Can EEG monitoring prevent delirium? The need for standardization and precision

Dear Editor, We read with great interest the recent systematic review and meta-analysis by Cao et al, titled “Effect of processed electroencephalography-guided anesthesia on postoperative delirium: an updated systematic review and meta-analysis,” published in your journal[1]. The authors provided a comprehensive and rigorous update of the evidence up to January 2025, confirming that processed electroencephalography (p-EEG)-guided anesthesia reduces the overall incidence of postoperative delirium (POD), particularly in elderly patients and non-cardiac surgeries. This study offers important evidence-based support for perioperative neuromonitoring. However, the significant regional heterogeneity and the limitations of current p-EEG technology revealed in the analysis merit further discussion to guide future clinical practice and research. First, one of the most striking findings of this study is the regional heterogeneity in efficacy. The pooled analysis showed clear benefits of p-EEG in European (RR 0.79) and East Asian/Australasian (RR 0.65) trials, whereas no advantage was observed in North American trials (RR 1.01)[1]. This discrepancy is unlikely coincidental and likely reflects systematic differences in anesthesia practices, drug selection, hemodynamic management targets, and the rigor of delirium assessment across health care systems. For instance, large North American trial may have adhered more strictly to hemodynamic-based fluid and vasopressor management protocols, which could partially offset the neuroprotective effects achieved solely through EEG-guided anesthesia depth control[2]. This suggests that p-EEG should not be viewed as an isolated technology, its effectiveness is highly dependent on the overall perioperative care pathway in which it is embedded[3]. Future guideline development and international multicenter studies must strive to describe and standardize synergistic factors, rather than merely prescribing a target range for EEG indices. Second, the review appropriately highlights the inherent limitations of current commercial p-EEG monitoring devices in the discussion. Reliance on single, processed indices (such as BIS, PSI) is susceptible to signal artifacts, pharmacological variability, and may overlook age-related electroencephalographic features (such as attenuated alpha power in elderly patients)[4]. As cited in the article, different commercial “depth of anesthesia” monitors may provide conflicting recommendations for identical EEG signals[5]. This strongly supports the argument that, for more reliable POD prevention, clinical practice may need to transition from dependence on processed indices toward the ability to interpret raw EEG. Raw EEG allows direct visualization of burst suppression duration and patterns, assessment of alpha-delta coherence, and other neurophysiological biomarkers closely linked to delirium risk[6]. Although this poses higher training demands for anesthesiologists, it may be the key to overcoming inter-device algorithmic variability and achieving truly individualized brain function monitoring. Therefore, we endorse the authors’ conclusion regarding the need for standardization and wish to emphasize that standardization should occur at two levels. The first is standardization of operational protocols, including unified thresholds for raw EEG burst suppression (the >4.4-minute high-risk threshold mentioned in the article), risk stratification incorporating preoperative cognitive baseline assessments, and clear intraoperative multimodal management protocols (integrating EEG, hemodynamic, and analgesic monitoring)[7,8]. The second is standardization of clinician competency, specifically enhancing training for anesthesiologists in raw EEG interpretation, enabling them to move beyond reliance on a single numerical index and understand the dynamic network states of the brain under anesthesia[4]. In summary, the study by Cao et al affirms the potential value of p-EEG monitoring while clearly delineating the challenges to realizing its universal benefit. The future direction should shift from a single index target toward a precision neuroprotection strategy that integrates raw EEG interpretation, embeds standardized perioperative pathways, and is guided by individual patient EEG characteristics. We look forward to more prospective studies validating such integrated approaches. We confirm that our article adheres to the TITAN Guidelines 2025, as evidenced by the completed TITAN Guideline Checklist 2025[9].