摘要
目的观察与儿童大叶性肺炎(湿热壅盛证)辨证相关的特征性临床数据,构建诊断预测模型并评价模型的预测效能。方法以山东中医药大学附属医院儿科电子病历系统为资料来源,回顾性抽取2019年1月1日至2022年12月31日大叶性肺炎住院患儿病历资料254例,按是否为湿热壅盛证分为湿热壅盛组112例、非湿热壅盛组142例。采用单因素和多因素分析方法筛选儿童大叶性肺炎(湿热壅盛证)的辨证相关因素,根据筛选结果初步建立该证型的诊断预测模型并通过Hosmer-Lemeshow拟合度检验和受试者工作特征曲线(ROC曲线)对模型进行评估。结果单因素和多因素分析结果显示起病时以发热为主症、发热持续时间、肺炎支原体、EB病毒、总胆汁酸是儿童大叶性肺炎(湿热壅盛证)的影响因素(P<0.05),构建的诊断预测模型为P=1/[1+exp(0.962X1+0.112X2+1.446X3+0.770X4+0.078X5-3.372)]。Hosmer-lemeshow检验结果显示该模型的理论预测与实际情况有较好的拟合度(χ^(2)=7.155,P=0.520);ROC曲线下面积为0.724,诊断敏感度为79.0%,特异性为55.1%。结论建立的诊断预测模型与临床辨证切合良好,以最优变量子集实现了儿童大叶性肺炎(湿热壅盛证)的辨证预测。
Objective:To analyze the characteristic clinical data about pattern identification of pattern of damp⁃ness-heat congestion and excessiveness in children with lobar pneumonia,to establish a diagnostic prediction mod⁃el and evaluate its efficiency.Methods:Based on the electronic medical record system in the pediatric depart⁃ment of the Affiliated Hospital of Shandong University of Traditional Chinese Medicine,254 cases with lobar pneu⁃monia from January 1st,2019 to December 31st,2022 were retrospectively selected.According to pattern identifica⁃tion,112 cases were enrolled in the dampness-heat congestion and excessiveness pattern group and 142 cases were in the other patterns group.Monofactor and multifactor analysis were adopted to screen pattern identification factors of dampness-heat congestion and excessiveness pattern in children with lobar pneumonia.The model was established according to the screening results and was evaluated by the Hosmer-Lemeshow test and ROC curve.Results:Monofactor and multifactor analysis showed that the influencing factors of dampness-heat congestion and excessiveness pattern in children with lobar pneumonia included fever as the main symptom at the disease onset,the duration of fever,mycoplasma pneumoniae,EB virus,and total bile acid(P<0.05).The established diagnostic prediction model was P=1[/1+exp(0.962X1+0.112X2+1.446X3+0.770X4+0.078X5-3.372)].The Hosmer-Leme⁃show test showed a good model fit with the actual situation(χ^(2)=7.155,P=0.520).The area under the ROC curve was 0.724,the diagnostic sensitivity was 79.0%and the specificity was 55.1%.Conclusions:The diagnostic pre⁃diction model fits in well with clinical pattern identification and can predict the pattern identification of dampnessheat congestion and excessiveness in children with lobar pneumonia by best subset selection.
作者
王兴丽
劳慧敏
王梦琦
黄娟
Wang Xingli;Lao Huimin;Wang Mengqi;Huang Juan(Shandong University of Traditional Chinese Medicine,Shandong,Jinan 250013,China)
出处
《中国中医急症》
2023年第12期2069-2072,2077,共5页
Journal of Emergency in Traditional Chinese Medicine
基金
国家中医药管理局中医药循证能力建设项目(2019XZZX-EK004)
山东省科学技术厅山东省重点研发计划项目(2016GSF202043)
山东省自然科学基金面上项目(ZR2022MH081)
山东省中医药管理局山东省中医药科技发展计划项目(2019-0124,2019-0206)。
关键词
大叶性肺炎
儿童
湿热壅盛证
预测模型
Lobar pneumonia
Children
Dampness-heat congestion and excessiveness pattern
Prediction model
