Nineteen RCTs with 2,218 participants were included in the review. Sample sizes ranged from 23 to 291 participants. One trial was good quality,11 were fair quality, and seven were poor quality. Only 12 trials described an adequate randomisation method. Only seven trials reported the number of drop-outs and described use of intention-to-treat analysis.
Event rates were low; some studies had no events in either group. There was no significant difference between therapeutic hypothermia and normothermia in rates of acute kidney injury (OR 1.01, 95% CI 0.68 to 1.51; 12 RCTs; 1,839 participants; Ι² =0%), serum creatinine (five RCTs; 522 participants), creatinine clearance (three RCTs; 141 participants), or need for dialysis (three RCTs; 509 participants). Therapeutic hypothermia was associated with a significant reduction in the mortality rate (OR 0.69, 95% CI 0.51 to 0.92; 16 RCTs; 2,077 participants, Ι²=0%).
In meta-regression for the outcome of acute kidney injury, lower target cooling temperature was associated with lower odds of injury (p=0.01). However, the findings were no longer significant if a trial that delivered infra-renal arterial cooling of 15°C was excluded. Duration of therapeutic hypothermia or duration of cardiac arrest did not have a significant effect. The results of other sensitivity and subgroup analyses were also reported.
There was some asymmetry in the funnel plot for the primary outcome.