Net Ultrafiltration During Continuous Kidney Replacement Therapy for Acute Kidney Injury: From Fixed Rates to Precision Fluid Removal

Abstract

Fluid overload is common among critically ill patients with acute 
kidney injury (AKI), particularly those requiring continuous kidney 
replacement therapy (CKRT), and is consistently associated with 
adverse clinical outcomes. Net ultrafiltration (UFNET), defined 
as the extracorporeal removal of fluid from the patient to reduce 
extracellular volume, is therefore a fundamental component of 
CKRT. However, unlike solute clearance, UFNET has historically 
been prescribed without well-established dosing principles. 
Insufficient fluid removal may prolong exposure to pulmonary, 
systemic, and organ edema, whereas overly rapid fluid removal 
may exceed vascular refill, reduce effective circulating volume, 
provoke hemodynamic instability, and potentially impair organ 
perfusion and kidney recovery. Observational studies have 
suggested a J- or U-shaped relationship between UFNET rate 
and mortality, with intermediate rates potentially indicating a 
favorable outcome with lower mortality.
More recent evidence challenges the interpretation of these 
observational thresholds as causal treatment targets. A recent 
multicenter analysis using hourly CKRT and fluid-balance 
found that the association between UFNET and mortality was 
substantially mediated by achieved fluid balance. Furthermore, 
the RELIEVE-AKI feasibility randomized trial demonstrated that 
maintaining targeted restrictive and liberal UFNET was challenging, 
as clinicians repeatedly adjusted the prescribed regimen. These 
findings suggest that UFNET should not be conceptualized as a 
fixed dose analogous to small-solute clearance. Rather, UFNET is 
a dynamic intervention whose appropriate intensity depends on 
the competing objectives of relieving congestion and preserving 
circulatory and organ perfusion.
This review examines the terminology, physiology, observational 
and interventional evidence, challenges in causal inference, 
bedside assessment, clinical phenotypes, kidney recovery, and 
future research priorities related to UFNET during CKRT. We 
propose a conceptual framework in which fluid-removal need and 
fluid-removal tolerance could determine individualized UFNET 
prescription, with repeated reassessment based on achieved fluid 
balance and physiologic response.