2型糖尿病患者再入院风险预测模型的构建  

Construction of readmitted risk prediction model for patients with type 2 diabetes mellitus

作　　者：岳江红 彭曦[2] 蔡梦文 李鹃 蔡钢 张选明 王恒[1] YUE Jianghong;PENG Xi;CAI Mengwen;LI Juan;CAI Gang;ZHANG Xuanming;WANG Heng(College of Pharmacy,Shihezi University,Xinjiang Shihezi 832003,China;Department of Pharmacy,The First Affiliated Hospital of Shihezi University,Xinjiang Shihezi 832008,China;Department of Traditional Chinese Medicine,The First Affiliated Hospital of Shihezi University,,Xinjiang Shihezi 832008,China)

机构地区：[1]石河子大学药学院,新疆石河子832003 [2]石河子大学第一附属医院药学部,新疆石河子832008 [3]石河子大学第一附属医院中医科,新疆石河子832008

出　　处：《中国医院药学杂志》2024年第14期1687-1692,共6页Chinese Journal of Hospital Pharmacy

基　　金：新疆生产建设兵团重大科技项目资助(编号:2020AA005)。

摘　　要：目的:探究2型糖尿病患者1年内再入院的危险因素,构建2型糖尿病患者1年内再入院的风险预测模型,并评估其区分度和有效性。方法:采用回顾性病例对照研究,收集石河子大学第一附属医院2019年6月至2022年6月经诊疗后出院的2型糖尿病患者的病例资料,按照纳入和排除标准,筛选1年内再入院患者117例纳入再入院组,按约1∶3的比例随机抽取同期未再入院患者360例纳入未再入院组,以性别作为匹配因素。采用单因素分析与多因素logistic回归分析筛选出院后1年内再入院的危险因素,构建风险预测模型,采用受试者工作特征(ROC)曲线、校准曲线、决策曲线评估该模型的区分度及有效性。结果:空腹血糖高、有冠心病病史、入院类型为急诊、病程长、有体质量变化、发生低血糖、有并发症为2型糖尿病患者1年内再入院的独立危险因素。规律监测血糖为2型糖尿病患者1年内再入院的保护因素。ROC曲线下面积(AUC)为0.747;约登指数最大值为0.374,灵敏度为82.1%,特异度为55.3%,截断值为0.187。结论:构建的2型糖尿病患者1年内再入院风险预测模型具有良好的预测效果,能有效识别再入院高危人群,为临床早期预防及干预2型糖尿病患者再入院提供了参考依据。OBJECTIVE To explore the risk factors for readmission within 1 year in patients with type 2 diabetes mellitus(T2DM)to construct a risk prediction model for readmission within 1 year and assess its differentiation and validity.METHODS From June 2019 to June 2022,a retrospective case-control study was performed.The relevant case data were collected from discharged T2DM patients after treatment.And 117 patients readmitted within 1 year were selected as readmission group while 360patients not readmitted during the same time period were randomized into non-readmitted group in a ratio of approximately 1:3with gender as a matching factor.The risk factors for readmission within 1 year post-discharge were screened by univariate and multivariate Logistic regression analyses.And a risk prediction model was constructed for assessing the differentiation and validity of model by the curves of ROC,calibration and decision.RESULTS High fasting glucose,history of coronary artery disease,type of emergency admission,long duration of illness,change in body mass,occurrence of hypoglycaemia and complications were independent risk factors for readmission within 1 year in T2DM patients.Regular monitoring of blood glucose was a protective factor for readmission within 1 year.The area under the ROC curve(AUC)was 0.747.The maximal value of Jordon's index was 0.374 with a sensitivity of 82.1%,a specificity of 55.3%and a cut-off value of 0.187.CONCLUSION The prediction model of readmission risk for T2DM patients within 1 year yields excellent outcomes.It may effectively identify the high risk group of readmission and provide references for clinical early prevention and intervention of readmission of T2DM patients.

关 键 词：2型糖尿病 再入院 影响因素 风险预测模型 

分 类 号：R587.1[医药卫生—内分泌]