Objective: To review whether induced hypothermia after traumatic brain injury affects morbidity and mortality based on the results of two meta-analyses.
Design: Critical appraisals of McIntyre et al: Prolonged therapeutic hypothermia after traumatic brain injury in adults: A systematic review. JAMA 2003; 289:2992-2999, and Henderson et al: Hypothermia in the management of traumatic brain injury: A systematic review and meta-analysis. Intensive Care Med 2003; 29:1637-1644.
Findings: Both meta-analyses included trials of adult patients with severe traumatic brain injury randomized to induced hypothermia or normothermia and evaluated risk of death and poor neurologic outcomes. McIntyre et al. found the overall relative risk of mortality with induced hypothermia to be 0.81 (95% confidence interval 0.69-0.96). By designing a priori analyses, these authors also found that the relative risk of death was reduced in patients cooled for >48 hrs, and the risk of poor neurologic outcome was reduced with all durations of cooling, cooling to 32-33°C, and rewarming in <24 hrs. In contrast, Henderson et al. found that induced hypothermia did not change the odds of death after traumatic brain injury (odds ratio 0.81; 95% confidence interval 0.59-1.13) and that normothermic controls had an odds ratio of 0.42 (95% confidence interval 0.25-0.70) for developing intercurrent pneumonia. Both analyses found trials to be heterogeneous with respect to neurologic outcome.
Conclusions: The discrepancies in the results of these contemporaneous meta-analyses may stem, in part, from differences in their trial selection strategies as well as from sources of trial heterogeneity. Nevertheless, McIntyre et al. uncovered the equivalent of a dose-dependent reduction in the risk of death with induced hypothermia, supporting further study of this neuroprotective strategy. Although these meta-analyses included trials containing adult patients, a phase II trial of induced hypothermia in pediatric traumatic brain injury has established its feasibility and safety in infants and children. As in adult patients, induced hypothermia for traumatic brain injury in children can be considered an optional therapy for refractory intracranial hypertension but should not be regarded as standard of care.