Abstract
The most influential timing of continuous renal replacement therapy (CRRT) in gravely ill patients with acutely severe, uncompensated renal dysfunction continues to be a subject of ongoing debate. Although earlier initiation has been hypothesized to limit metabolic disturbances and prevent downstream organ dysfunction, clinical trials have yielded inconsistent findings, in part because of variability in how "early" initiation is defined across studies. This meta-analysis, in conjunction with a systematic review, sought to examine the correlation between early versus delayed CRRT initiation and mortality, with analyses stratified according to initiation strategy, including clinical staging-based benchmark by Kidney Disease: Improving Global Outcomes (KDIGO) and biomarker-driven approaches using neutrophil gelatinase-associated lipocalin (NGAL). A literature search was conducted in PubMed, EMBASE, and the Cochrane Library to identify studies published between January 2015 and June 2025. Eligible studies included randomized controlled trials and observational investigations that compared early and delayed CRRT initiation among critically ill adults with acute kidney injury (AKI). The primary outcome of interest was all-cause mortality measured between 28 and 90 days or at the time of intensive care unit (ICU) discharge. Pooled relative risks (RR) and odds ratios (OR) were estimated using random-effects meta-analytic models, with subgroup analyses performed according to initiation criteria. Nine studies encompassing 2,349 patients were included (six randomized trials and three observational studies). Overall, early CRRT initiation was not associated with a statistically significant decrease in mortality compared with delayed initiation (risk ratio (RR) = 0.87; 95% confidence interval (CI), 0.69-1.10; p = 0.25; I(2) = 90.4%). Subgroup analysis demonstrated no significant mortality benefit with biomarker-guided (NGAL-based) early initiation (RR = 0.90; 95% CI, 0.41-2.01), whereas KDIGO-based initiation showed a borderline association favoring early therapy (RR = 0.75; 95% CI, 0.57-0.99), though heterogeneity remained substantial. No meaningful interaction was observed between initiation strategy and mortality.