Comment on "Trajectory pattern of serum urea nitrogen to creatinine ratio in acute kidney injury requiring continuous renal replacement therapy: A post hoc analysis of a retrospective multicenter cohort study".

We read with interest the recent multicenter post hoc cohort study evaluating serum urea nitrogen-to-creatinine ratio (UCR) trajectories in patients with acute kidney injury receiving continuous renal replacement therapy (CRRT). The authors identified three groups using group-based multitrajectory modeling and reported higher mortality in patients with elevated UCR or lactate. While the study addresses an important question in critical care risk stratification, several methodological and interpretive issues merit clarification. First, the identified classes appear to be driven primarily by baseline differences in UCR and lactate rather than clearly distinct longitudinal trajectories, raising the possibility that the findings reflect baseline prognostic stratification rather than true temporal phenotypes. Comparison with simpler baseline models incorporating day-1 variables would help determine the added value of the trajectory approach. Second, the requirement for at least two days of CRRT and the use of data up to day 3 for classification may introduce survivor or guarantee-time bias, as phenotype assignment depends on continued survival and treatment exposure. Analytical approaches such as landmark analysis or time-to-event modeling from a uniform baseline could improve validity. Third, UCR may be influenced by CRRT related factors, including modality, effluent dose, and nutritional support, complicating its biological interpretation as a marker of catabolism or illness severity. Adjustment for these variables would help distinguish treatment effects from underlying pathophysiology. Addressing these issues would strengthen the clinical applicability of UCR in risk stratification during CRRT.