Neurosurgery ICU, Xiangya Hospital, Central South University, Changsha, China.
General ICU/Department of Critical Care Medicine, Xiangya Hospital, Central South University, Changsha, China.
Clin EEG Neurosci. 2022 Sep;53(5):452-459. doi: 10.1177/1550059420984934. Epub 2021 Jan 6.
This study aimed to explore the effectiveness of quantitative electroencephalogram (EEG) and EEG reactivity (EEG-R) to predict the prognosis of patients with severe traumatic brain injury.
This was a prospective observational study on severe traumatic brain injury. Quantitative EEG monitoring was performed for 8 to 12 hours within 14 days of onset. The EEG-R was tested during the monitoring period. We then followed patients for 3 months to determine their level of consciousness. The Glasgow Outcome Scale (GOS) score was used. The score 3, 4, 5 of GOS were defined good prognosis, and score 1 and 2 as poor prognosis. Univariate and multivariate analyses were employed to assess the association of predictors with poor prognosis.
A total of 56 patients were included in the study. Thirty-two patients (57.1%) awoke (good prognosis) in 3 months after the onset. Twenty-four patients (42.9%) did not awake (poor prognosis), including 11 cases deaths. Univariate analysis showed that Glasgow coma scale (GCS) score, the amplitude-integrated EEG (aEEG), the relative band power (RBP), the relative alpha variability (RAV), the spectral entropy (SE), and EEG-R reached significant difference between the poor-prognosis and good-prognosis groups. However, age, gender, and pupillary light reflex did not correlate significantly with poor prognosis. Furthermore, multivariate logistic regression analysis showed that only RAV and EEG-R were significant independent predictors of poor prognosis, and the prognostic model containing these 2 variables yielded a predictive performance with an area under the curve of 0.882.
Quantitative EEG and EEG-R may be used to assess the prognosis of patients with severe traumatic brain injury early. RAV and EEG-R were the good predictive indicators of poor prognosis.
本研究旨在探讨定量脑电图（EEG）和脑电图反应性（EEG-R）对预测严重创伤性脑损伤患者预后的有效性。
这是一项关于严重创伤性脑损伤的前瞻性观察研究。在发病后 14 天内进行 8-12 小时的定量 EEG 监测。在监测期间进行 EEG-R 测试。然后，我们对患者进行了 3 个月的随访，以确定他们的意识水平。采用格拉斯哥结局量表（GOS）评分。GOS 评分 3、4、5 定义为预后良好，评分 1 和 2 为预后不良。采用单变量和多变量分析评估预测因子与预后不良的相关性。
共有 56 例患者纳入研究。32 例患者（57.1%）在发病后 3 个月苏醒（预后良好）。24 例患者（42.9%）未苏醒（预后不良），包括 11 例死亡。单变量分析显示，格拉斯哥昏迷评分（GCS）、振幅整合脑电图（aEEG）、相对频带功率（RBP）、相对阿尔法变异性（RAV）、频谱熵（SE）和 EEG-R 在预后不良组和预后良好组之间差异有统计学意义。然而，年龄、性别和瞳孔对光反射与预后不良无显著相关性。此外，多变量逻辑回归分析显示，只有 RAV 和 EEG-R 是预后不良的显著独立预测因子，包含这 2 个变量的预后模型的曲线下面积为 0.882，具有良好的预测性能。
定量脑电图和脑电图反应性可用于早期评估严重创伤性脑损伤患者的预后。RAV 和 EEG-R 是预后不良的良好预测指标。
