Hyperoncotic colloids and acute kidney injury: a meta-analysis of randomized trials

INTRODUCTION It has been hypothesized that hyperoncotic colloids might contribute to acute kidney injury (AKI). However, the validity of this hypothesis remains unclear. METHODS A meta-analysis was conducted of randomized controlled trials evaluating AKI after infusion of hyperoncotic albumin and hydroxyethyl starch (HES) solutions. Mortality was a secondary endpoint. Eligible trials were sought by multiple methods, and the pooled odds ratios (OR) for AKI and death and 95% confidence intervals (CI) were computed under a random effects model. RESULTS Eleven randomized trials with a total of 1220 patients were included: 7 evaluating hyperoncotic albumin and 4 hyperoncotic HES. Clinical indications were ascites, surgery, sepsis and spontaneous bacterial peritonitis. Hyperoncotic albumin decreased the odds of AKI by 76% (OR, 0.24; CI, 0.12-0.48; P < 0.0001), while hyperoncotic HES increased those odds by 92% (OR, 1.92; CI, 1.31-2.81; P = 0.0008). Parallel effects on mortality were observed, with hyperoncotic albumin reducing the odds of death by 48% (OR, 0.52; CI, 0.28-0.95; P = 0.035) and hyperoncotic HES raising those odds by 41% (OR, 1.41; CI, 1.01-1.96; P = 0.043). CONCLUSIONS This meta-analysis does not support the hypothesis that hyperoncotic colloid solutions per se injure the kidney. Renal effects appear instead to be colloid-specific, with albumin displaying renoprotection and HES showing nephrotoxicity.