Commentary on “The Red Blood Cell Distribution Width‐Albumin Ratio Predicts Postoperative Delirium in the Elderly With Gastrointestinal Tumors: A Retrospective Cohort Study”

We read with great interest the recent article by Du et al. 1 investigating the red blood cell distribution width-albumin ratio (RAR) as a predictor of postoperative delirium (POD) in older adults undergoing gastrointestinal tumor surgery. The authors address an important clinical question with a rigorous analytical approach, including propensity score matching to balance confounding variables. Their finding that elevated preoperative RAR is independently associated with POD—and demonstrates superior predictive performance compared to hypertension, creatinine, or mean corpuscular hemoglobin alone—offers a clinically valuable contribution given the routine availability of these laboratory parameters. We respectfully offer several methodological considerations to enhance the interpretation and future application of these findings. First, while propensity score matching was appropriately employed, the reduction in sample size from 203 to 108 patients (54 pairs) raises questions about estimate stability. Only 17 POD events occurred in the matched cohort (16 in the high RAR group vs. 1 in the low RAR group), yet no post hoc power analysis was reported for this reduced sample. The wide confidence interval for the post-PSM odds ratio (2.462; 95% CI 1.356–4.467) suggests some imprecision. A power calculation or explicit acknowledgment of this limitation would contextualize the reliability of the matched analysis findings. Second, the exclusion of patients with preoperative psychiatric or neurological disorders, while methodologically sound, may inadvertently bias the sample toward healthier individuals. Subclinical cognitive impairment—highly prevalent in older surgical populations and a powerful predictor of POD—was not assessed. If RAR is also associated with underlying cognitive vulnerability, the exclusion of overt neurological disease may actually attenuate the observed RAR-POD association. Future studies incorporating formal preoperative cognitive screening would clarify whether RAR adds predictive value beyond cognitive status. Third, POD was treated as a binary outcome, but emerging evidence suggests that delirium duration and severity are clinically meaningful endpoints associated with long-term outcomes 2. The authors had access to medical and nursing documentation yet did not characterize delirium severity or duration. RAR might differentially predict transient versus persistent POD, and examining these granular outcomes in future work would provide a more nuanced understanding of its prognostic significance. We commend the authors for an important and well-executed study addressing a practical clinical question. Addressing these methodological considerations in future prospective work would further strengthen the evidence base for RAR as a perioperative risk stratification tool. J.H. and L.W. were responsible for literature collection and drafting the initial manuscript. X.G. supervised the study. The order of co-first authors was determined by flipping a coin. L.W. and J.H. made equal contributions to this manuscript. All authors read and approved the final version of the manuscript. The authors have nothing to report. The authors have nothing to report. The authors declare no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.