目的 探讨恶性血液病患者发生跌倒的危险因素，构建风险预测模型，为预防跌倒提供参考。方法 2021年6月至2022年6月，采用便利抽样法选取山东省某三级甲等医院住院的恶性血液病患者411例为研究对象，分析跌倒的危险因素，并建立跌倒风险预测模型，采用受试者工作特征（receiver operating characteristic，ROC）曲线下面积（area under curve，AUC），评估风险预测模型区分度与校准度。结果 调查期间发生跌倒事件9例，近乎跌倒事件54例。恶性血液病患者跌倒风险预测模型纳入近3个月跌倒史、疾病诊断、腹泻、疼痛、骨髓抑制天数等5个预测因子。内部验证显示AUC为0.840[95%CI（0.880~0.732）];外部验证显示AUC为0.802[95%CI(0.866~0.727)]。结论 构建的恶性血液病患者发生跌倒的危险预测模型预测效果好，可帮助医护人员识别跌倒高风险人群。

Objective To explore the risk factors for falls in patients with hematologic malignancies，to construct a risk prediction model，and to provide a reference for fall prevention.Methods From June 2021 to June 2022，411 patients with hematologic malignancies hospitalized in a tertiary A hospital in Shandong Province were selected by the convenience sampling method.The risk factors of falls were analyzed and a fall risk prediction model was established.The area under the receiver operating characteristic (ROC) area under curve (AUC) was used to evaluate the discrimination and calibration of the risk prediction model.Results During the investigation period，there were 9 cases of falls and 54 cases of near falls.The fall risk prediction model for patients with patients with hematologic malignancies incorporated 5 predictive factors：fall history in the past 3 months，disease diagnosis，diarrhea，pain，and the number of days of bone marrow suppression.Internal validation showed that the AUC was 0.840[95%CI(0.880-0.732)].External validation showed that the AUC was 0.802[95%CI(0.866-0.727)].Conclusions The constructed risk prediction model for falls in patients with hematologic malignancies demonstrates a good predictive effect and can help medical staff identify high-risk groups for falls.

R47；R473.55