Falls in elderly patients are an important cause of fractures, functional impairment, and mortality. In this paper, a questionnaire was used to collect information on fall history, balance function and sensory function from patients over 65 years of age. In the analysis, the presence or absence of falls was used as a factor, and a corresponding prediction model was constructed using methods such as univariate analysis and regression analysis. This survey found that in the past year, 60% of the patients had fallen, 16.67% had one fall, 33.33% had two falls, and 50% had three or more falls; model specificity is 61.54%, the sensitivity is 71.43%, and the misjudgment is 38.46%. The model has good specificity and sensitivity and a small misjudgment rate; so, the model is more reasonable. This paper selects several sensitivity indices that have a certain impact on the risk of falling and makes a satisfactory forecast, which can provide a theoretical basis for the prevention of the risk of falls in elderly patients during hospitalization.

中文翻译：

---

基于危险因素构建的老年患者住院跌倒预测模型的临床干预效果

老年患者跌倒是骨折、功能障碍和死亡的重要原因。本文采用问卷调查方式收集 65 岁以上患者的跌倒史、平衡功能和感觉功能信息。分析中以有无跌倒为因素，采用单因素分析、回归分析等方法构建相应的预测模型。本次调查发现，在过去一年中，60%的患者跌倒，16.67%跌倒1次，33.33%跌倒2次，50%跌倒3次及以上；模型特异性为61.54%，敏感性为71.43%，误判率为38.46%。该模型具有良好的特异性和敏感性，误判率小；所以，模型更合理。