|
|
|
|
|
| Analysis of related factors of different types of acute cerebral apoplexy patients |
| LUO Song-Ping, ZHANG Yan-Kai, LOU Yun, ZHU Hong, BAI Gao-Wei, YANG Bing, DUAN Bao-Min |
|
|
|
|
Abstract Objective To analyze the risk factors of acute stroke in pre-hospital emergency, establish the risk prediction model of hemorrhagic stroke, and evaluate its predictive ability. Methods This retrospective study included 1536 cases according to the inclusion/exclusion criteria from the acute stroke patients who admitted in Kaifeng City Emergency Center Hospital from January 2013 to December 2015. According to the order of treatment group, 1075 cases（70%）were derived, and 461 cases（30%）were verified. Inclusion criteria: ①acute onset（onset within 6 h）; ②patients diagnosed with stroke. Exclusion criteria: ①low blood sugar; ②cerebrovascular disease sequelae; ③intracranial infection or systemic infection; ④liver and kidney dysfunction, ⑤brain trauma. Single factor screening and logistic multiple regression analysis were carried out on 20 factors which may influence the risk model. Lemeshow -Hosmer test and receiver operating characteristic curve（ROC）were used to fit the goodness of fit test and to determine the model. Results The incidence, headache, vomiting and convulsions were the independent risk factors for hemorrhagic stroke. Derived group Hosmer- Lemeshow test showed that P＝0.496, χ2＝4.380, prediction of hemorrhagic stroke rate and the actual hemorrhagic stroke rate difference was not statistically significant; the model has a good fit. It was better in the derivation group（AUC＝0.875, P＜0.01, 95%CI 0.793～0.965）and validation group（AUC＝0.846, P＜0.01, 95%CI 0.828～0.922）.Conclusion Although the prediction model has certain limitations, it is still able to make a more accurate estimate of hemorrhagic stroke.
|
|
Received: 01 June 2016
|
|
|
|
|
|
|
Address: 1 North 2nd Street,
Zhongguancun, 