A NEW study introduces a practical bedside tool to help clinicians determine when patients with severe acute kidney injury (AKI) can be safely weaned off renal replacement therapy (RRT).
Uncertainty Around Stopping RRT
RRT is a life-saving intervention in critically ill patients with advanced AKI, but deciding when to discontinue treatment remains a major clinical challenge. Even after clinicians stop RRT, there is significant uncertainty about whether kidney function will recover sufficiently to avoid restarting therapy. To address this gap, researchers from the DOORS study sought to identify reliable predictors of successful RRT weaning and to develop a tool to guide post-discontinuation decisions.
The investigators conducted a post-hoc analysis of two large multicentre trials (AKIKI and AKIKI2), focusing on ICU patients with severe (KDIGO stage 3) AKI managed using a conservative approach to RRT initiation. Among 554 patients who received RRT, 180 underwent an attempt at weaning, defined as discontinuation for at least three consecutive days. Of these, 101 patients (56%) were successfully weaned, meaning they did not require RRT again within seven days.
Using multivariable logistic regression, the team identified six key predictors of successful weaning. These included factors present before the weaning attempt, such as duration of RRT, presence of septic shock on admission, and baseline serum creatinine, as well as clinical parameters assessed after discontinuation, namely vasopressor use, need for invasive mechanical ventilation, and urine output.
Development of the UNDERSCORE Model
These variables were combined into a predictive model called the UNDERSCORE. The tool demonstrated strong performance in the original study population, with an area under the curve (AUC) of 0.86, indicating high discriminative ability. External validation in an independent Swiss ICU cohort of 415 patients showed more modest but still clinically useful performance (AUC 0.73), despite a broader and more heterogeneous patient population in which 81% were successfully weaned.
The authors conclude that the UNDERSCORE provides a simple and pragmatic method for estimating the likelihood of sustained RRT discontinuation. By supporting bedside decision-making, the tool may help clinicians better tailor patient monitoring and resource use following RRT cessation, potentially improving outcomes in this high-risk population.
Reference
Chaïbi K et al. Predictive factors of successful renal replacement therapy weaning. Intensive Care Med. 2026;DOI: 10.1007/s00134-026-08357-x.
Featured image: zlikovec on Adobe Stock
