Comparative efficacy of balanced crystalloids versus 0.9% saline on delayed graft function and perioperative outcomes in kidney transplantation: a meta-analysis of randomised controlled trials

Tzu Chang,Ming-Chieh Shih,Yi-Luen Wu,Tsung-Ta Wu,Jen-Ting Yang,Chun-Yu Wu
DOI: https://doi.org/10.1016/j.bja.2024.08.008
Abstract:Background: Delayed graft function after kidney transplantation is linked to poor graft survival and increased chronic allograft injury. Recent guidelines suggest using balanced crystalloids over 0.9% saline owing to better metabolic profiles, but their impact on DGF remains unclear. Methods: We searched PubMed, Embase, and Cochrane Central Registry of Clinical Trials from inception until February 29, 2024, and included RCTs that randomised adult participants to receive either intravenous balanced fluids or 0.9% saline intraoperatively. We pooled data using a random-effects model and present risk ratios (RRs) or mean differences, with 95% confidence intervals (CIs). We assessed individual study risk of bias using the modified Cochrane tool and certainty of evidence using GRADE. Outcomes analysed were delayed graft function incidence, vasopressor requirements, length of hospital stay, and postoperative metabolic profiles. Results: Of 106 publications identified, we included 11 RCTs (n=1717). Pooled analysis showed that the use of balanced fluids was associated with a lower incidence of delayed graft function compared with 0.9% saline (RR 0.82, 95% CI: 0.69 to 0.98, P=0.01, moderate certainty). Balanced crystalloids were associated with higher postoperative serum pH, higher serum bicarbonate, and lower serum chloride concentration, but effects on vasopressor requirements, length of hospital stay, and serum creatinine were uncertain. Conclusions: Balanced crystalloid intravenous fluid therapy reduced delayed graft function incidence and maintained more favourable serum chemistry profiles compared with 0.9% saline in patients undergoing kidney transplantation. However, crystalloid type did not significantly influence vasopressor requirements and length of hospital stay.
What problem does this paper attempt to address?