基于logistic回归构建脑死亡患者家属器官捐献决策预测模型及其影响因素分析OA
Construction of predictive model for organ donation decisions of family members of brain death patients based on logistic regression and analysis of its influencing factors
目的 探讨影响脑死亡患者家属器官捐献决策的关键因素,构建预测模型,为临床沟通与捐献流程优化提供依据.方法 纳入2022年1月至2024年12月岑溪市人民医院ICU收治的150例脑死亡患者及其434名家属,收集患者、家属及决策过程特征.以是否同意捐献为因变量,采用单因素及多因素logistic回归分析筛选独立影响因素并构建预测模型,计算比值比(OR)及95％置信区间(CI).通过Hosmer-Lemeshow检验和受试者工作特征(ROC)曲线下面积(AUC)评估模型校准度和区分度,并采用Bootstrap方法进行内部验证.结果 150例脑死亡患者中,器官捐献同意率为14.7％.单因素logistic回归分析筛选出9个与捐献意愿相关的变量(P＜0.05).多因素分析表明,生前捐献意愿、展示无自主呼吸、男性患者、医疗团队沟通质量较优及家属无宗教信仰可显著促进捐献意愿(OR=4.78～6.90,均P＜0.05),而决策人数超过2人、医保类型为第三方责任及患者职业为个体户则为阻碍因素(OR=0.31～0.39,均P＜0.05).模型表现良好(AUC=0.889,95％CI:0.834～0.933;Hosmer-Lemeshow P=0.158),Bootstrap 校正后 AUC 为 0.838,提示预测效能稳定.结论 Logistic回归预测模型在预测器官捐献决策中表现良好,可为临床提供有效量化支持.生前意愿、无宗教信仰、男性患者、展示无自主呼吸、高质量沟通及决策人数不超过2人是促进捐献的关键因素.
Objective To explore key factors influencing family decision-making regarding organ donation in brain-dead patients,and to construct a predictive model,so as to inform clinical communication and donation process optimization.Methods A total of 150 brain-dead patients admitted to the ICU of Cenxi People's Hospital and their 434 family members from January 2022 to December 2024 were included,and the characteristics of patients,family members,and decision-mak-ing processes were collected.The consent to donate was used as dependent variable,and univariate and multivariate logistic regression analysis were performed to identify independent influencing factors and construct a predictive model,and then odds ratios(ORs)and 95％confidence intervals(CIs)were calculated.Model calibration and discrimination were assessed by Hosmer-Lemeshow test and the area under the receiver operating characteristic(RC)curve(AUC),and internal validation was performed using Bootstrap method.Results Among 150 brain death patients,organ donation consent rate was 14.7％.Univariate logistic regression identified 9 variables associated with donation willingness(P＜0.05).Multivariate analysis demonstrated that willingness to donate before death,absence of spontaneous respiration,male patients,high-quality commu-nication by the medical team,and absence of family religious belief among family members can significantly promote donation willingness(OR=4.78-6.90,all P＜0.05),while more than 2 decision-makers,third-party responsibility for medical in-surance,and individual patient occupation were obstructive factors(OR=0.31-0.39,all P＜0.05).The model showed good predictive performance(AUC=0.889,95％CI:0.834-0.933;Hosmer-Lemeshow P=0.158),and the Bootstrap-corrected A UC was 0.838,which indicated stable predictive ability.Conclusion The logistic regression-based predictive model demonstrates good performance in forecasting family decisions on organ donation,which can provide quantitative sup-port for clinical practice.Donation willingness before death,absence of religious belief,male patients,lack of autonomous breathing display,high-quality communication,and no more than 2 decision-makers are key factors in promoting donations.
张洪宾;高汉铭;廖吉祥;黄一洪;陈秋明;陆海生;罗庆教;刘远波
广西岑溪市人民医院重症医学科,广西岑溪 543200广西岑溪市人民医院重症医学科,广西岑溪 543200广西医科大学第二附属医院器官移植中心,广西南宁 530007广西岑溪市人民医院重症医学科,广西岑溪 543200广西岑溪市人民医院重症医学科,广西岑溪 543200广西医科大学第二附属医院器官移植中心,广西南宁 530007广西医科大学第二附属医院器官移植中心,广西南宁 530007广西岑溪市人民医院重症医学科,广西岑溪 543200
医药卫生
器官捐献脑死亡家属决策预测模型logistic回归
organ donationbrain deathfamily decision-makingpredictive modellogistic regression
《右江医学》 2026 (1)
19-26,8
广西壮族自治区卫生健康委员会自筹经费科研课题(Z20211305)广西医疗卫生重点学科建设项目(桂卫科发[2019]19号)
评论