预测老年慢性病共病患者抑郁、焦虑发生风险的列线图模型构建

Analysis of factors affecting depression and anxiety in elderly patients with multiple chronic diseases and construction of a nomogram prediction model

目的 分析老年慢性病共病患者抑郁、焦虑的影响因素并构建预测模型。方法 选取该院2021年1月—2023年12月就诊的387例老年慢性病共病患者为研究对象,采用随机抽样的方法按照7:3比例分为建模组(n=271)和验证组(n=116)。采用抑郁自评量表(SDS)和焦虑自评量表(SAS)评价患者抑郁、焦虑发生情况。采用调查问卷收集一般资料;采用多因素Logistic回归分析老年慢性病共病患者抑郁、焦虑的影响因素;预测老年慢性病共病患者抑郁、焦虑风险的列线图模型利用R软件中rms程序包构建,利用rmda程序包绘制临床决策曲线(DCA)评估其临床应用价值;Hosmer-Lemeshow检验、ROC曲线及校准曲线对模型进行验证评估。结果 单因素分析显示,抑郁焦虑组女性、独居、合并慢性病≥4种、低水平社会支持的占比较正常组高(P<0.05)。多因素Logistic回归分析显示,女性、独居、合并慢性病≥4种、低水平社会支持均为老年慢性病共病抑郁、焦虑的影响因素(P<0.05)。Hosmer-Lemeshow检验中建模组χ^(2)=11.135,P=0.133;验证组χ^(2)=5.710,P=0.574。ROC曲线显示,建模组和验证组曲线下面积分别为0.874和0.844,且校准曲线与理想曲线一致性较好。DCA曲线显示,列线图模型临床应用价值较高。结论 老年慢性病共病抑郁、焦虑受性别、居住情况、合并慢性病种类、社会支持度的影响,基于此构建的列线图模型具有较好的预测价值。

Objective To analyze the influencing factors of depression and anxiety in elderly patients with multiple chronic diseases and construct a prediction model.Methods A total of three hundred and eightyseven elderly patients with chronic diseases treated in this hospital from January 2021to December 2023were selected as the research objects,and were divided into modeling group(n=271)and verification group(n=116)by random sampling according to a ratio of 7:3.Self-rating depression scale(SDS)and self-rating anxiety scale(SAS)were used to evaluate the incidence of depression and anxiety of patients.General data were collected by questionnaires.Multivariate Logistic regression was used to analyze the influencing factors of depression and anxiety in elderly patients with chronic diseases.The nomogram model for predicting the risk of depression and anxiety in elderly patients with chronic diseases was constructed by rms program package in R software,and clinical decision curve(DCA)was drawn by rmda program package to evaluate its clinical application value.Hosmer-Lemeshow test,ROC curve and calibration curve were used to validate the model.Results Univariate analysis showed that the proportions of female(χ^(2)=13.899),living alone(χ^(2)=21.066),numbers of chronic diseases≥4(χ^(2)=13.108)and low-level social support(χ^(2)=13.594)in depression and anxiety group were higher than those in normal group(P<0.05).Multivariate Logistic regression analysis showed that being female,living alone,number of chornic diseases≥4,and low level of social support were all influencing factors of depression and anxiety in elderly chronic diseases(P<0.05).In the Hosmer-Lemeshow test,modeling groupχ^(2)=11.135,P=0.133;Verification groupχ^(2)=5.710,P=0.574.ROC curve showed that the area under the line of modeling group and verification suite was 0.874and 0.844,respectively,and the calibration curve was in good agreement with the ideal curve.DCA curve shows that the nomogram model has high clinical application value.Conclusion The depression and anxiety in elderly patients with multiple chronic diseases are influenced by gender,living conditions,numbers of chronic diseases and social support.The nomogram model constructed based on this has good predictive value.