Fig. 3

a. Effects of intravenous hydration lowering on risk of contrast-induced nephropathy (Hydration vs. No hydration, using random effects model) The event rate in different study arms is presented alongside the computed risk ratio (95% confidence interval [CI] (lower and upper limit) with p value. Forest plot shows effect size (solid squares) with 95% CI (black line through the solid squares), in terms of risk ratio for individual studies and pooled risk ratio (open diamonds) for random effects model at the bottom. Studies favouring reduction of risk with isotonic hydration are on the left of the centre line, and studies favouring control arm are on right of the centre line. b. Effects of intravenous saline hydration lowering on risk of contrast-induced nephropathy (Saline vs. No hydration) The event rate in different study arms is presented alongside the computed risk ratio (95% confidence interval [CI] (lower and upper limit) with p value. Forest plot shows effect size (solid squares) with 95% CI (black line through the solid squares), in terms of risk ratio for individual studies and pooled risk ratio (open diamonds) for random effects model at the bottom. Studies favouring reduction of risk with normal saline hydration are on the left of the centre line, and studies favouring control arm are on right of the centre line