Research register · frozen 10 August 2026
One corpus. Different levels of evidence.
Hypothermia is not a single intervention. Population, timing, target temperature, duration, route, comparison treatment, spatial distribution and rewarming determine what a study actually tests. This register makes those distinctions visible and keeps negative, neutral and unfinished work in the same map as positive studies.
How to read this site
Evidence is a ladder, not a pile.
A large database does not turn metadata into clinical proof. The same trial can appear as a protocol, registry entry, primary report, follow-up and secondary analysis. Medical Cooling therefore separates six statement types.
Biology is plausible
A measurable physical or biological process occurs. It can justify the next experiment, but not a patient-benefit claim.
The procedure can be delivered
A method can be performed under the tested conditions. Feasibility is not efficacy.
Limited data define a risk picture
The absence of a problem in a small cohort is not a general safety proof.
Patient-relevant outcomes improve
Requires a design capable of supporting the specific claim in the specific population.
The evidence does not carry the claim
The question may be untested, underpowered or contradicted by the available data.
The prespecified difference was not shown
Not automatically equivalence, but it narrows the claims that remain defensible.
Thermal dose
Temperature is not a number. It is a controlled trajectory.
The research synthesis treats therapeutic temperature as a multidimensional dose. Two trials that both say “33 °C” can still deliver very different interventions.
Clinical evidence map
Where hypothermia stands by indication.
| Field | Status | What the evidence supports | What it does not support |
|---|---|---|---|
| AIRCHILL | Preclinical / exploratory | Local temperature change; early technical feasibility under defined experimental conditions. | Patient safety, sufficient organ or whole-body cooling, clinical benefit. |
| Cardiac arrest | Temperature control established | Protocolised temperature management and fever control; current target choice remains context-dependent. | Universal superiority of 33 °C; benefit of large cold-fluid boluses; benefit of simply extending 33 °C duration. |
| Transnasal intra-arrest cooling | Efficacy open | Earlier temperature reduction; PRINCESS2 pilot supports protocol feasibility. | Established clinical benefit or transfer to AIRCHILL. |
| Neonatal HIE | Established in a narrow indication | Protocolised whole-body hypothermia in carefully selected term / near-term neonates with moderate-to-severe HIE in the early treatment window. | Routine extension to mild HIE, younger preterm infants or late treatment without new evidence. |
| Ischaemic stroke | Experimental | Feasibility; heterogeneous signals; selective approaches are under active study. CHILL-ART currently remains a conference-level result in this corpus. | Established standard, proven mortality reduction or equivalence of local and systemic methods. |
| Severe TBI | No general benefit established | Hypothermia can lower intracranial pressure in some settings. | Improved functional outcome from general prophylactic or systemic hypothermia. |
| Spinal cord injury | Experimental | Preclinical plausibility and small feasibility signals. | Established clinical efficacy. |
| Accidental hypothermia | Rewarming / resuscitation medicine | Continued resuscitation in appropriate cases and extracorporeal support / rewarming in selected patients. | One universal randomized effect estimate for every rewarming strategy. |
| Perioperative hypothermia | Prevention established | Active warming and prewarming can reduce perioperative hypothermia and shivering. | Transfer of warming evidence to therapeutic neuroprotection. |
| Sepsis / ARDS | Open | Controlled programmes are testing fever control and hypothermia. | Clinical benefit before results are available. |
Trial register
What the next generation of studies is testing.
Status below is a frozen research snapshot. A registry entry describes planning and status, not efficacy. Completion dates can become stale and publications can appear before registry result tables are updated.
