结核性胸膜炎数学诊断模型的建立及评估  

Establishment and evaluation of mathematical diagnosis model of tuberculous pleurisy

作　　者：刘丽波[1] 王楠[1] LIU Li-bo(Department of clinical laboratory,the First Affiliated Hospital of Dalian Medical University,Dalian,Liaoning,116023,China)

机构地区：[1]大连医科大学附属第一医院检验科,辽宁大连116023

出　　处：《齐齐哈尔医学院学报》2020年第7期818-821,共4页Journal of Qiqihar Medical University

摘　　要：目的通过对本院结核性胸膜炎患者实验室诊断指标进行数学分析,初步建立结核性胸膜炎数学诊断模型,辅助鉴别诊断早期结核性胸膜炎,提高临床诊断效率及准确性。方法选择2015年1月-2017年5月在本院主要诊断为结核性胸膜炎的30例患者作为研究对象,年龄在60岁以上,收集病历中ESR、CRP、ADA、PCR等结果,以及是否有胸水量大于1000 ml,发热、胸痛、呼吸困难等体征,同时选择同期30例非结核性胸膜炎患者的相同指标作为对照,阳性者赋值为"1",阴性赋值为"0",用SPSS 17.0对数据进行统计分析。结果通过应用两类疾病临床指标所占例数,建立非条件logistic回归模型可知,γ-IFN释放试验、PCR、ADA、ESR、胸水量五项结果有明显差异(P<0.05),对模型建立有意义,因此得到方程式Y=1.727+3.186×T-SPOT+5.347×PCR+6.278×ADA+3.601×ESR+3.277×胸水量,cut-off值为11.700,即Y值大于11.700即为结核性胸膜炎。结论经过对15例结核性胸膜炎和20例非结核性胸膜炎患者的效能验证证明,本模型对结核性胸膜炎的诊断是具有可行性,可为临床早期诊断提供新帮助。Objective To establish an initiatory mathematical diagnostic model of tuberculous pleurisy through mathematical analysis of laboratory diagnostic indexes of these patients in the first affiliated hospital of Dalian Medical University,and assist differential diagnosis of early tuberculous pleurisy,so as to improve the efficiency and accuracy of clinical diagnosis.Methods From 2015 to February 2017,30 cases in the first affiliated hospital of Dalian Medical University mainly diagnosed with tuberculous pleurisy patients were collected as study objects,those patients were over 60 years old.Concrete results including ESR,CRP,ADA,and PCR,and whether the chest water was greater than 1000 ml,signs including fever,chest pain,difficulty breathing were collected from the medical record,and 30 patients with non-tuberculous pleurisy were selected as control and the same indicators were collected.Positive assignments were“1”,otherwise negative,assigned“0”.Then,those results were analyzed by SPSS 17.0.Results By applying the number of clinical indicators of the two types of diseases of the unconditional logistic regression model,γ-IFN release test,PCR,ADA,ESR,pleural effusion had a significant difference(P<<0.05),then we got an equation,Y=1.727+3.186×T-SPOT+5.347×PCR+6.278×ADA+3.601×ESR+3.277×pleural effusion.And the cut-off value is 11.700,means that when the Y is greater than 11.700,it is diagnosed as tuberculous pleurisy.Conclusions The efficacy of 15 cases of tuberculous pleurisy and 20 cases of non-tuberculous pleurisy proves that this model is feasible for the diagnosis of tuberculous pleurisy and could provide a new and convenient method for early clinical diagnosis.

关 键 词：结核性胸膜炎 数学诊断模型 实验室指标 

分 类 号：R446.9[医药卫生—诊断学]