急性白血病患儿化疗期间合并血流感染的危险因素分析及风险列线图模型的建立  

Analysis of risk factors and establishment of risk nomogram model for bloodstream infection in children with acute leukemia during chemotherapy

作　　者：周晓红 张春旭[1] 宋丽琴 ZHOU Xiaohong;ZHANG Chunxu;SONG Liqin(Children′s Hospital of Soochow University,Jiangsu 215000 China)

机构地区：[1]苏州大学附属儿童医院,215000

出　　处：《全科护理》2022年第17期2305-2310,共6页Chinese General Practice Nursing

摘　　要：目的:调查急性白血病(acute leukemia,AL)患儿化疗期间发生血流感染的危险因素并建立列线图模型。方法:选取医院2019年6月-2021年5月收治的443例AL患儿作为研究对象,根据患儿化疗期间是否发生血流感染,分为血流感染组(97例)和未感染组(346例)。通过对发生血流感染的相关临床因素进行单因素和多因素分析,得出独立预测因素,基于此构建列线图模型,并对模型进行评价。结果:住院时间≥30 d、30 d内抗菌药物暴露史≥3个品种、血白细胞计数<4.0×10^(9)/L、血清白蛋白<30 g/L、化疗阶段处于诱导期及经外周静脉置入中心静脉导管(PICC)置管时间≥20 d是AL患儿化疗期间合并血流感染的独立危险因素。基于以上影响因素建立列线图模型,模型验证结果显示,初始模型C-index为0.878,与内部验证的C-index 0.876较为接近,校正曲线显示具有良好的区分度及一致性。结论:研究建立的AL患儿血流感染预测模型具有较高的诊断价值,可以以更加直观和简洁的形式为病人提供个体化血流感染风险预测。Objective:To investigate the risk factors of bloodstream infection in children with acute leukemia(AL)during chemotherapy and establish a nomogram model.Methods:A total of 443 children with AL admitted to hospital from June 2019 to May 2021 were selected as the research objects.According to whether the children had bloodstream infection during chemotherapy or not,they were divided into bloodstream infection group(97 cases)and non-infected group(346 cases).Through single-factor and multi-factor analysis of the relevant clinical factors of bloodstream infection,independent predictive factors were obtained,and a nomogram model was constructed based on this,and the model was evaluated.Results:Hospitalization time≥30 days,antibacterial drug exposure history≥3 varieties within 30 days,white blood cell count<4.0×10^(9)/L,albumin<30 g/L,chemotherapy phase in the induction phase,and catheterization time of peripherally inserted central venous catheter(PICC)≥20 d were the independent risk factors for bloodstream infection in children with AL during chemotherapy.The nomogram model was established based on the above influencing factors.The model verification results showed that the initial model C-index was 0.878,which was closer to the internally verified C-index 0.876.The calibration curve showed good discrimination and consistency.Conclusion:The bloodstream infection prediction model established in study has higher diagnostic value,and can provide the individualized bloodstream infection risk prediction for patients with more intuitive and concise forms.

关 键 词：急性白血病 血流感染 列线图 风险因素 

分 类 号：R733.71[医药卫生—肿瘤]