作者：黄欢,杨佳佳,孙利,殷安康

单位：

225000 江苏省扬州市，扬州大学附属医院急诊科 通信作者：孙利，E-mail：sunli19830824@163.com

单位（英文）：

Department of Emergency, Affiliated Hospital of Yangzhou University, Yangzhou 225000, China Corresponding author: SUN Li, E-mail: sunli19830824@163.com

关键词：

卒中; 创伤后应激障碍; 影响因素分析; 列线图;

关键词（英文）：

Stroke; Post-traumatic stress disorder; Root factors analysis; Nomograms

中图分类号：

DOI：

10.12114/j.issn.1008-5971.2021.00.272

基金项目：

江苏省 2018 年医院管理创新研究课题（JSYGY-2-2018-59）

摘要：

背景通过构建列线图模型可直观、个体化地展示脑卒中患者创伤后应激障碍(PTSD)发生风险,这对改善脑卒中患者预后具有重要的临床意义。目的 分析脑卒中患者发生PTSD的影响因素,并构建其风险预测列线图模型。方法 选取2019年1月至2020年10月扬州大学附属医院急诊科收治的脑卒中患者308例为研究对象,根据患者是否发生PTSD将其分为PTSD组(n=62)和非PTSD组(n=246)。采用研究小组自行设计的一般资料调查问卷收集患者的一般资料,脑卒中患者发生PTSD的影响因素分析采用多因素Logistic回归分析;使用R软件(R 3.6.3)rms程序包构建脑卒中患者发生PTSD风险预测列线图模型;采用Bootstrap法重复抽样100次对列线图模型进行内部验证,采用Hosmer-Lemeshow拟合优度检验和校准曲线评估其一致性;绘制ROC曲线以评估该列线图模型对脑卒中患者发生PTSD的预测价值。结果 308例患者中发生PTSD 62例,PTSD发生率为20.13%(62/308)。PTSD组和非PTSD组患者文化程度、家庭功能、日常生活能力(ADL)、吞咽功能障碍情况及医保类型比较,差异有统计学意义(P <0.05)。多因素Logistic回归分析结果显示,文化程度为大专及以上是脑卒中患者发生PTSD的保护因素,内向人格、ADL重度障碍、吞咽功能障碍是脑卒中患者发生PTSD的危险因素(P <0.05)。基于上述影响因素构建脑卒中患者发生PTSD风险预测列线图模型,内部验证结果显示,该列线图模型的校准曲线趋近于理想曲线,该列线图模型的预测值与实际值比较,差异无统计学意义(P <0.05);ROC曲线分析结果显示,该列线图模型预测脑卒中患者发生PTSD的ROC曲线下面积为0.815〔95%CI(0.765,0.864)〕。结论 文化程度为大专及以上是脑卒中患者发生PTSD的保护因素,内向人格、ADL重度障碍、吞咽功能障碍是脑卒中患者发生PTSD的危险因素;基于上述影响因素构建的脑卒中患者发生PTSD风险预测列线图模型具有良好的一致性与区分度,可以辅助临床医护人员筛选PTSD高风险的脑卒中患者。

英文摘要：

【Abstract】　Background We can intuitively and individually show the risk of post-traumatic stress disorder (PTSD)in stroke patients by constructing nomogram model, which has important clinical significance to improve the prognosis of strokepatients. ObjectiveTo analyze the influencing factors of PTSD in stroke patients and construct its risk prediction nomogrammodel.MethodsFrom January 2019 to October 2020, a total of 308 stroke patients treated in the Department of Emergency,Affiliated Hospital of Yangzhou University were selected as the research objects, and they were divided into PTSD group (n=62)and non PTSD group (n=246) according to whether the patients had PTSD. The general data of patients were collected by thegeneral data questionnaire designed by the research team. The influencing factors of PTSD in stroke patients were analyzed bymultivariate Logistic regression analysis; the nomogram model of PTSD in stroke patients was constructed by R software (R 3.6.3)rms package; Bootstrap method was used to repeatedly sample 100 times to verify the nomogram model internally, and Hosmer-Lemeshowgoodness of fit test and calibration curve were used to evaluate the consistency of the nomogram model; ROC curvewas drawn to evaluate the predictive value of the nomogram model for PTSD in stroke patients. Results The incidence of PTSDin 308 stroke patients was 20.13% (62/308) . There were significant differences in educational level, family function, ability ofdaily living (ADL) , swallowing dysfunction and medical insurance type between PTSD group and non PTSD group (P < 0.05) .Multivariate Logistic regression analysis showed that, college education or above was the protective factor of PTSD in strokepatients, introverted personality, ADL severe disorder and swallowing dysfunction were the risk factors of PTSD in stroke patients(P < 0.05) . The nomogram model for predicting PTSD risk in stroke patients was constructed based on the above influencingfactors. The internal verification results showed that the calibration curve of the nomogram model was close to the ideal curve,and there was no significant difference between the predicted value of the nomogram model and the actual value (P < 0.05) ; theROC curve analysis showed that the under ROC curve area of the nomogram model for predicting PTSD in stroke patients was0.815 [95%CI (0.765, 0.864) ] . Conclusion College education or above is the protective factor of PTSD in stroke patients,and introverted personality, ADL severe disorder and swallowing dysfunction are the risk factors of PTSD in stroke patients.The nomogram model for PTSD risk prediction of stroke patients constructed based on the above influencing factors has goodconsistency and discrimination, and can assist clinical medical staff in screening high-risk PTSD of patients with stroke.

参考文献：