摘要
目的基于分类回归树(CART)模型对卒中后抑郁(PSD)进行预测研究。方法 320例脑卒中患者随机平均分为训练集和预测集,采用CART模型对训练集的数据进行评估,并提取引起PSD的风险因子,计算各风险因子的贡献率。用预测集对预测结果进行检查,以确定引起脑卒中幸存者PSD障碍的风险因子。结果采用CART模型对训练集的数据进行评估发现,既往脑卒中病史、体质量指数(BMI)、社会支持评定量表(SSRS)、多伦多述情障碍量表(TAS-20)、汉密尔顿焦虑量表(HARS)、NIHSS六项风险因子对引发PSD障碍的贡献率分别为22.82%、15.47%、24.12%、10.50%、16.27%、10.82%。预测集中,PSD患者既往脑卒中病史、BMI、SSRS、TAS-20、HARS、NIHSS六项风险因子均明显高于非PSD患者(χ2/t=4.327,2.40,2.24,5.84,3.29,13.23;均P<0.05)。结论引起脑卒中患者PSD的风险因子包括既往脑卒中病史、BMI、SSRS、TAS-20、HARS、NIHSS六项,对以上风险因子的有效控制有望降低脑卒中患者PSD的发生率。
Objective To predicte the risk factors of post-stroke depression based on classification and regression tree （CART） model. Methods Three hundred and twenty cases of patients with stroke from May 2016 to December 2017 were randomly divided into training set and prediction set on average. CART model was used to evaluate the data of the training set, and extract the risk factor of post-stroke depression factors, and calculate the contribution of each risk factor. Finally, the prediction set was checked the prediction results to determine the risk factors for post-stroke depression. Results The CART regression tree model was used to evaluate the training set data, prior history of stroke, body mass index （ BMI）, social support rating scale （ SSRS）, Toronto alexithymia scale （TAS-20）, Hamilton anxiety rating scale （HARS）, NIHSS scores were risk factors, and the contribution rate were 22.82% , 15.47% , 24.12% , 15.47% , 16.27% and 10.82% respectively. In prediction set, six risk factors as prior history of stroke, BMI, SSRS, Toronto alexithymia scale （ TAS-20 ）, Hamilton anxiety rating scale （ HARS）, NIHSS in PSD group were significantly higher than those in non-PSD group （X2/t = 4. 327, 2.40, 2.24, 5.84, 3.29, 13.23, all P 〈 0. 05）. Conclusion Risk factors for post-stroke depression include prior history of stroke, BMI, SSRS, TAS-20, HARS, NIHSS, and effective control of those risk factors are expected to reduce the chance of post-stroke depression.
作者
钱淑霞
庄建华
岳卫清
吴晓强
步益峰
任莉琼
方莉萍
顾静霞
QIAN Shu-xia;ZHUANG Jian-hua;YUE Wei-qing(Department of Neurology, the Second Affiliated Hospital of Jiaxing University, Jiaxing 314000, China)
出处
《临床神经病学杂志》
CAS
2018年第2期143-145,共3页
Journal of Clinical Neurology
基金
浙江省嘉兴市社会发展领域研究与示范应用项目(2016AY23048)
关键词
回归树
卒中后抑郁
风险因子
预测
regression tree
post-stroke depression
risk factor
prediction
