论文部分内容阅读
目的:构建急性缺血性脑卒中（AIS）早期神经功能恶化（END）的模型，并评价该模型的预测价值。方法:选择2018年2月至2020年12月在应急总医院就诊的263例AIS患者作为研究对象。根据是否发生END将其分为END组（80例）和非END组（183例）。用Logistic回归分析AIS并发END的独立危险因素，并构建上述因素的预测模型。用ROC曲线、校正曲线和决策曲线分析法评价该模型价值。结果:END组的收缩压、舒张压、空腹血糖、总胆固醇、美国国立卫生研究院卒中量表（NIHSS）评分和入院至溶栓时间（DNT）分别为（144.66 ± 18.08） mmHg（1 mmHg=0.133 kPa）、（84.04 ± 8.47） mmHg、（6.15 ± 1.00） mmol/L、（4.82 ± 1.08） mmol/L、（14.90 ± 4.95）分、（4.58 ± 1.02） h，均高于非END组的（132.32 ± 15.53） mmHg、（81.19 ± 11.00） mmHg、（5.53 ± 0.98） mmol/L、 （4.51 ± 1.15） mmol/L、（11.37 ± 3.32）分、（3.22 ± 1.08） h，END组的HDL-C水平为（1.09 ± 0.07） mmol/L，低于非END组的（1.15 ± 0.08） mmol/L，差异有统计学意义（n t值为2.07 ~ 9.53，均n P<0.05）。Logistic回归分析显示，空腹血糖（n OR=2.25，95% n CI 1.50~3.40，n P<0.05）、收缩压（n OR=1.04，95% n CI 1.02~1.07，n P<0.05）、NIHSS评分（n OR=1.26，95% n CI 1.14~1.39，n P<0.05）和DNT（n OR=3.44，95% n CI 2.30~5.14，n P<0.05）均是AIS并发END的独立危险因素。由空腹血糖、收缩压、NIHSS评分和DNT构成的模型X诊断AIS并发END的ROC曲线下面积、灵敏度、特异度分别为0.90（95%n CI 0.87 ~ 0.94）、90.00%、77.6%，模型X的校准曲线与理想曲线重合度高。n 结论:由空腹血糖、收缩压、NIHSS评分和DNT构成的模型X对于END的预判有较高的应用价值，可辅助临床做出更好的决策。“,”Objective:To construct a model of early neurological deterioration (END) in acute ischemic stroke (AIS), and to evaluate the predictive value of this model.Methods:From February 2018 to December 2020, 263 cases of AIS patients admitted to Response General Hospital were selected as research objects. According to whether END occurs, they were divided into END group (80 cases) and non-END group (183 cases). Logistic regression was used to analyze the independent risk factors of AIS concurrent END, and the predictive model of the above factors was constructed. The value of the model was evaluated by receiver operating characteristic (ROC) curve, correction curve and decision curve analysis (DCA).Results:Systolic blood pressure, diastolic blood pressure, fasting plasma glucose, total cholesterol, National Institutes of Health Stroke Scale (NIHSS) and door to needle time (DNT) in the END group were (144.66 ± 18.08) mmHg(1 mmHg=0.133 kPa), (84.04 ± 8.47) mmHg, (6.15 ± 1.00) mmol/L, (4.82 ± 1.08) mmol/L, (14.90 ± 4.95) points, (4.58 ± 1.02) h, all higher than those in the non-END group (132.32 ± 15.53) mmHg, (81.19 ± 11.00) mmHg, (5.53 ± 0.98) mmol/L, (4.51 ± 1.15) mmol/L, (11.37 ± 3.32) points, (3.22 ± 1.08) h, and the level of high density lipoprotein cholesterol was (1.09 ± 0.07) mmol/L in the END group,lower than that in the non-END group (1.15 ± 0.08) mmol/L, the differences were statistically significant (n t values were 2.07-9.53, all n P<0.05). Logistic regression analysis showed that fasting plasma glucose (n OR=2.25, 95% n CI 1.50-3.40, n P<0.05), systolic blood pressure (n OR=1.04, 95% n CI 1.02-1.07, n P<0.05), NIHSS score (n OR=1.26, 95% n CI 1.14-1.39, n P<0.05) and DNT (n OR=3.44, 95% n CI 2.30-5.14, n P<0.05) were independent risk factors for AIS concurrent END. The area under ROC curve, sensitivity and specificity of the model X composed of fasting plasma glucose, systolic blood pressure, NIHSS scores and DNT for diagnosing AIS concurrent END were 0.90 (95%n CI 0.87-0.94), 90.00% and 77.60% respectively. The calibration curve of model X had a high coincidence with the ideal curve.n Conclusions:The model X composed of fasting plasma glucose, systolic blood pressure, NIHSS score and DNT has high application value for the prediction of END, which can assist the clinic to make better decision.