Prognostic Predictive Value Of EEG Classification Based On EEG Reactivity In Patients With Disorders Of Consciousness Caused By Cerebrovascular Diseases

BackgroundAcute cerebrovascular diseases are clinically common with high morbidity, high mortality and high disability, which become a serious threat to the lives of patients. Acute cerebrovascular diseases are heavy burdens to their families and our country. Cerebrovascular diseases, also known as stroke, are caused by a variety of vascular diseases of the brain, mainly including the ischemic and hemorrhagic cerebrovascular diseases. Patients with disorders of consciousness caused by acute cerebrovascular diseases are badly and always have poor prognosis. So the prognostic predictive value of these patients disorders of consciousness caused by acute cerebrovascular diseases is extremly importent. Imagings such as CT, MRI and others are valuable to the patients, which are commonly used by dorctors. However, patients can not be thansfered to imaging for disease severity sometimes. Also, imaging can’t be used as a bedside complement continuously. So these examnations have certain restrictions for the patients because of depending on the diseases of these patients. Neurophysiological techniques developed rapidly in recent years. The examnations are safe, non-invasive, objective and easy and been widely used for patients with cerebrovascular diseases. EEG(Electroencephalogram, EEG) are commonly used currently. Many studies show that EEG are critically importent in patients with disorders of consciousness particularly. EEG is one of the most useful methods to evaluate a patient’s state of consciousness and can assess the prognosis of the patients. EEG is easy to operate and you can monitor the process of illnesses, especially for patients in neurological intensive care unit (NICU)Classifications of EEG standard appeared because the EEG is more difficult and complex to learn for clinic doctors. Dorctors can assess patients easily and have a better communication with the patients and their relatives if they can understand the EEG classifications. Doctors can carry out the treatment plan as soon as possible.Synek classification standards and Wang Xiaomei standards are usely common. Both classifications are already proved to be better reflected the severity of illness of patients and can assess the patient’s prognosis. However, there are some shortcomings for the two classifications. Synek classification is aimly primarily to the patients with diffused brain damage such as patients suffered cardiac arrest. Wang Xiaomei published articles to prove that Synek classification is applicable to patients with cerebrovascular diseases. Synek classification is not suitable to some patients with slow-wave EEG. In addition, it is not suitable to patients with RAWOD and RAWCS modes. So Synek classification has some certain restrictions and many patients cannot be classified by it. Young classification can assess the prognosis of patients.The biggest feature contained three-phase wave coma mode, while the main shortcoming is that the distinction of the slow-wave system is rough.9/8 wave of> 50% is defined as Class I and many cases are focused on the level I. If 0/8 wave<50%, cases will not be included. RAWOD and RAWCS modes of acute massive cerebral infarction were not included. The effect of this classification is poor to evaluate the prognosis of patients. Wang Xiaomei proposed a new classification standard in 2010 for patients with severe cerebrovascular diseases. The advantages of the standard were based on the characteristics of cerebrovascular diseases.The EEG classification furtherly divided slow waves into focal and diffuse, and included RAWOD and RAWCS modes. It is suitable for almost all the patients and much more valuable than Synek standard. Otherwise, it is not fully reflect the EEG reactivity. Wang Xiaomei defined a special coma mode as class III in the NICU.The spindle wave coma was rare occurred in NICU. These factors may reduce the accuracy of prognosis of patients. All the classifications were not fully reflect the importance of EEG reactivity.Foreign research has proved that EEG reactivity is a strong positive prognostic factor for patients with disorder of consciousness.The patients with EEG reactivity is 90% in sensitivity,94% in specificity; 96% in mortality while 93% in patients without EEG reactivity.It is much more valuable than GCS score and brainstem reflex. EEG reactivity is important to predict the prognosis. Because of the importance of EEG reactivity in terms of judging the prognosis of patients and the characteristics in neurological intensive care unit, we proposed the classification of EEG based on EEG reactivity (ECBER)in 2012. And we have proved that the EEG reactivity is importent to the prognosis of patients with disorders of consciousness by small-scale clinical studies.We proposed new EEG classification based on EEG reactivity (ECBER) in 2012 in department of Neurology NICU, Nanfang Hospital,considering the importance of EEG reactivity. The classification is as follows:EEG reactivity to the first category (Ⅰ-Ⅲ grade); RAWOD or RAWCS and other special modes are the second class (IV grade); No EEG reactivity to the third category (Ⅴ-Ⅵ level). EEG of patients were classified according to Synek, Wang Xiaomei and ECBER standards. The relationships between the three different EEG grading standards and the prognosis of patients three months later after the onset are analyzed. Survival and mortality are outcome measures.The purpose of this research:1. To investigate the value of ECBER classification on