Balanced crystalloids versus 0.9% saline for fluid resuscitation in children with septic shock: an updated systematic review and meta-analysis of randomized controlled trials

Why clinicians should know about it

• Picked for Emergency Medicine (top studies of the week, 27 September 2026).

Abstract

BACKGROUND: 0.9% (normal) saline remains the most widely used bolus resuscitation fluid in paediatric septic shock, despite a supraphysiological chloride content that may impair renal perfusion and cause hyperchloraemic acidosis. Whether balanced crystalloids, which more closely approximate plasma, improve clinical outcomes in children remains uncertain. METHODS: Following PRISMA 2020 guidelines, PubMed/MEDLINE, Scopus, ScienceDirect and Web of Science were searched upto July 2026 for randomized trials of balanced or buffered crystalloids versus 0.9% saline in children (< 18 years) with sepsis or septic shock. Random-effects meta-analyses used RevMan 5.4. RESULTS: Six trials (10,308 children) were included. Balanced crystalloids did not reduce in-hospital or ICU mortality (3.7% vs. 3.9%; RR 0.95, 95% CI 0.75-1.21), at 30 days (RR 1.08, 0.71-1.65) or at 90 days (RR 1.10, 0.83-1.45). Acute kidney injury was not reduced (RR 0.76, 0.46-1.27), but renal replacement therapy was (RR 0.67, 0.48-0.92). Hyperchloraemia (RR 0.65, 0.61-0.70), 6-hour chloride rise (MD - 3.99 mEq/L, - 5.57 to - 2.42) and hypernatraemia (RR 0.59, 0.42-0.83) were reduced, whereas hyperlactataemia was more frequent (RR 1.17, 1.00-1.38). No other safety, organ-support or resource-use outcome differed. CONCLUSION: Balanced crystalloids reliably reduce hyperchloraemia, chloride loading, and hypernatraemia and are associated with less renal replacement therapy but confer no advantage in mortality or length of stay; the evidence now excludes a clinically important mortality benefit. They remain a physiologically preferable and equally safe default where available, but fluid choice should not displace early recognition, timely antimicrobials and haemodynamic support. PROSPERO REGISTRATION: CRD420261468983.

Abstract as published, via PubMed.