Fig. 2. Pooled data evaluating the effect of dexamethasone dose on early pain scores (4 h or less) at rest compared with placebo. Data evaluated using a random effects model. Point estimate (95% CI) for overall effect was −0.32 (−0.46 to −0.17). Standardized mean difference for individual study represented by small circles on Forrest plot with 95% CI of the difference shown as solid line. Larger sized circle and thicker 95% CI line denote larger sample size. The diamond represents the pooled estimate and uncertainty for the effects of low- (0.1 mg/kg or less), intermediate- (0.11–0.2 mg/kg), and high-dose (more than 0.2 mg/kg) dexamethasone, respectively. Sample heterogeneity as assessed by the I 2for the low-, intermediate-, and high-dose grouping of studies was 68, 52, and 42, respectively.