The Applications Of Color Density Spectral Array For Prognosis Evaluation Of Patients With Coma And Identification Epileptiform Events

Background and purpose : Correct identification prognosis for patients with coma from discriminating epileptic events are common challenges in PICU ward.This study evaluates the quantitative prognosis value of EEG CDSA in PICU coma patients and its effect on epilepsy monitoring.In order to establish a simple and intuitive brain CDSA PICU bedside bed.Methods:1)collect 42 cases of coma in our hospital PICU.In the first three days after admission we grade every pationt according to Glasgow(GCS),then the brain electrical monitor records CDSA not less than 16 hours.follow-up visit ends after 3 months later.pationts are devided into two groups.Good prognosis group(n=21)includs pationts of normal,mild disabled and moderate disabled,and poor prognosis group(n=21)includs pationts of severely disabled,coma or vegetative state,brain death,clinical death.We review CDSA and original EEG data and evalue pationts from the background activities,sleep-wake cycle,sleepstages,fast wave with drug,symmetry of bilateral cerebral hemisphere.Then we compared the receiver operating Characteristic(ROC)curve with GCS score.at last we evaluated effectiveness the two methods in assessing prognosis.2)Twenty seven cases of video EEG records of pediatric EEG room were randomly selected and conversed to4 h/screen CDSA graphics by FFT algorithm.A total of 256 graphics changes differented from background are marked,among them,109 are attacks,147 for reference.Six subjects(including 1 EEG doctors,1 EEG technicians,2 PICU doctors and 2 PICU nurses)were asked to undergo a2-hour training in three aspact of basic principle,episodes CDSA graphics,poor identification of pseudo contents.After trainning they were tested to mark the 256 events separately without checking the original EEG data.to interpretation of 256 attacks,epilepsy mark think attacks.After the test,the consistency of results,the misdiagnosis rate and the reason of errors were analyzed.Results:1)The 42 patients with coma was comprised of 21 girls and21 boys.The average age of them was 89.81±46.39 months.There were21 cases in benign prognosis group and poor prognosis group,respectively.The poor prognosis rate was 50%.There were no differences between Different outcomes between the benign prognosis group and the poor prognosis group refer to their sex,age,original disease(P> 0.05).Our results show that waking sleep stages,sleep stages,drug fast waveand CDSA classification were significantly correlated with the prognosis of coma patients.Logistic regression analysis suggest that the sleep-wake cycle was an independent risk factor for the prognosis of coma patients.The ROC curve was established using poor prognosis as the gold standard,the sleep-wake cycle,and GCS score,indicating that the sleep-wake cycle was more effective than GCS score for prognosis evaluation of coma patients in PICU.2)The average number of correct identification the epileptic events by CDSA graphics was nearly 82.6 in the 6 medical workers.The sensitivity of them was about 75.7% in average,and the 4 events of status epilepticus were all distinguished.The average missing rate of the 6 medical workers was nearly 24.1%,and the number of missing events was 26.3 in average.The specificity of CDSA method was 80%.Based on the result of Kappa consistency test,there were no significant differences of sensitivity and specificity between the6 medical workers(P>0.05).Conclusion : 1)Those factors are close correlated with benign prognosis in coma patients,including the changeable type of CDSA,having a sleep-wake cycle,clear sleep stages,having drug reactivity.Other factors were indicated to poor prognosis or death,including no spectral type and single slow type of CDSA.The disappearance of sleep-wake cycle was an independent risk factors for poor prognosis of coma patients.The effectiveness of sleep-wake cycle was better forprognosis evaluation than that of GCS score.2)The CDSA method has medium sensitivity for detecting epileptic events,and has lower rate of misdiagnosis or missed diagnosis.The medical workers in PICU will easily obtain the screening skill for CDSA after short time training.The CDSA method maybe a promising and feasible tool to monitor patients with coma in PICU.