老年脑卒中后偏瘫患者康复训练效果影响因素分析及列线图预测模型构建

Analysis of Influencing Factors of Rehabilitation Training Effect and Construction of Nomogram Prediction Model in Elderly Post-Stroke Hemiplegia Patients

目的:分析老年脑卒中后偏瘫患者康复训练效果的影响因素,并构建预测康复训练效果的列线图模型。方法:回顾性收集2017年7月至2021年6月桂平市人民医院接受康复训练的286例老年急性脑卒中后偏瘫患者的相关临床资料,按照7∶3的比例随机分为建模组(n=200)与验证组(n=86),利用建模组数据构建风险预测列线图模型,利用验证组数据评价模型性能。两组均根据康复训练效果分为效果良好组、效果不佳组。采用单因素及多因素Logistic回归分析康复训练效果不佳的危险因素,R3.6.3软件构建预测康复效果不佳的列线图模型,并验证该列线图模型预测康复效果不佳的效果。结果:286例偏瘫患者中康复效果不佳占42.66%(122/286)。效果不佳组患者年龄、男性比例、饮酒比例、高血压比例、康复训练开始时间(>2周)以及居住环境为农村比例均显著高于效果良好组,差异有显著性(P<0.05)。通过多因素Logistic回归模型发现,居住环境、年龄、高血压、康复训练开始时间>2周、饮酒是影响偏瘫患者康复效果不佳的危险因素(P<0.05)。建模组的受试者工作特征曲线(ROC曲线)下面积为0.801,验证组的曲线下面积为0.851;建模组与验证组的校准曲线斜率均接近1。结论:基于年龄、饮酒、高血压、康复训练开始时间以及居住环境构建的预测老年脑卒中后偏瘫患者康复训练效果不佳的列线图模型,可较好地个体化预测康复训练效果不佳的风险,对及早干预康复训练有一定指导意义。

Objective:To analyze the influencing factors of rehabilitation training effect in elderly post-stroke hemiplegia patients,and to construct a nomogram model to predict the effect of rehabilitation training.Methods:The clinical data of 286 elderly patients with hemiplegia after acute stroke who received rehabilitation training in Guiping People's Hospital from July 2017 to June 2021 were retrospectively collected and randomly divided into modeling group(n=200)and validation group(n=86)with the ratio of 7∶3.The data of the modeling group were used to construct a risk prediction nomogram model,and the data of the validation group were used to evaluate the performance of the model.The two groups were divided into a good effect group and a poor effect group according to the effect of rehabilitation training.Univariate and multivariate Logistic regression analyses were used to identify the risk factors for poor rehabilitation training effect,R3.6.3 software was applied to build a nomogram model for predicting poor rehabilitation effect,and the effect of the nomogram model in predicting poor rehabilitation effect was verified.Results:Among the 286 hemiplegic patients,42.66%(122/286)had poor rehabilitation effect.The age,male proportion,alcohol consumption proportion,hypertension proportion,rehabilitation training start time>2 weeks,and rural living environment proportion of patients in the poor effect group were significantly higher than those in the good effect group,with significant differences(P<0.05).Through multivariate Logistic regression model,it was found that living environment,age,hypertension,rehabilitation training start time>2 weeks,and alcohol consumption were the risk factors for poor rehabilitation effect of hemiplegic patients(P<0.05).The area under the ROC of the modeling group was 0.801,and the area under the curve of the validation group was 0.851;the slopes of the calibration curves of both the modeling group and the validation group were close to 1.Conclusion:Based on age,drinking,high blood pressure,rehabilitation training start time,and living environment,the nomogram model for predicting poor rehabilitation training effect in elderly post-stroke hemiplegia patients can better predict poor rehabilitation training effect individually,and it has certain guiding significance for early intervention in rehabilitation training.