|
|
|
| Optimum Predictive Value of the Risk Factors of Progressive Stroke According to the ROC Curve |
| Department of Neurology,TongRen Hospital Affiliated to Shangshi Jiaotong University School of Medicine,Shanghai 200336, China |
|
|
|
|
Abstract 【Objective】 To investigate the predictors of progressive stroke and to obtain the optimum predictive value of progressive stroke. 【Methods】Patients admitted consecutively with acute stroke within 24 hours not taking rtPA between 2014.1.12014.12.31 were included. Progressive stroke was defined as an increase in the National Institutes of Health Stroke Scale≥1 point. Patients were divided into a progressive group (PS group) and a nonprogressive group (nonPS group). The differences in demographics,history, imaging data (MRA), and related risk factors (blood sugar, lipids and pressure) were compared between the two groups in order to screen for the predictors of progressive stroke. Independent predictors for progressive stroke were analyzed using stepwise logistic regression. The receiver operating characteristic (ROC) curve was drawn to evaluate the optimum predictive value of progressive stroke. 【Results】A total of 93 patients with acute infarction within 24 hours after onset not taking rtPA were enrolled. 36 were in PS group, and 57 of them were in nonPS group. Univariate analysis showed that there were no significant differences in the proportions of sexes and history of previous hypertension, diabetes mellitus, hypercholesterolemia, or smoking between two groups. No difference was found in , blood pressure, plasma glucose, HbA1C, Triglyceride, or HDLc between two groups. Serum total cholesterol (TC) and LDLc of the PS group were significantly higher than those of the nonPS group(P<0.05). The proportions of LAA of the PS group were higher than those of the nonPS group, without significant meaning. Multivariate stepwise logistic regression analysis showed that LDLc was the independent predictor for PS. A ROC (receiver operating characteristic) curve provided a cutpoint of 2.45 mmol/L,which the optimum prediction effect could be achieved. As LDL was stratified into high LDL (LDL >2.45 mmol/L) and low LDL (LDL<2.45 mmol/L), multivariate stepwise logistic regression analysis showed that high LDL was still the independent predictor for progressive stroke. 【Conclusion】LDLc is the independent predictor for progressive stroke. Stratified by the optimum predictive value, high LDLc was still independent predictor for progressive stroke. Therefore clinically LDLc could be a useful indicator to evaluate PS prognosis.
|
|
Received: 10 September 2015
|
|
|
|
|
|
|
2016, Vol. 33 
