Authors: Khoudrietchan C et al.
Anesthesiology. January 2026.
Summary:
This Letter to the Editor comments on Li et al.’s recent study evaluating whether intraoperative midazolam increases postoperative delirium in elderly patients. The authors commend the investigators for addressing an important clinical question using a multicenter cohort but raise several methodological concerns that may limit interpretation of the findings.
They note that delirium was assessed using the 3D-CAM twice daily for seven postoperative days or until hospital discharge. Because delirium fluctuates and can appear and resolve within hours, assessments spaced 12 hours apart could easily miss episodes—especially within the high-risk window of the first 24–72 hours. They emphasize the balance between detection and practicality: more frequent checks might improve accuracy but also increase staff workload or disturb patient rest.
The letter highlights that most delirium occurred on postoperative day 1 with a median onset around midday. This supports their concern that infrequent assessments may underdetect cases and may also contribute to inaccurate incidence reporting.
The authors further discuss potential confounding. They note that patients selected to receive midazolam tended to be younger and healthier—groups inherently at lower risk for postoperative delirium. Without adjusting for these baseline differences using robust statistical methods (such as matching or weighting), the results may be biased.
They also critique the secondary analysis examining postoperative anxiety. Although the study reported lower anxiety scores in patients without midazolam, the authors caution against overinterpreting this finding. The anxiety measure used (GAD-7) is not validated specifically for immediate postoperative use, and the timing of assessments may not reliably capture true perioperative anxiety.
Overall, while supporting the importance of the question studied, the authors warn that methodological limitations—assessment frequency, confounding, and interpretation of secondary outcomes—prevent strong conclusions about the effect of midazolam on delirium.
What You Should Know
• Twice-daily delirium assessments risk missing fluctuating episodes, particularly during the critical early postoperative period.
• Most delirium occurred on postoperative day 1, reinforcing the need for high-frequency early assessments.
• Potential confounding exists because midazolam was given to younger, healthier patients—groups less likely to develop delirium.
• The exploratory anxiety outcome is difficult to interpret due to timing and limitations of GAD-7 in postoperative settings.
• Overall results should be interpreted cautiously due to underdetection concerns and lack of adjustment for baseline differences.
Key Points
• Delirium assessments every 12 hours may substantially under-detect postoperative delirium.
• The highest-risk time window (first 24–72 hours) requires more rigorous monitoring to capture transient episodes.
• Baseline patient differences between groups introduce confounding not fully addressed by the study.
• The anxiety outcome has limited clinical validity in this setting and may not be reliable.
• Methodological concerns limit the strength of conclusions regarding midazolam’s true impact on delirium.
Thank you for allowing us to use this article from Anesthesiology.