| BackgroundDisorders of consciousness are caused by a variety of pathogenic factors which injure the brain function, representing decrease of wakeful arousal state and impairment to the content of awareness. Prognosis of unconscious patients with different etiologies varies. The mortality in hospital reported in different clinical studies ranges from0%to90%. As reported in other countries, an average of4-10%of the patients admitted to emergency department had unconsciousness. The time of patients in coma lasted more than3days, the cost of quality-adjusted life years (QALYs) would reach about87000-140000dollars a year. The coma patients survived but with long-term coma, vegetative state, highly life-dependent usually obtained poor prognosis. The clinical manifestation of the unconscious patient changes but the treatment is timely. In general, the comatose patients take up more medical resource. Therefore, accurate assessment of neurological function to predict prognoses is significant for screening patients, clinical decision making, and rational allocation of medical resource.Neural electrophysiological examinations had been proved the predictive value for the prognosis of patients with disorders of consciousness.40Hz auditory steady state evoked potential (40Hz aSSEP) first described by Galambos et al. in1981, is elicited by sounds delivered repeatedly at rate of40c/s, and the respond waveform approximates a40-Hz sinusoid. The40Hz aSSEP is a kind of middle latency response. Some previous studies had indicated that the midbrain, thalamus and the temporal lobe play a fundamental role in generating40Hz aSSEP. While this composite response was considered to be induced according to multiple generator theory at the present times. The generators of the40Hz aSSEP were comprehensive. Thus, this responce could evaluate the global brain function. Several clinical researches had declared40Hz aSSEP could be a feasible method to predict outcomes of patients with massive cerebral infarction or brainstem stroke. Animal studies were conducted to point out the response thresholds of40Hz aSSEP increase following the prolongation of ischemia hypoxia. Therefore, we had a hypothesis that not only the40Hz aSSEP can assess the whole brain function, but also its thresholds are able to reflect the degree of brain damage.Our target was to analyze the relationship between40Hz aSSEP and its classification of thresholds with the prognosis of unconscious patients. In order to confirm the predictive value, we performed a compare of the prediction capability between40Hz aSSEP with Acute Physiology and Chronic Health Evaluation Ⅱ (APAHCEⅡ) and Glasgow Coma Scale (GCS).Subjects and methodsThis observational study was approved by Medical Ethics Committee of Nanfang Hospital, an affiliated teaching hospital of Southern Medical University. This study was registered on the Chinese Clinical Trial Registration. The registration number was:ChiCTR-DDT-13003682.1. SubjectsSubjects were recruited from the neurology intensive care units (NICU) of Nanfang Hospital between July2012and July2013. Inclusion criteria were (1) meeting the criteria of admission to NICU;(2) acute onset of unconsciousness;(3) age>18years;(4) no hearing disorder history. Exclusion criteria were (1) abandoning treatment;(2) failed to follow-up;(3) sedation within24h before evaluation;(4) apparent interference artifact on40Hz aSSEP. APACHE Ⅱand GCS were recorded within24h after admissions.40Hz aSSEP was evaluated within72h after admissions. General information of the patients such as gender, age, hypertension, diabetes, intubation and ventilation were collected too.2. The evaluation of40Hz aSSEPThe evaluation of40Hz aSSEP was measured with Nicolet Evoked Potential Equipment (Viking Quest, N/SM165597, USA). Acoustic stimuli pips were delivered through inserted earphones (TIP-300, USA) at frequency of500Hz, rate of39.1/s. Filter was set at5-100Hz. Time based was10ms. Recording electrode, grounding electrode and reference electrode was placed on vertex (Cz), forehead (FPz) and homo-lateral earlobe of stimuli, respectively. Repetitions were500-1000times twice on each side.3. The classification of40Hz aSSEPThe stimulus was pip pure tone. The initial stimulus intensity was90dB nHL with60dB nHL contra-lateral white noise masking. We adopted a "terraced search" for searching the thresholds of40Hz aSSEP. If an identifiable response was achieved, the evaluation was repeated with stimulus intensity reduced by10dB gradually, meanwhile, for obtaining an appropriate signal-to-noise ratio; we made sure that the intensity of the contra-lateral white noise is at least15dB lower than the stimulus by