Aim: Postoperative delirium (POD) is a common complication with significant adverse effects in elderly patients. Electroencephalography (EEG) provides a promising approach for predicting the risk of POD. This study aims to elucidate the correlation between intraoperative EEG spectrum and the incidence of POD in elderly patients undergoing orthopedic surgery.
Methods: A single-center prospective observational cohort study was conducted at Zhongda Hospital, Southeast University, from September 2022 to March 2023, registered on Chinese Clinical Trail Registry (ChiCTR2300069548). Among the 172 patients who underwent orthopedic surgery, 125 completed the study with available data. Preoperative baseline cognitive function was assessed using the mini-mental state examination (MMSE). An intraoperative 4-channel EEG was recorded. Total power spectra and power spectral density for beta, alpha, theta, and delta bands were calculated. Spectral edge frequency, burst suppression ratio, and patient state index were directly extracted from the EEG monitor. The primary outcome was POD, assessed using a 3-minute Diagnostic Interview for Confusion Assessment Method (CAM)-Defined Delirium scale or the CAM for the Intensive Care Unit. Patients were divided into POD and non-postoperative delirium (non-POD) groups. Logistic regression analysis was conducted to evaluate the independent predictive effect of intraoperative spectral distribution and other quantitative EEG data for POD.
Results: The incidence of delirium within 72 hours after surgery was 8.8%. Compared to the non-POD group, the POD group showed significantly lower absolute power in the beta [0.06 (0.03, 0.09) dB vs 0.14 (0.08, 0.26) dB, p < 0.001], alpha [0.41 (0.25, 0.71) dB vs 1.24 (0.55, 2.57) dB, p = 0.008], and theta [0.77 (0.44, 1.01) dB vs 1.19 (0.72, 2.02) dB, p = 0.035] bands. Logistic regression analysis identified several independent risk factors for POD, including lower intraoperative alpha power [odds ratio (OR) 10.210, 95% confidence interval (CI) 1.233-84.568, p = 0.031], advanced age (OR 7.713, 95% CI 1.022-58.204, p = 0.048), preoperative anemia (OR 6.636, 95% CI 1.091-40.358, p = 0.040), and preoperative depression (OR 10.089, 95% CI 1.029-98.909, p = 0.047). In contrast, higher preoperative MMSE scores appeared to be a protective factor for POD (OR 0.130, 95% CI 0.021-0.820, p = 0.030).
Conclusions: Intraoperative EEG frontal low alpha power demonstrated a significant independent association with POD in elderly patients after orthopedic surgery. This parameter may serve as an intraoperative neurophysiological marker of cerebral vulnerability to POD. Additionally, advanced age, lower preoperative MMSE scores, preoperative anemia, and preoperative depression were independent risk factors for POD.
Clinical trial registration: Chinese Clinical Trial Registry (ChiCTR2300069548).