Construction and evaluation of an anxiety risk prediction model for the elderly

Objective To construct a risk prediction model and evaluating its predictive forecasting performance to predict the occurrence of anxiety in older adults. Methods The convenience sampling method was used to select 500 eligible elderly people in Hangzhou City as study subjects from September 2023 to February 2024, which were divided into a training set and a validation set. Binary logistic regression was utilized to determine independent risk factors, construct a risk prediction model draw a column-line diagram, and evaluate its prediction effect through the subjects' work characteristic curve, Hosmer-Lemeshow test, and calibration curve. Results A total of 475 older adults were recruited for final analysis. The incidence of anxiety in the elderly people was 36. 00%. The independent risk factors of anxiety were living years (years), average daily exercise time, marital status and communication with children, whether children's occupation is related to medicine, cardiac arrest and other symptoms requiring first aid, the risk of conscious infection with COVID-19, and the degree of anxiety in COVID-19. The area under the ROC curve for the training set was 0. 835, with a specificity of 73. 7% and a sensitivity of 75. 8%. This indicates good predictive performance of the model Hosmer-Lemeshow test χ2 = 9. 653, p = 0. 2902. Conclusion The incidence of anxiety in the elderly was high. The risk prediction value of the constructed nematic chart model is convenient for the early identification of the elderly with anxiety and for taking corresponding intervention measures. (目的 构建老年人产生焦虑的风险预测模型并评价其预测性能, 以预测老年人焦虑的发生。方法 2023年9月—2024年2月, 采用便利抽样法选取浙江省杭州市内年龄≥60周岁、意识清楚、有语言表达能力且同意参与研究的500名老年人作为研究对象, 分为训练集和验证集。利用二元Logistic回归确定独立危险因素, 构建风险预测模型并绘制列线图, 通过受试者工作特征曲线、Hosmer-Lemeshow检验以及校准曲线评价其预测效果。结果 本研究共取得有效样本475例, 老年人焦虑的发生率为36. 00%。焦虑的独立危险因素为现住地生活年限(年)、每日平均运动时间、婚姻状况、与子女的交流频率、子女职业是否与医学相关、心脏骤停等需急救的症状、自觉感染新型冠状病毒的风险、新型冠状病毒感染担忧程度。训练集ROC曲线下面积为0. 835, 特异度为73. 7%, 灵敏度为75. 8%。说明模型的预测性能良好, Hosmer-Lemeshow检验χ2 = 9. 653, P=0. 2902。结论 老年人焦虑的发生率较高, 构建的列线图模型风险预测价值较好, 便于早期识别产生焦虑的老年人并采取相应的干预措施。)