decreasing its intensity. When a stable response could no longer obtain or the intensity of the stimulus was decreased to20dB, the evaluation was completed. The threshold was defined as the minimum stimulus intensity inducing a clear respond, ranging from20to90dB nHL in this study. We classified the40Hz aSSEP into five grades according to bilateral threshold values. Grade0:bilateral response abolished at90dB; Grade1:one response preserved at60-90dB with the other abolished; Grade2:bilateral response preserved at60-90dB; Grade3:one response preserved at60-90dB with the other preserved at20-50dB; Grade4:bilateral response preserved at20-50dB.4. Follow-up and the main outcome measuresPatients were followed up with Glasgow Outcome Scores (GOS) recorded in the first month and in third month after the patients’admission. Score1was defined as mortality and score1,2and3were considered as poor prognosis. Considering the time of follow-up and outcomes, we defined our main outcome measures as short term (the first month after admission) mortality; long term (the third month after admission) mortality; short term poor prognosis and long term poor prognosis.5. Statistical analysisStatistical analyses were performed by using SPSS software (version19.0, SPSS Inc., IBM Company, USA). Measurement data was described with mean±standard deviation and categorical data was described with rate. Correlation ship was analyzed with Pearson or Spearman correlation analysis. Hosmer-Lemeshow goodness-of-fit test was applied to measure the calibration. Kruskal-Wallis H test was applied on ranked data. Nemenyi test was performed to reveal difference in grades of40Hz aSSEP. Area under Receiver Operating Characteristic curve (AUC) was employed to assess prediction accuracy. Z test was applied to pair comparion of the AUC.Statistical significance was considered as P<0.05.Results1. The general information of patientsA total of106patients meet the inclusion criterion. Fifteen patients were excluded:10for apparent interference artifact on tracing,4for abandoned treatment and one for failing to follow up. In the remaining91patients, there were56males (61.5%) and35females (38.5%). The causes of unconsciousness were cerebrovascular disease57(62.6%), intracranial infection15(16.5%), metabolic encephalopathy8(8.9%), drug and poison6(6.7%), and hypoxic-ischemic encephalopathy5(5.6%). Age was range of20-88years,55.1±19.2years; GCS was range of3-14,7.6±3.2; APACHE II was range of5-32,18.8±6.7. Patients with hypertension51(56.0%), diabetes20(22.0%), intubation65(71.4%), and ventilator51(56.0%). In the classification of40Hz aSSEP, there were13patients in Grade0(14.3%),14patients in Grade1(15.4%),27patients in Grade2(29.6%),16patients in Grade3(17.6%) and21patients in Grade4(23.1%). Mortality after admission:28Patients (30.8%) death and63patients (69.2%) survived in the first month;31patients (34.1%) death and60patients (65.9%) survived in the third month. Prognosis after admission:poor prognosis67patients (73.6%) and favorable prognosis24patients (26.4%) in the first month; poor prognosis61patients (67.0%) and favorable prognosis30patients (33.0%) in the third month.2. Correlation ship between40Hz aSSEP, APACHE Ⅱ and GCS with outcomes40Hz aSSEP, APACHE Ⅱ and GCS had a significant relationship with the outcomes. With mortality,40Hz aSSEP had nice correlation ship with short term or long term mortality. The r value was-0.535and-0.566, respectively. Besides, this correlation had no difference with APACHE Ⅱ (r=0.548,0.550). However,40Hz aSSEP and APACHE Ⅱ had statistic significance difference comparing with GCS (r=-0.368,-0.368), respectively. Both with short term and long term poor prognosis,40Hz aSSEP (r=-0.593,-0.622) had stronger correlations then APACHE Ⅱ (r=0.351,0.597) and GCS (r=-0.398,-0.393), respectively. In our study, the patients obtained poorer outcomes following the decrease of grade of the40Hz aSSEP. In addition,40Hz aSSEP had correlation with APACHE Ⅱ (r=-0.625, P=0.000) and with GCS (r=0.583, P=0.000).3. The goodness-of-fit test of40Hz aSSEP, APACHE Ⅱ，and GCS with outcomes40Hz aSSEP had the best calibration for short term mortality, long term mortality, short term poor prognosis, and long term poor prognosis, respectively. The P values of40Hz aSSEP were0.601,0.242,0.922, and0.913, respectively. The P values of APACHE Ⅱ were0.496,0.161,0.901, and0.743, respectively. Although GCS had statistically significant P values (P=0.146,0.052,0.145, and0.211), it had poor calibration for outcomes. The calibration for mortality of these three methods was higher than the calibration for poor prognosis.4. Prediction capacity of40Hz aSSEP, APACHE Ⅱ and GCSThe