| Study | Field | Status | n | Planned / reported completion | What it addresses | Gewonnenes Wissen |
|---|---|---|---|---|---|---|
| ICECAP · NCT04217551Duration of Therapeutic Hypothermia After OHCA | Cardiac arrest | Completed | 1,158 actual | 2026-02-01 | Adaptive comparison of durations at 33 °C. The 2026 result did not show benefit from longer cooling; there was no normothermia arm. | Duration is a separate causal question from temperature target. More hours at 33 °C did not rescue efficacy; AIRCHILL must demonstrate an early thermal separation first and should not assume that a longer exposure compensates for a weak or late dose. |
| P-ICECAP · NCT05376267 | Cardiac arrest | Recruiting | 900 estimated | 2028-03-31 | Cooling duration after paediatric cardiac arrest. | Paediatric arrest is not an adult surrogate. A separate population, physiology, consent pathway and outcome model are required; AIRCHILL should not transfer adult assumptions into paediatric development. |
| PRINCESS2 · NCT06025123 | Cardiac arrest | Recruiting | 1,022 estimated | 2028-05 | On-scene transnasal cooling in initially shockable OHCA; pilot addressed adherence and device-related safety, not efficacy. | The programme learned from PRINCESS by narrowing to an initial shockable rhythm, separating a 100-patient feasibility/adherence pilot from efficacy, and powering the main trial for a prespecified 9-point absolute difference. For AIRCHILL, phenotype, timing, adherence and safety should be proven before a pivotal outcome claim. |
| NCT06776549 | Cardiac arrest | Recruiting | 468 estimated | 2030-01-31 | Hypothermia versus normothermia after ECPR for OHCA. | ECPR should be treated as a distinct study stratum because its workflow and timing differ materially from conventional resuscitation. |
| OverCool · NCT06798818 | Cardiac arrest | Not yet recruiting in snapshot | 24 estimated | 2026-03 | Ultra-rapid cooling induction using total liquid ventilation; small feasibility programme. | A very small study can be appropriate when the question is technical feasibility rather than clinical benefit. AIRCHILL should use early cohorts to establish deliverability and safety, not to infer neurological efficacy from a handful of patients. |
| NCT07086703 | Cardiac arrest | Not yet recruiting | 788 estimated | 2029-05-01 | 33 °C versus controlled normothermia after in-hospital cardiac arrest. | In-hospital arrest differs from OHCA in delay, monitoring and cause. AIRCHILL should avoid combining fundamentally different care pathways unless interaction by setting is prespecified and adequately powered. |
| PRECEDENCE · NCT07421882 | Cardiac arrest | Recruiting | 50 estimated | 2027-03 | Small prehospital cooling-vest feasibility study. | Prehospital feasibility should quantify set-up time, treatment interruptions and achieved thermal separation. A device that is easy to deploy but produces little early temperature difference cannot answer the neuroprotection hypothesis. |
| CoolPrime · NCT04621279 | Neonatal HIE | Recruiting | 460 estimated | 2029-01-01 | Cooling versus normothermia for mild HIE. | Evidence from moderate/severe neonatal HIE does not automatically extend to mild disease. AIRCHILL should not broaden an indication merely because a related population benefits from cooling; each severity phenotype needs its own benefit–risk test. |
| COMET · NCT05889507 | Neonatal HIE | Recruiting | 426 estimated | 2030-01-01 | Whole-body hypothermia in mild neonatal encephalopathy with long-term follow-up. | Long-term outcomes require long-term follow-up. The timing of AIRCHILL endpoints should match the claimed patient benefit rather than rely only on early hospital outcomes. |
| COTTIS-2 · NCT06301412 / DRKS00031086 | Stroke | Recruiting in snapshot | 400 estimated | 2026-06-18 | Targeted temperature management integrated with thrombectomy; completion date is already past and therefore needs live re-verification before reuse. | Live status and final result sources should be re-checked before a study is used to set AIRCHILL design assumptions. |
| CHILL-ART · NCT06758609 | Stroke | Registry / publication conflict | 262 estimated | 2026-06-30 | Registry still showed recruiting while a 2026 conference abstract reported randomized results; full publication remains the higher-priority next source. | Use the final peer-reviewed report as the preferred result source when status metadata and publications conflict. |
| COOLHEAD-2b · NCT07526649 | Stroke | Not yet recruiting | 182 estimated | 2029-06-30 | Non-invasive convective head cooling during endovascular thrombectomy. | A device study should show that the intended local thermal effect is achieved during the clinically relevant procedural window before asking a patient-outcome question. |
| NCT04494074 | Sepsis | Recruiting | 820 estimated | 2026-12-01 | External fever control in mechanically ventilated septic-shock patients. | Temperature manipulation can have indication-specific effects. AIRCHILL should not transfer efficacy assumptions from post-arrest neuroprotection into inflammatory critical illness. |
| NCT04545424 | ARDS | Recruiting | 340 estimated | 2026-10-31 | Therapeutic hypothermia in ARDS. | Pulmonary disease changes the safety question. For AIRCHILL, respiratory indications require separate gas-exchange and lung-safety evidence rather than borrowing from neurological indications. |
| NCT02991690 | Spinal cord | Recruiting | 120 estimated | 2026-09 | Systemic hypothermia after acute cervical spinal cord injury. | Mechanistic plausibility in another nervous-system injury does not establish transferability. Each AIRCHILL indication needs its own population, timing window and patient-relevant endpoint. |
| NCT06200285 | Accidental hypothermia | Recruiting registry | 500 estimated | 2054-12 | International observational registry; not a randomized intervention. | Observational registries are valuable for incidence, safety patterns and prognosis, but they cannot substitute for randomized evidence when estimating an AIRCHILL treatment effect. |
Register snapshot: ClinicalTrials.gov API v2 data state 7 August 2026, research extraction 10 August 2026. Status is intentionally shown with conflicts instead of silently overwriting them.
