Nine RCTs were included in the review (788 participants, range 15 to 273). The quality of included studies was low: more than three-quarters did not describe the method of randomisation and all studies used an open-label design.
Renal replacement therapy was associated with a statistically insignificant lower incidence of CI-AKI than control (RR 0.74 95% CI 0.35 to 1.60; eight RCTs, significant heterogeneity Ι²=84%). Renal replacement therapy was associated with significantly lower in-hospital mortality than control (RR 0.33, 95% CI 0.14 to 0.77; four RCTs, no significant heterogeneity).
Subgroup analyses found that continuous veno-venous haemofiltration was associated with a significant decrease in CI-AKI compared with control (10.8% versus 51.1%, p=0.006). In patients with chronic kidney disease stage 3, the control group was associated with significantly decreased CI-AKI compared with the haemodialysis group (RR 1.53, 95% CI 1.10 to 2.12; five RCTs). In patients with a baseline chronic kidney disease over 3, renal replacement therapy was associated with decreased CI-AKI compared with control (RR 0.19, 95% CI 0.08 to 0.43; three RCTs). These analyses were not associated with significant heterogeneity.
There was no significant difference in the incidence of permanent haemodialysis between haemodialysis and control groups.
Meta-regression analysis suggested that heterogeneity was explained primarily by baseline chronic kidney disease stage (co-efficient -1.67, 95% CI 2.32 to -1.02).
There was no evidence of significant publication bias.