Correlation study between entropy index and Glasgow coma scale score in patients with acute brain injury / 中华危重病急救医学

ABSTRACT

Objective To investigate the correlation between the entropy index and Glasgow coma scale (GCS) score in patients with acute brain injury (ABI), and to explore the feasibility of using entropy index to assess the extent of brain damage. Methods A prospective study was conducted. 101 patients with ABI admitted to intensive care unit (ICU) of the First People's Hospital of Changshou from May 2015 to May 2017 were enrolled. The values of reaction entropy (RE) and state entropy (SE) within 24 hours after ICU admission were collected, and the GCS score was recorded. A linear regression analysis of RE and SE values and GCS scores was conducted. Receiver operating characteristic (ROC) curve was plotted, and the value of entropy index in evaluating the severity of ABI was analyzed. Results 101 patients with ABI were finally enrolled in the analysis, with mild in 22 patients (GCS score 13-15), moderate in 35 patients (GCS score 9-12), and severe in 44 patients (GCS score 3-8). With the increase severity of brain damage, the values of RE and SE, and GCS score were gradually decreased (RE values in patients with mild, moderate and severe brain injury were 89.77±5.28, 75.37±4.86 and 50.61±9.89, SE values were 79.91±6.11, 65.20±5.44 and 40.68±9.41, and GCS scores were 14.00±0.82, 10.43±1.15 and 5.32±1.75). There was no significant difference in entropy index or GCS score among patients with different degrees of brain injury (all P < 0.01). Correlation analysis showed that the RE value was positively correlated with the GCS score (r2= 0.974, P = 0.000), SE and GCS scores also showed linear positive correlation (r2= 0.962, P = 0.000). ROC curve analysis showed that the area under ROC curve (AUC) of RE for diagnosis of ABI in patients with severe brain injury was 0.914, when the optimal cut-off value was 53.5, the Youden index was maximum (0.753), the sensitivity was 91.2%, and the specificity was 84.1%. The AUC of SE was 0.916, when the optimal cut-off value was 41.5, the Youden index was maximum (0.730), the sensitivity was 91.2%, and the specificity was 81.8%. The AUC of RE for diagnosis of mild brain injury was 0.884, when the optimal cut-off value was 69.5, the Youden index was maximum (0.651), the sensitivity was 72.7%, and the specificity was 92.4%. The AUC of SE was 0.897, when the optimal cut-off value was 58.5, the Youden index was maximum (0.676), the sensitivity was 72.7%, and the specificity was 94.9%. Conclusions The entropy index RE and SE values of ABI patients were gradually decreased with the severity of brain damage, which was positively correlated with GCS score. As an objective, continuous and real-time method, entropy index could be used as a new tool for the evaluation and monitoring of ABI.