Analysis of risk factors of delayed encephalopathy after acute carbon monoxide poisoning

To analyze the risk factors of delayed encephalopathy in patients with acute carbon monoxide poisoning (DEACMP) and to investigate the clinical significance and clinical value.

The baseline data from 68 patients admitted in the 981st Hospital of Chinese People’s Liberation Army Joint Logistics Support Force were collected, including sex, age, Glasgow Coma Scale (GCS), history of smoking/drinking, history of DEACMP. Laboratory examination records including cranial CT/MRI, cervical vascular color ultrasonography, cardiac color ultrasonography, blood biochemical markers, blood routine, and blood coagulation function were also collected. Additionally, patients were followed-up visited at 1 month and 6 months. Patients were divided into either the DEACMP group or the NDEACMP group according to the occurrence of DEACMP. The risk factors of DEACMP were identified by univariate and logistic regression analyses. The area under the curve (AUC), sensitivity, and specificity of each index were compared by the receiver operating characteristic (ROC) curve.

Among the 68 patients, 13 patients suffered from DEACMP with an incidence of 19.1%. Univariate analysis indicated that there were statistically significant differences in patients’ age, blood glucose, blood glucose/potassium, GCS scores, abnormal cranial CT/MRI, and coma time more than 4 h (P < 0.05). Logistic regression analysis showed that B, SE, Wald, df, P, Exp (B) of GCS score were -0.489, 0.208, 5.55, 1, 0.018 and 0.634, respectively. Overall, GCS score is correlated with DEACMP’s occurrence. ROC curve analysis showed that the sensitivity, specificity, and the AUC of GCS scores were 0.769, 0.873 and 0.843, respectively. The AUC of the indices in descending order are: GCS score, blood glucose, blood glucose/potassium and age.

GCS score, blood glucose, blood glucose/potassium, and age are useful predictive factors for the occurrence of DEACMP. Among these, GCS score is the most significant predictive factor.