Meta-analysis of the efficacy and safety of therapeutic hypothermia in children with acute traumatic brain injury.

Objective: To evaluate the efficacy and safety of therapeutic hypothermia in children with acute traumatic brain injury (TBI).

Methods: A systematic literature review using PubMed, Embase, Cochrane Library, Chinese National Knowledge Infrastructure, Wanfang, VIP, and Chinese Biomedical Database was performed to retrieve studies of randomized controlled trials (RCTs) on therapeutic hypothermia for children with TBI published before March 2014. Data extraction and quality evaluation of RCTs were performed by 2 investigators independently. A meta-analysis was performed by RevMan 5.2.7.

Results: There were 7 RCTs comprising 442 children (218 in hypothermia group and 224 in normothermia group). Meta-analysis showed therapeutic hypothermia could increase mortality compared with the normothermia group (relative risk [RR] = 1.84, 95% confidence interval [CI] = 1.15-2.93, P = 0.01). On the Glasgow Outcome Scale (GOS), the following scores did not differ between the hypothermia group and normothermia group: 3-month GOS 4-5 (RR = 0.89, 95% CI = 0.68-1.16, P = 0.39), 3-month GOS 1-3 (RR = 1.19, 95% CI = 0.80-1.76, P = 0.39), 6-month GOS 4-5 (RR = 0.91, 95% CI = 0.78-1.07, P = 0.26), and 6-month GOS 1-3 (RR = 1.18, 95% CI = 0.88-1.59, P = 0.27). Hypothermia did not increase the rate of pneumonia (RR = 0.84, 95% CI = 0.63-1.12, P = 0.23) or bleeding (RR = 0.94, 95% CI = 0.39-2.26, P = 0.89), but the incidence of arrhythmias was higher in the hypothermia group (RR = 2.60, 95% CI = 1.06-6.41, P = 0.04).

Conclusions: No benefit of therapeutic hypothermia in children with TBI is shown in this study; therapeutic hypothermia may increase the risk of mortality and arrhythmia. There is no evidence that therapeutic hypothermia improves prognosis of children with TBI; there is also no evidence that therapeutic hypothermia increases the risk of pneumonia and coagulation dysfunction. These results are limited by the quality of the included studies and need to be considered with caution. Further large-scale, well-designed RCTs on this topic are needed.