patients with disorders of consciousness by cerebralvascular diseases.2.The relationships between the three different EEG grading standards and the prognosis of patients 3 months later are analyzed and we compare the prognostic value, accuracy and their advantages and disadvantages.We aim to provide better classification method for the patients in neurological intensive care unit (NICU).Subjects and methods1. SubjectsA study was carried out at the NICU of Nanfang hospital, Southern Medical University of China from July 2013 to October 2014. Patients included:1) Adult patients, age range from 18-90 years.2) disorders with consciousness during the EEG examination.3) patients who received EEG examination in the first three days after admitting NICU.4) CT or MRI was performed to confirm the cerebrovascular diseases. Exclusion criteria:(1) patients with previous history of stroke; (2) patients with histories of serious heart, liver, kidney disease, cancers and other diseases affecting the prognosis of patients; (3) EEG examination interfered obviously; (4) the influence of EEG state activities, such as shock, hypothermia and other states;(5) patients using antiepileptic drugs or drugs significantly affect brain function, such as a sedative narcotic drugs within the 24 h before EEG examinations; (6) patients with arterial or venous thrombolysis. (7) surgery or minimally invasive cranial surgery passing. Exclusion criteria:patients lost, or abandoned. All patients admitted to the NICU were recorded demographic data, including gender, age, diagnosis, medical history and other relevant information and GCS score.2. The specific EEG methodPatients selected got EEG examinations by the same Nicolet EEG instrument. Ensuring the vital signs of patients relatively stable, EEG examinations were operated when patients were quiet and with small interference. Each EEG record was for 30 minutes. Silver-chloride EEG electrodes were applied to the scalp according to the currently internationally accepted standard 10-20 electrode placement. Before examining the patients we remove grease and cuticle (all are controlled under 10 kQ). Select the 10 unipolar system:F1-Al, C3-A1, T3-A1, P3-A1, O1-A1;F2-A2, C4-A2, T4-A2, P4-A2, O2-A2. Point the ground electrode Fpz, Fz electrode for the reference electrode. Range of parameters: frequency range 0.5Hz-30Hz, voltage 10uV/mm, time parameters 0.3s, paper speed 3cm/s. EEG reactivity was tested by giving pain (pressing nail bed). Reactive electroencephalographic pattern was defined as a change in the the background(including amplitude or frequency); On the contrary, there is no reactivity. EEG reactivity should be judged by two doctors.3. EEG ClassificationSpecific grading criteria:grade I:dominant α rhythm or with focal δ/θ wave and δ/θ<50%; grade II:focal or one side of δ/θ events (≥50%); grade Ⅲ:diffuse δ or θ activity (≥50%); grade Ⅳ:RAWOD or RAWCS; V grade: diffuse α,θ or δ activity, not reactive; outbreak-suppressed epileptiform activity (universality paroxysmal or more spikes or sharp waves); low-voltage electrical activity (≤20uV); VI level:EEG activity≤10uV.4. Follow-up and outcome measuresRespectively Synek, Wang Xiaomei and ECBER standards for patients admitted for the first EEG examnation. The relationships between the three different EEG grading standards and the prognosis of patients 3 months later after the onset were analyzed. Survival and mortality were outcome measures.5. StatisticsSPSS 13.0 software was used for analysis. Measurement data were presented as mean ± standard deviation (x ± s). Correlation between different EEG gradings and prognosis of patients was evaluated using correlation analysis. We used logistic regression analysis to predict the prognosis of patients with different EEG classifications. P<0.05 was considered statistically significant definition.Results1. The general data of patients114 patients were included from July 2013 to December 2014.13 cases were lost and 10 cases were given up treatments in the 23 cases which were excluded.91 cases were left. A total of 52 cases were male (57%),39 cases (43%) were female.62 cases were ischemic cerebrovascular disease (68%) (including 5 hemorrhagecases after cerebral infarction),29 cases of hemorrhagic cerebrovascular disease (32%). The average age is 62.52 ± 14.69 years and mean GCS score is 8.51 ± 2.24 points. Prognosis:24 patients (26%) died within three months,67 cases of survival (74%).According to Synke grading standards,11 patients could not be included (12%). Wang Xiaomei and ECBER standards can be included all patients.2. Synek, Wang Xiaomei, ECBER classifications and prognosis of patients had significant correlation, correlation coefficients were respectively 0.554,0.580,0.589 (all P<0.001).3. Synek, Wang Xiaomei and ECBER classifications predicting prognosis overall accuracy rate was 78.8%,81.3% and 85.7%, respectively.Conclusion1. ECBER classification and prognosis of patients had significant correlation and this classification can be used as one of the prognostic indicators for patients with disorders of consicousness caused by cerebrovascular diseases.2. Some patients can not be graded (12%) according Synek standards. While Wang Xiaomei and ECBER grading standards can be fully included all the patients.3. The three EEG classifications can well predicte the prognosis of patients with disorders of consciouseness caused by cerebrovascular diseases and can evaluate the brain function effectively.4. Wang Xiaomei and ECBER grading standards are more appropriate to patients with disorders of consciousness caused by cerebrovascular diseases compared to Synek standard.