A Cochrane review included 22 RCTs with a total of 1 589 patients. The patients with any closed traumatic head injury were randomised to be cooled to a maximum of 35ºC for at least 12 hours or to the control group. There were fewer deaths in hypothermia patients, but the result was not statistically significant (OR 0.83, 95% CI 0.66 to 1.04; 20 trials, 1 562 patients). In 8 trials with good allocation concealment there was no decrease in the risk of death with hypothermia (OR 1.08, 95% CI 0.79 to 1.47). Hypothermia patients were less likely to have an unfavourable outcome (death, vegetative state or severe disability) (OR 0.75, 95% CI 0.61 to 0.93; 20 trials, 1 562 patients). In 8 trials with good allocation concealment the hypothermia patients were non-significantly less likely to have an unfavourable outcome (OR 0.90, 95% CI 0.68 to 1.20). Hypothermia was associated with a slight, non-significant increase in the risk of pneumonia (OR 1.33, 95% CI 0.94 to 1.89) but in trials with good allocation concealment there was a non-significant reduction in in the risk of pneumonia (OR 0.81, 95% CI 0.50 to 1.31; 3 trials, 281 patients).
Comment: The quality of evidence is downgraded by inconsistency (heterogeneity in interventions and study populations) and study quality (inadequate allocation concealment).