定量脑电图参数对重症动脉瘤性蛛网膜下腔出血患者预后的预测价值

Predictive value of quantitative EEG parameters on prognosis of patients with severe aneurysmal subarachnoid hemorrhage

目的探讨定量脑电图参数对重症动脉瘤性蛛网膜下腔出血(SaSAH)患者发病后90 d预后预测的可行性。方法前瞻性连续纳入2022年9月至2023年9月河南省人民医院神经外科重症监护室(NSICU)收治的SaSAH患者,收集患者的基线资料,包括年龄、性别、病史(高血压病、糖尿病、冠心病、卒中)、吸烟史、饮酒史、动脉瘤位置(前循环、后循环)、手术方式(开颅手术、介入手术、复合手术)、入院Hunt-Hess分级、格拉斯哥昏迷量表(GCS)评分、急性生理学与慢性健康状况评分系统Ⅱ(APACHEⅡ)评分、蛛网膜下腔出血早期脑水肿评分(SEBES)、入NSICU后首次随机血糖、乳酸水平及NSICU住院时间。所有患者入NSICU后48 h内均行定量脑电图监测,收集振幅整合脑电图(aEEG)上、下边界、95%频谱边缘频率(SEF95)、α变动、(δ+θ)与(α+β)功率比(DTABR)、大脑对称性指数(BSI)、光谱熵。根据发病90 d后的改良Rankin量表(mRS)评分将患者分为预后良好(mRS评分≤2分)组和预后不良(mRS评分3~6分)组。采用Spearman等级相关法分析SaSAH患者定量脑电图参数与mRS评分的相关性。采用多因素Logistic回归分析法分析预后不良的相关因素,并绘制受试者工作特征(ROC)曲线以评价各指标预测患者预后不良的效能。结果(1)共纳入SaSAH患者72例,预后不良组47例,预后良好组25例,发病后90 d预后不良占65.3%。两组患者在性别、年龄、高血压病、糖尿病、冠心病、卒中史、吸烟史、饮酒史、动脉瘤位置、手术方式、乳酸水平、NSICU住院时间方面差异均无统计学意义(均P>0.05);Hunt-Hess分级、SEBES、随机血糖差异均具有统计学意义(均P<0.05);与预后良好组相比,预后不良组患者的aEEG上、下边界及SEF95、α变动、光谱熵更低,DTABR、BSI更高(均P<0.05)。(2)Spearman等级相关分析显示,aEEG上边界(r=-0.41,P<0.01)、aEEG下边界(r=-0.54,P<0.01)、SEF95(r=-0.46,P<0.01)、α变动(r=-0.53,P<0.01)和光谱熵(r=-0.39,P<0.01)与SaSAH患者mRS评分呈负相关,DTABR(r=0.52,P<0.01)、BSI(r=0.33,P<0.01)与SaSAH患者mRS评分呈正相关。(3)多因素Logistic回归分析结果表明,Hunt-Hess分级(Ⅳ级比Ⅲ级:OR=1.203,95%CI:1.005~1.441,P=0.044;Ⅴ级比Ⅲ级:OR=1.661,95%CI:1.109~2.487,P=0.014)、SEBES(OR=1.647,95%CI:1.050~2.586;P=0.030)、aEEG下边界(OR=0.687,95%CI:0.496~0.953;P=0.024)、SEF95(OR=0.436,95%CI:0.202~0.937;P=0.034)、α变动(OR=0.368,95%CI:0.189~0.717;P=0.003)、DTABR(OR=1.324,95%CI:1.064~1.649;P=0.012)和BSI(OR=1.513,95%CI:1.026~2.231;P=0.036)是SaSAH患者预后不良的相关因素。ROC曲线分析显示,上述7项指标均对SaSAH患者预后不良有一定预测价值,其中DTABR预测SaSAH患者预后不良的效能最高,其曲线下面积为0.862(95%CI:0.761~0.932),敏感度为85.11%,特异度为80.00%。结论定量脑电图参数aEEG下边界、SEF95、α变动、DTABR、BSI可能对SaSAH患者的短期预后具有一定的预测价值,有待于未来多中心大样本研究进一步证实。

