Fig. 2.
Forest plots for meta-analyses of the adjusted estimates for the association between vasectomy and prostate cancer by study design. Data were pooled separately by study design. As significant heterogeneity (I2 > 50%) was found, a pooled estimate was calculated with a random-effect model (DerSimonian and Laird method). CI = confidence interval; df = degrees of freedom; IV = inverse variance; SE = standard error.