Cooling heatstroke patients to 35°C showed more complications and no clear gain in a pilot

Design

Multicentre pilot randomised trial of two temperature management strategies

Population

35 intubated adults with heatstroke at 10 tertiary ICUs in China

Primary outcome

Feasibility (recruitment, retention, adherence), with 28-day function and survival as exploratory outcomes

Effect

Poor function 17.6% vs 16.7%; survival 82.4% vs 83.3%; severe gastrointestinal injury 52.9% vs 0%; arrhythmia 41.2% vs 11.1%

This pilot trial, published on 6 August 2026, randomised 35 intubated adults with heatstroke at ten Chinese intensive care units. One group had protocolised rapid cooling to 35.0°C within two hours, held for 48 hours; the other had cooling only if temperature exceeded 38.0°C, aiming for 36.5 to 37.5°C. The aim was to test feasibility: recruitment was 83.7%, retention 97.2%, and protocol adherence above 88%.

There was no significant difference in poor functional outcome at 28 days (17.6% vs 16.7%) or survival (82.4% vs 83.3%). Severe gastrointestinal injury (grade 3 to 4) was far more frequent with cooling to 35°C (52.9% vs 0%, p < 0.001), as were arrhythmias (41.2% vs 11.1%, p = 0.045). A rescue sedative cocktail for shivering was needed in all patients in the hypothermia group against 27.8% in the other group.

With 35 patients the trial cannot rule out benefit or harm on survival, but the authors advise against routine adoption. Prompt cooling to a normal temperature remains the principle; this trial concerns going lower.

• Cooling to 35°C for 48 hours did not change survival (82.4% vs 83.3%) or function at 28 days.
• Severe gastrointestinal injury occurred in 52.9% vs 0%, and arrhythmias in 41.2% vs 11.1%.
• All patients cooled to 35°C needed drug treatment for shivering.
• Aim for prompt cooling to a normal temperature and avoid routine cooling below it.
• The pilot enrolled only 35 patients, so larger trials may refine this.

Why it matters

The pilot found harm signals from deeper cooling with no offsetting benefit, which is relevant wherever a target below normal temperature is considered.

Don't overread it

A 35-patient feasibility trial was not powered for survival, and its findings are preliminary.

The statistics, in plain English

With 35 patients, a survival difference of one percentage point is meaningless either way. The complication differences are large enough to reach statistical significance even in a small sample, which is why the safety signals are the part of the result to take seriously.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.