Timing of kidney replacement therapy in critically ill adults: an umbrella review of systematic reviews and meta-analyses

Abstract

Abstract Background In critically ill adults with acute kidney injury (AKI) and no urgent indications, the optimal timing of kidney replacement therapy (KRT) initiation remains unresolved. Systematic reviews have reached conflicting conclusions, and no trial has prospectively validated a structured deferred strategy. We conducted an umbrella review to map the totality of evidence and appraise the certainty of findings on this question. Methods We searched PubMed, Scopus, Web of Science, and the Cochrane Library (2014 to 1 April 2025) for systematic reviews and meta-analyses (SR/MAs) of randomized controlled trials comparing early versus delayed KRT in ICU adults with AKI. Two reviewers independently performed selection, extraction, AMSTAR 2 quality assessment, and GRADE certainty ratings. Primary study overlap was quantified using the Corrected Covered Area (CCA). Findings were synthesized narratively given the high overlap. The protocol was registered on PROSPERO (CRD42024558822). Results A total of 25 SR/MAs were included. The CCA was 49.75% (very high overlap). AMSTAR 2 quality was high in 2, low in 9, and critically low in 14 reviews. Early KRT was not associated with reduced mortality in any included review. The two reviews rated high on AMSTAR 2 reported risk ratios close to unity: Fayad et al. (Cochrane; RR 0.97, 95% CI 0.87–1.09) and Li et al. (RR 1.00, 95% CI 0.95–1.04). Certainty of evidence was predominantly low or very low. ICU length of stay was modestly reduced in three reviews (MD − 0.69 to −3.24 days); other secondary outcomes were inconsistent. Early KRT consistently increased procedure-related harm: hypotension (6/7 reviews), catheter-related infection (5/6), and hypophosphatemia (5/6). Conclusions Across 25 SR/Mas, early KRT is not associated with a survival benefit and consistently increases procedure-related adverse events in critically ill adults with AKI. The current practice of deferring KRT is supported only by the failure to demonstrate benefit from early initiation, not by prospective evidence of its own safety: neither strategy has been prospectively validated. The marked heterogeneity in how “early” and “delayed” were defined further indicates that a binary timing paradigm inadequately captures the complexity of AKI, and that future research should move toward more individualized initiation strategies.