Publication lessons
Why earlier cooling studies were positive, neutral or inconclusive.
A neutral trial is not automatically “too small”. The decisive question is whether the study created the biological contrast it intended, in the right population, early enough, with an endpoint and sample size capable of detecting a realistic effect. Post-hoc and subgroup findings are labelled as such and are not treated as confirmatory evidence.
| Publication / study | N | What was tested | Result / why interpretation is difficult | Gewonnenes Wissen |
|---|---|---|---|---|
| PRINCE · 2010Circulation | ~200 randomized | Intra-arrest transnasal evaporative cooling during CPR. | Cooling was clearly accelerated, but the pilot was not powered for patient-relevant clinical outcomes. | A small randomized device study can establish feasibility and thermal separation; it should not be used to estimate a stable neurological effect size for a pivotal trial. |
| PRINCESS · 2019JAMA | 677 randomized; 671 completed | Transnasal intra-arrest cooling versus standard care in a broad witnessed OHCA population. | Core temperature <34 °C was reached earlier (median 105 vs 182 min), but CPC 1–2 at 90 days was 16.6% vs 13.5% (absolute difference 3.1 points; 95% CI −2.3 to 8.5). The primary endpoint was neutral. | The study was not simply “negative because too small”: the observed overall effect was much smaller than a large pivotal-effect assumption, and the population mixed prognostically different rhythms. AIRCHILL should prespecify phenotype and power to a conservative effect, not to a post-hoc signal. |
| TTM · 2013NEJM | 939 primary analysis | 33 °C versus actively controlled 36 °C after OHCA. | No mortality or neurological advantage of 33 °C. Both groups received protocolised temperature control, so the comparison was not cooling versus uncontrolled fever. | Comparator quality matters. If both arms receive effective temperature management, the incremental effect can be small. AIRCHILL must define what physiological contrast it adds on top of modern standard care. |
| Kim et al. · 2014JAMA | 1,359 randomized | Up to 2 L of 4 °C saline immediately after ROSC versus standard care. | Hospital temperature was lowered and target was reached about an hour earlier, but survival and neurological status did not improve; the fluid strategy also added a treatment burden distinct from selective/device cooling. | Earlier temperature reduction is not sufficient evidence of benefit if the delivery method introduces competing physiological effects. AIRCHILL should separate the effect of thermal dose from fluid, pressure, ventilation or other co-interventions. |
| HYPERION · 2019NEJM | 584 randomized; 581 analysed | 33 °C for 24 h versus targeted 37 °C after nonshockable cardiac arrest. | Favourable CPC at day 90 was 10.2% vs 5.7% (difference 4.5 points; 95% CI 0.1–8.9), with no mortality difference. The signal was statistically fragile because event counts were low and the confidence interval was wide. | Low baseline event rates make binary neurological endpoints sample-hungry and unstable. AIRCHILL should obtain contemporary control-event rates before locking sample size and should avoid treating one borderline positive trial as a transferable effect estimate. |
| TTM2 · 2021NEJM | 1,900 enrolled; 1,861 ITT | 33 °C versus normothermia with early fever treatment after OHCA. | No reduction in 6-month mortality or poor functional outcome; haemodynamically important arrhythmia was more frequent with hypothermia. Median arrest-to-randomisation was about 135 min, so this trial chiefly answers later post-ROSC systemic cooling, not an intervention that produces target-organ cooling during the first minutes. | A large, well-conducted neutral trial can rule against the tested strategy without answering a materially different timing/route hypothesis. AIRCHILL must prove that it actually creates an earlier and different thermal exposure rather than cite TTM2 either for or against that untested exposure. |
| PRINCE + PRINCESS pooled analysis · 2021Critical Care | 851 as-treated; 325 initial shockable rhythm | Post-hoc pooled individual-patient analysis by initial rhythm. | Among shockable-rhythm patients, favourable outcome at discharge was 34.2% vs 24.0% (RR 1.43, 95% CI 1.01–2.02). This was post-hoc, pooled and as-treated rather than the primary randomized endpoint of either trial. | This is hypothesis-generating phenotype information, not confirmatory efficacy. It can justify prospective enrichment such as PRINCESS2, but AIRCHILL must not use the observed 10-point-plus subgroup difference as its assumed treatment effect without independent confirmation. |
