Effect and Feasibility of Therapeutic Hypothermia in Patients with Hemorrhagic Stroke: A Systematic Review and Meta-Analysis.

Abstract

Therapeutic hypothermia (TH) has shown good results in experimental models of hemorrhagic stroke. The clinical application of TH, however, remains controversial, since reports regarding its therapeutic effect are inconsistent. We conducted a systematic review based on Preferred Reporting Items for Systematic Reviews and Meta-analyses comparing TH with a control group in terms of mortality, poor outcome, delayed cerebral ischemia (DCI), and specific complications. The subgroup analyses were stratified by study type, country, mean age, hemorrhage type, cooling method, treatment duration, rewarming velocity, and follow-up time. Nine studies were included, most of which were of moderate quality. The overall effect demonstrated insignificant differences in mortality (risk ratio [RR] 0.78; 95% confidence interval [CI] 0.58-1.06; P = 0.11) and poor outcome rate (RR 0.89; 95% CI 0.70-1.12; P = 0.32) between TH and the control group. However, sensitivity analyses, after we omitted 1 study, achieved a statistically significant difference in poor outcome favoring TH. Moreover, in the subgroup analyses, the results derived from randomized studies revealed that TH significantly reduced poor outcomes (RR 0.40; 95% CI 0.22-0.74; P = 0.003). In addition, TH significantly reduced DCI compared with control (RR 0.61; 95% CI 0.40-0.93; P = 0.02). The incidence of specific complications (rebleeding, pneumonia, sepsis, arrhythmia, and hydrocephalus) between the 2 groups were comparable and did not reach significant difference. The overall effect showed TH did not significantly reduce mortality and poor outcomes but led to a decreased incidence of DCI. Compared with control, TH resulted in comparable incidences of specific complications.