Balanced crystalloids versus saline and mortality in hospitalized patients: a hierarchical Bayesian meta-analysis of cluster-randomized trials.

BACKGROUND

Balanced crystalloids are widely used as alternatives to 0.9% saline, yet their effects on patient-important outcomes remain uncertain. Cluster-randomized trials provide a pragmatic framework for evaluating fluid strategies as implemented in routine clinical practice. We performed a hierarchical Bayesian meta-analysis to estimate the effectiveness of balanced versus saline fluid policies.

METHODS

We searched MEDLINE, Embase, and CENTRAL through September 2025 for cluster-randomized and cluster-allocated trials comparing balanced crystalloids with saline across perioperative, emergency, and critical care settings. The primary outcome was all-cause mortality; acute kidney injury (AKI) was secondary. Hierarchical Bayesian random-effects models were fitted using published study-level estimates that accounted for clustering whenever available.

RESULTS

Six trials involving 42,223 patients met the inclusion criteria. For mortality, the pooled Bayesian relative effect was 0.92 (95% CrI 0.77-1.18), corresponding to an 83% posterior probability of benefit. For AKI, the pooled Bayesian relative effect was 1.02 (95% CrI 0.86-1.23), corresponding to a 40% posterior probability of benefit. Sensitivity analyses yielded similar findings.

CONCLUSIONS

Balanced crystalloid policies were not associated with a clear reduction in mortality or AKI in cluster-randomized evaluations of routine care. Any true effect is likely modest and dependent on exposure and clinical context. The available evidence remains compatible with modest benefit or no meaningful effect and complements patient-level evidence by informing the real-world effectiveness of fluid policies under routine implementation conditions.

PROSPERO ID: CRD420251146439.