Stoughton J, Nath R L, Abbott W M
Departments of Surgery, The Lawrence Memorial Hospital of Medford and Winchester, and Massachusetts General Hospital, USA.
J Vasc Surg. 1998 Dec;28(6):1014-21; discussion 1021-3. doi: 10.1016/s0741-5214(98)70027-8.
This study examines the accuracy of intraoperative electroencephalographic (EEG) monitoring for the detection of cerebral ischemia by comparing EEG with simultaneous mental status evaluation (MSE) during carotid endarterectomy in awake patients.
Between 1994 and 1997, 208 consecutive carotid endarterectomies were prospectively evaluated for cerebral function during surgery with simultaneous MSE and EEG monitoring. Regional anesthesia (RA), which consisted of superficial cervical block, was chosen preferentially in 75% of the cases, with general anesthesia (GA) reserved for the patients who did not fulfill the criteria for RA. When available, 8-channel EEG monitoring was performed (59% with RA and 55% with GA).
The EEG was a reliable predictor in comparison with MSE in most but not all cases of cerebral ischemia. Significant neurologic changes were noted using MSE in 4 of 89 patients (4.5%) that were not detected using EEG (false negative results). Conversely, 6 of 89 cases (6.7%) showed unilateral slowing without associated changes in MSE (false positive results). For the awake patients, 21 of 150 cases (14%) showed MSE changes that required a shunt. By contrast, 9 of 32 GA cases (28%) showed EEG changes that would have led to shunting (P = NS). In the RA group, there were no strokes versus 3 of 58 cases (5.2%) with strokes in the GA group. Two of 150 cases (0.1%) had transient ischemic attacks in the RA group. There was 1 myocardial infarction in the GA group; no deaths occurred in this series.
EEG monitoring yielded a significant number of false positive (6.7%) and false negative (4.5%) results in the detection of neurologic deficits when compared with MSE in the awake patients. In this series, the preferential use of RA resulted in less shunt use and was possibly associated with a lower stroke rate.
本研究通过在清醒患者颈动脉内膜切除术期间，将术中脑电图（EEG）监测与同步精神状态评估（MSE）进行比较，来检验EEG监测检测脑缺血的准确性。
1994年至1997年间，对208例连续的颈动脉内膜切除术患者在手术期间进行了同步MSE和EEG监测，以对脑功能进行前瞻性评估。75%的病例优先选择由颈浅丛阻滞组成的区域麻醉（RA），全身麻醉（GA）则留给不符合RA标准的患者。若条件允许，进行8通道EEG监测（RA组为59%，GA组为55%）。
与MSE相比，EEG在大多数但并非所有脑缺血病例中都是可靠的预测指标。89例患者中有4例（4.5%）使用MSE时发现有明显的神经学变化，但EEG未检测到（假阴性结果）。相反，89例中有6例（6.7%）显示单侧减慢，而MSE无相关变化（假阳性结果）。对于清醒患者，150例中有21例（14%）显示MSE变化需要分流。相比之下，32例GA病例中有9例（28%）显示EEG变化本会导致分流（P=无显著差异）。RA组无中风病例，而GA组58例中有3例（5.2%）中风。RA组150例中有2例（0.1%）发生短暂性脑缺血发作。GA组有1例心肌梗死；本系列无死亡病例。
与清醒患者的MSE相比，EEG监测在检测神经功能缺损时产生了大量假阳性（6.7%）和假阴性（4.5%）结果。在本系列中，优先使用RA导致分流使用减少，并且可能与较低的中风发生率相关。