Objective To explore the feasibility of quantitative EEG parameters for prognostic prediction of patients with severe aneurysmal subarachnoid hemorrhage(SaSAH)90 d after the onset of the disease.Methods Patients with SaSAH admitted to the Neurosurgical Intensive Care Unit(NSICU)of Henan Provincial People′s Hospital from September 2022 to September 2023 were prospectively consecutively enrolled,and baseline data were collected,including age,gender,medical history(hypertension,diabetes mellitus,coronary artery disease,and stroke),history of smoking,history of drinking,location of aneurysm(anterior circulation,posterior circulation),surgical modality(craniotomy,interventional surgery,hybrid surgery),Hunt-Hess classification,Glasgow coma scale(GCS)score,acute physiology and chronic health status scoring systemⅡ(APACHEⅡ)score,subarachnoid hemorrhage early brain edema score(SEBES),first randomized blood glucose level after admission to NSICU,lactate level,and duration of NSICU stay.Quantitative EEG monitoring was performed in all patients within 48 h after admission to the NSICU,and amplitude-integrated electroencephalogram(aEEG)upper and lower boundaries,95%spectral edge frequency(SEF95),αchange,(δ+θ)to(α+β)power ratio(DTABR),brain symmetry index(BSI),and spectral entropy were collected.Based on modified Rankin scale(mRS)scores 90 d after onset,patients were categorized into good prognosis(mRS score≤2 points)and poor prognosis(mRS score 3-6 points)groups.Spearman rank correlation was used to analyze the correlation between quantitative EEG parameters and mRS scores in SaSAH patients.Multifactorial Logistic regression analysis was used to screen for correlates of poor prognosis,and receiver operating characteristic(ROC)curves were plotted to evaluate the efficacy of each index in predicting patients′poor prognosis.Results(1)A total of 72 patients with SaSAH were included,with 47 in the poor prognosis group and 25 in the good prognosis group,and the poor prognosis rate at 90 d after the onset was 65.3%.There was no statistically significant difference between the two groups in terms of gender,age,hypertension,diabetes mellitus,coronary artery disease,history of stroke,history of smoking,history of drinking,location of aneurysm,surgical modality,lactate level,and length of hospitalization in the NSICU(all P>0.05);the differences between the Hunt-Hess grading,SEBES,and random blood glucose were statistically significant upon comparison(all P<0.05).Compared with the good prognosis group,the changes of aEEG upper and lower boundary,SEF95,αchange and spectral entropy were lower in the poor prognosis group,but DTABR and BSI were higher(all P<0.05).(2)Spearman rank correlation analysis showed that the upper border of aEEG(r=-0.41,P<0.01),lower border of aEEG(r=-0.54,P<0.01),SEF95(r=-0.46,P<0.01),αchange(r=-0.53,P<0.01)and spectral entropy(r=-0.39,P<0.01)were negatively correlated with the mRS scores of SaSAH patients,and DTABR(r=0.52,P<0.01)and BSI(r=0.33,P<0.01)were positively correlated with poor prognosis of SaSAH patients.(3)The results of multifactorial Logistic regression analysis showed that Hunt-Hess grading(levelⅣvs.Ⅲ:OR,1.203,95%CI 1.005-1.441,P=0.044;levelⅤvs.Ⅲ:OR,1.661,95%CI 1.109-2.487,P=0.014),SEBES(OR,1.647,95%CI 1.050-2.586;P=0.030),aEEG lower border(OR,0.687,95%CI 0.496-0.953l;P=0.024),SEF95(OR,0.436,95%CI 0.202-0.937;P=0.034),αchange(OR,0.368,95%CI 0.189-0.717;P=0.003),DTABR(OR,1.324,95%CI 1.064-1.649;P=0.012),and BSI(OR,1.513,95%CI 1.026-2.231;P=0.036)were influencing factors of poor prognosis in SaSAH patients.ROC curve analysis showed that all of the above seven indicators had a certain predictive value for poor prognosis in SaSAH patients,among which the area under the curve of DTABR was the highest as 0.862(95%CI 0.761-0.932),with sensitivity 85.11%and specificity 80.00%.Conclusion Quantitative EEG parameters aEEG lower border,SEF95,αchange,DTABR,and BSI may have certain predictive value for the short-term prognosis of SaSAH patients,which needs to be further confirmed in future multi-center large-sample studies.