Fig. 2From: Meta-analysis of targeted temperature management in animal models of cardiac arrestFunnel plot and Trim and Fill analysis of neurobehavioural outcome. Open circles are missing studies according to Trim and Fill-analysis. Estimate of efficacy with missing studies included: 0.76 SMD [95% CI 0.59 to 0.93]. Efficacy without missing studies: 0.93 SMD [95% CI 0.79 to 1.08]. Standardized mean difference = SMD. Standard error = SE. Confidence interval = CIBack to article page