摘要
目的研究进展型脑梗死的危险因素及对脑梗死分型进行预测的可行性。方法回顾性分析2005年1月至2006年12月某医院收治的200例脑梗死患者的临床资料,其中进展型脑梗死69例,非进展型脑梗死(对照组)131例,研究其人口学资料、病史、神经功能缺损程度评分、体检指标、头颅CT检查情况等方面共20个因素,并用logistic回归法进行相关分析研究。结果Logistic回归分析结果显示,进入模型的各因素按其OR值大小排列顺序为脑梗死类型(OR=40.914)、短暂性脑缺血发作史(OR=35.804)、WBC升高(OR=3.151)、入院时高血糖(OR=3.020)、脑梗死部位总数(OR=2.456)、血钙水平(OR=0.031)及入院时神经功能缺损程度评分(OR=0.908);用logistic回归模型进行预测,显示对脑梗死分型的平均准确率为83.4%,对脑梗死进展型的准确率为73.5%。结论大面积梗死、有短暂性脑缺血发作史、WBC升高、入院时高血糖、多梗死部位、低血钙及入院时神经功能缺损程度评分低可能为进展型脑梗死的危险因素;用logistic回归模型对脑梗死进行分型的临床应用价值有待进一步研究。
Objective To identify risk factors and feasibility in classification of progressive ischemic stroke. Methods Clinical records of 200 cases of ischemic stroke including 69 progressive and 131 non-progressive ischemic stroke patients were reviewed in a hospital during January, 2005 to December, 2006. 20 factors including demographic data, disease history, neurological deficits score, physical examination and head CT results et al were analyzed retrospectively. Logistic regression analysis was performed to identify relationship with these factors. Results Logistic regression indicated factors entered model and ranked descendently by OR value were infarction category （OR = 40. 914）, TIA medical record （OR = 35. 804） , higher blood leucocytes （OR = 3.151） and hyperglycemia （OR = 3. 020） , number of infarction sites （OR = 2.456）, level of blood calcium（OR =0. 031） and neurological deficits score at time of hospital admission（OR = 0.908）. The positive correction rate of ischemic stroke classified by logistic regression model was 83.4%, and the progressive ischemic stroke was 73.5%. Conclusion Larger infarction area, TIA medical record, higher blood leucocytes and hyperglycemia, multiple infarction, hypocalcemia and lower neurological deficits score were the main risk factors of progressive ischemic stroke and application of logistic regression model to classify ischemic stroke required further study.
出处
《华南预防医学》
2007年第3期5-7,共3页
South China Journal of Preventive Medicine
