Does treatment with essential amino acids and hypertonic glucose improve survival in acute renal failure?: A meta-analysis

Abstract

We performed a meta-analysis of four randomized trials (RCTs) and one concurrent control study which addressed the following question: Does treatment with amino acid solutions and hypertonic glucose improve survival of seriously ill patients with acute renal failure who are unable to take oral or enteral feedings? Three RCTs compared outcomes with essential amino acid solutions (EAA) and hypertonic glucose against hypertonic glucose alone, and their results were pooled. Use of EAA was associated with an absolute increase in initial survival of 0.24 as compared to glucose (confidence interval: +.015 to +.446). Without weighting the data to reflect the quality of the RCTs, this effect was significant (p = .017). Patients needing dialysis showed a significant treatment effect (p = .015), while those not requiring dialysis did not (p = .11). However, survival to hospital discharge was not significantly improved for the patient population as a whole (p = .10). Using a standardized quality assessment protocol, the four RCTs received scores ranging from .188 to .357 out of a possible 1.00. Any statistically significant treatment effects were abolished by factoring the quality scores of the studies into the data-pooling process. Sensitivity analysis was performed to determine the effect of including the nonrandomized study. Although it had more patients than the other studies combined, the nonrandomized study was of low quality (score: .032) and its inclusion did not change the outcome of data pooling once quality weighting was applied. We conclude that the efficacy of these parenteral nutritional regimens remains uncertain in this clinical setting.