Early renal replacement therapy and short-term mortality in higher-risk KDIGO stage 2 sepsis-associated acute kidney injury with sustained oliguria: a target trial emulation

Abstract Whether proactive early renal replacement therapy (RRT) improves outcomes in higher-risk patients with KDIGO stage 2 sepsis-associated acute kidney injury (SA-AKI) and sustained oliguria remains uncertain. We evaluated its association with short-term mortality. We emulated a target trial using the Medical Information Mart for Intensive Care IV database, version 3.1. Adults who developed KDIGO stage 2 SA-AKI based on sustained oliguria within 7 days after sepsis onset and had a SOFA score ≥ 8 or were receiving vasopressors were included after exclusion of baseline urgent indications for RRT. Time zero (T0) was defined as the first time after sepsis onset at which the urine-output KDIGO stage 2 criterion was met. We compared proactive initiation of any RRT modality within 12 h and before the first prespecified rescue indication with a no early RRT strategy that permitted RRT after a rescue indication. A clone-censor-weight approach using stabilized inverse probability-of-censoring weights, weighted pooled logistic regression, and standardization was applied. The primary and secondary outcomes were 28-day and 7-day all-cause mortality. Among 3,523 patients, 42 (1.2%) initiated RRT within 12 h and 30 (0.9%) initiated proactive RRT before a rescue indication. Estimated 7-day mortality risks were 18.2% with no early RRT and 21.0% with proactive early RRT, corresponding to a risk difference of 2.8% points (95% CI, − 5.5 to 14.1) and a risk ratio of 1.15 (95% CI, 0.70 to 1.79). Corresponding 28-day risks were 32.1% and 26.4%, with a risk difference of − 5.8% points (95% CI, − 16.4 to 9.7) and a risk ratio of 0.82 (95% CI, 0.50 to 1.30). Sensitivity analyses did not materially change the overall interpretation. Diagnostic analyses indicated substantial practical positivity limitations for the proactive early RRT strategy. We found no clear evidence that routine proactive RRT initiation within 12 h reduced 7-day or 28-day mortality in this higher-risk population. The estimates were imprecise, these findings should be considered exploratory and hypothesis-generating rather than providing strong causal evidence.