensitivity. CDC heat surveillance · Destatis diagnosis data