[关键词]
[摘要]
目的 构建中老年急性白血病患者医院感染预测模型并验证,以期为减少该类人员并发医院感染风险提供参考。 方法 回顾性收集某综合性三级甲等医院2018年6月至2021年6月收治的≥45岁的急性白血病患者613例的病历资料,将纳入研究患者随机抽取70%(429例)作为训练集构建预测模型,30%(184例)进行内部验证。采用受试者工作特征 (receiver operating characteristic,ROC)曲线下面积评估模型临床效能,并将预测模型以公式和列线图形式呈现。 结果 613例患者中有100例发生了医院感染,感染率约为16.3%。不同一般情况(年龄、性别、住院时长等)、实验室指标水平(中性粒细胞绝对值、粒细胞缺乏持续时间等)及治疗措施(化疗、合并激素治疗等)的中老年急性白血病患者在是否发生院内感染上的差异均有统计学意义(均P<0.05),住院时长、化疗、粒细胞缺乏持续时间、中心静脉置管、侵入性操作等是独立危险因素(均P<0.05),而年龄和经验性抗感染治疗是保护性因素(均P<0.05)。预测模型的ROC曲线下面积为0.8997,内部验证的ROC曲线下面积为0.8704。 结论 构建的模型可以作为有效采取措施预防中老年急性白血病患者发生医院感染的参考依据,值得推广。
[Key word]
[Abstract]
Objective To construct and analyze of nosocomial infection prediction model in middle-aged and elderly patients with acute leukemia(AL),and to provide reference for reducing the risk of nosocomial infection. Methods The medical records of 613 patients with acute leukemia aged ≥ 45 years from June 2018 to June 2021 in a general tertiary A hospital were retrospectively collected.70% (429 cases) of the included patients were randomly selected as the training set to construct the prediction model,and 30% (184 cases) were internally validated.The area under the receiver operating characteristic(ROC) curve was used to assess the model clinical efficacy,and the prediction model was presented as a formula and nomogram. Results 100 of the 613 patients had nosocomial infection during hospitalization,the infection rate was about 16.3%.There were statistically significant differences in nosocomial infection among middle-aged and elderly AL patients with different general conditions (age,gender,length of hospital stay,etc.),laboratory indexes (absolute neutrophil value,duration of granulocyte deficiency,etc.) and therapeutic measures (chemotherapy,combined with hormone therapy,etc.) (all P<0.05).Length of stay,chemotherapy,duration of granulocytosis,central venous catheterization,and invasive procedures were independent risk factors (all P<0.05),while age and empirical anti-infective therapy were protective factors (all P<0.05).The area under ROC curve of the prediction model was 0.8997,and the area under ROC curve of internal verification was 0.8704. Conclusions The established model can be used as a reference for effective measures to prevent nosocomial infection in middle-aged and elderly patients with acute leukemia.
[中图分类号]
R473.55
[基金项目]
四川省科技厅计划项目(2020YJ0339);泸州市健康社会工作研究课题(JK-2022-45)