analysis of the ROCs had indicated that for outcomes prediction (mortality or poor prognosis),40Hz aSSEP had favorable prediction capacity (AUCs all greater than0.8). Compared the AUCs with APACHE Ⅱ,40Hz aSSEP had no difference for short term and long term mortality while had significant for short term and long term poor prognosis (P=0.006,0.010). While compared the AUC for short term mortality, long term mortality, short term poor prognosis, and long term prognosis with GCS,40Hz aSSEP had a higher prediction accuracy, respectively (P=0.041,0.021,0.030, and0.014). For short term and long term mortality, APACHE Ⅱ was superior to GCS (P=0.017,0.036), restectively. But there was no difference between them for short term or long term poor prognosis.5. Differences between grades of40Hz aSSEPThe significant distinction of outcomes between the grades of40Hz aSSEP was revealed by Kruskal-Wallis H test. The P values of short term and long term outcomes (mortality or poor prognosis) were all0.000. Besides, the APACHE Ⅱ, GCS, intubation, and ventilator had statistic difference between grades(P all=0.000) while the other variables such as age, gender, hypertension, and diabetes had no significant difference.The short term mortality of patients in Grade0was84.6%, and the long term mortality was92.3%. The short term mortality of Grade0had significant difference compared with Grade2,3, and4. The P values were0.030,0.000, and0.000, respectively. Similarity, the long term mortality was different between Grade0with2,3, and4, respectively. The P values were0.036,0.000, and0.000. The overall poor prognosis of Grade0,1, and2was above the average of85.2%. Moreover, the short term and long term poor prognosis of Grade0were all100%; the short term and long term poor prognosis of Grade1were100%and92.9%. Patients in Grade4obtained the most favorable outcomes (long term mortality was19.0%). Grade0,1, and2had significant difference compared the short term or long term poor prognosis with Grade4(P all=0.000).Compared in each grades of40Hz aSSEP, patients in Grade0,1, and2had no significant difference between the short term with the long term outcomes (mortality and poor prognosis). Compared patients in Grade3and4, no difference were found between their short term with long term mortality. However, patients achieved a better long term prognosis when compared with their short term poor prognosis (P=0.023in Grade3, P=0.007in Grade4).Conclusion 1.40Hz aSSEP had significant correlation ship with the short term mortality, long term mortality, short term poor prognosis, and long term poor prognosis. The strength of correlation of40Hz aSSEP was similar to APACHE Ⅱ but stronger than GCS. Patients in lower grades of40Hz aSSEP obtained poorer outcomes.2. Although had significant calibration for short term and long term outcomes (mortality and poor prognosis),40Hz aSSEP was a bit better then APAHCHEⅡ. The calibration for outcomes of GCS was poor. The calibration for poor prognosis of40Hz aSSEP, APAHCHE Ⅱ, and GCS was better than the calibration for mortality, respectively.3.40Hz aSSEP and APACHE Ⅱ had favorable prediction capacity for short term and long term mortality. No diffidence was found between them. The prediction accuracy of them was significant better than GCS, respectively. That indicated the40Hz aSSEP could be a stable and reliable predictor for mortality as APACHE Ⅱ. For short term and long term poor prognosis,40Hz aSSEP was superior to APAHCHE Ⅱ and GCS. That revealed40Hz aSSEP had a better distinction in poor prognosis.4. Patients in different grades had distinguished outcomes. Patients in Grade0were in extremely highly risk to death (short term and long term mortality was84.6%and100%). Patients in Grade1and2were in highly risk to death (general mortality was about35%to45%). Patients in Grade3and4had the lowest mortality (less than10%). Patients in Grade0,1, and2were more likely to obtain a long term poor prognosis (the rate was85.2%or higher). About50%of patients in Grade3achieved long term poor prognosis. Patients in Grade4had the best outcomes (long term poor prognosis is less than19%).5. After comparing the short term and long term outcomes in each grades of40Hz aSSEP, patients in Grade0,1, and2not only had higher mortality but also the vast majority of the survivors remained in long term severity disability. While patients in Grade3and4had lower mortality and most of the survivors could have a better prognosis after an appropriate treatment.6. The40Hz aSSEP could be a feasible and reliable predictor for short term and long term mortality, short term and long term poor prognosis in unconscious patients... |
PDF Full Text Request