| PRINCESS2 design + pilotNCT06025123 | 1,022 planned; first 100 prespecified pilot | On-scene transnasal cooling in initial shockable OHCA; mRS 0–1 at 90 days. | The programme narrowed the phenotype, created a feasibility/adherence pilot, prespecified DSMC safety/futility review and powered for an absolute 9-point difference (45% to 54%). Efficacy remains unproven while the trial is ongoing. | The strongest transferable lesson is programme architecture: qualify sites and workflow first, measure protocol adherence and real treatment timing, then run the efficacy test. The 9-point assumption belongs to PRINCESS2 and is not an AIRCHILL effect estimate. |
Double-counting protection
One study can have many records.
Primary RCT + school-age follow-up
The follow-up is clinically important, but it is not a second randomized population.
Separate trials, linked secondary analyses
A TTM2 shockable-rhythm secondary analysis is not counted as a new independent RCT.
Development family
PRINCESS2 pilot participants flow into the main programme; pilot and main trial must not be double-counted as independent efficacy datasets.
Registry + 2026 article
Two records, one adaptive randomized study.
Registry + result article
Same trial family; the adverse functional result remains visible.
Pilot + COTTIS-2
Different development phases should be distinguished but described as one translational programme.
Registry + conference result
One RCT; conference data remain preliminary until a full publication is appraised.
Report + poster + peer-reviewed subset
Overlapping samples are reported transparently and not simply added together.
Safety and hidden treatment dose
The cooling device is only part of the intervention.
Arrhythmia, circulation, coagulation, infection, electrolytes, drug handling, shivering, sedation and rewarming can all move the benefit–risk balance. Selective systems add route-specific risks.
Rhythm and haemodynamics
Deeper cooling can increase bradycardia, arrhythmia burden and the need for haemodynamic support.
Bleeding must be a prespecified safety endpoint
Especially relevant in trauma, intracranial bleeding, thrombectomy and anticoagulated patients.
Fever suppression can hide a signal
Cooling can alter immune response; pneumonia recurs as a concern in several evidence families.
Electrolytes, glucose and drug clearance change
Rewarming can reverse shifts that developed during cooling, making the end phase part of the dose.
Shivering changes the protocol
Analgesia, sedation and neuromuscular blockade can affect blood pressure, neurological examination and ventilation.
Airway and gas-path safety
Mucosa, secretion handling, condensation, resistance, pressure, gas exchange and lung tissue need dedicated testing.
The 33,808-record corpus
The large table is a screening instrument, not a scoreboard.
The uploaded research register contains journal records, reviews, case reports, randomized trials, registry records, observational studies and guidelines. It is intentionally broader than the set that can support website claims. The on-site evidence map and trial table therefore expose the manually appraised decision points rather than forcing a visitor to load more than 20 MB of raw metadata.
The full raw research register remains a versioned research appendix. Because it contains 33,808 metadata rows, possible false positives and repeated study families, it is deliberately not injected into every public page load. The public table above is the claim-relevant navigation layer.
What this means for AIRCHILL.
The defensible hypothesis is narrow: a very early, sufficiently dosed and safely controllable intervention through an already required airway could change target-organ temperature with less additional treatment burden than some systemic approaches. AIRCHILL has not demonstrated clinical benefit, patient safety in emergency patients or sufficient cooling performance of the future product. The research map defines the tests required to move those statements.
Open the clinical study plannerOpen global impact modelsSources & assumptions