Analysis Of Heart Rate Variability In Patients With Brain Death

BackgroundLegal definition of brain death refers to the irreversible loss of whole brain function including the brainstem.With the loss of brainstem function,the function of the autonomic nervous system will definitely be affected.At present,the commonly used clinical autonomic nervous function evaluation methods mainly rely on clinical evaluation,and the objective evaluation means are limited.The evaluation system of brain death does not include the evaluation of autonomic nervous system.Heart rate variability(HRV)is considered a rapid,reliable,and non-invasive method for the assessment of the autonomic nervous system.HRV analysis based on 24-hour Holter monitoring can be used to assess autonomic function and predict long-term prognosis in critically ill patients,but the predictive value for the diagnosis of brain death in neurocritically ill patients are still unclear.ObjectiveThis study aims to evaluate the sensitivity,specificity and accuracy of HRV analysis based on 24-hour Holter monitoring as a diagnosis tool for brain death in neurocritical ill patients.MethodsThis study is a single-center,retrospective,case-control study.Patients admitted to the Neurocritical Care Unit of Nanfang Hospital from December 2019 to June 2021 were included to assess cardiac autonomic function.HRV analysis based on 24-hour Holter monitoring to assess cardiac autonomic function in patients with brain death,deep coma(Glasgow Coma Scale,GCS=3),and mild disturbance of consciousness(Glasgow Coma Scale,GCS≥8).ResultsMean heart rate and minimum heart rate in patients of brain death were significantly different compared with patients of deep and mild disturbance of consciousness coma(P<0.05).In time-domain analysis,mean of the standard deviation of normal to normal intervals for each 5-min segment,SDNNI was significantly lower in patients of brain death than in patients of deep coma(P=0.016).In frequencydomain analysis,total power(P-0.004),the very low frequency(VLF)power(P=0.001),the low frequency(LF)power(P=0.007),normalized units of the low frequency(LFnu)(P=0.001),and LF/HF(P<0.001)were lower in patients of brain death than patients in deep coma.In addition,normalized units of high frequency(HFnu)(P<0.001)was higher in patients of brain death than in patients of deep coma,indicating cardiac autonomic dysfunction especially in sympathetic system.Binomial logistic regression was performed by excluding parameters without significant differences and backward likely ratio(LR),to establish a regression model(χ2=45.016;P<0.001)for prediction of brain death.The regression equation for the prediction was K=-3.978-0.003~*VLF+0.002~*LF+0.083~*HFnu.ROC curve revealed that the optimal cut-off value of the predicted probability for predicting brain death was 0.57(sensitivity 76.47%,specificity 91.23%,accuracy 85.71%),which gives an AUC=0.894,95%CI 0.826-0.962.ConclusionThis study showed that cardiac autonomic function was disturbed in patients of brain death.Our study indicated that long-term HRV analysis based on 24h Holter is a supplemental method for assessment of brain death.