Font Size: a A A

Analysis Of Clinical Characteristics Of Lung Cancer And Lung Cancer Complicated With Chronic Obstructive Pulmonary Disease And Establishment And Verification Of Prediction Model

Posted on:2020-04-03Degree:MasterType:Thesis
Country:ChinaCandidate:H ChenFull Text:PDF
GTID:2494306728998979Subject:Internal medicine (pulmonology)
Abstract/Summary:PDF Full Text Request
ObjectiveTo collect clinical data of patients with simple lung cancer and chronic obstructive pulmonary disease(COPD)complicated with lung cancer,apply statistical analysis(single factor,multiple factors and correlation)to find independent risk factors affecting the occurrence of lung cancer complicated with COPD,establish and verify a mathematical model that can predict lung cancer with COPD,provide objective model support for the early diagnosis of lung cancer with COPD,reduce misdiagnosis and missed diagnosis of the disease,and guide clinicians to make early diagnosis and formulate the next treatment plan.MethodsData of 299 cases of lung cancer patients with definite pathological diagnosis for the first time through surgery or non-surgery means,and patients with definite pulmonary function diagnosed as COPD during the period from March 2014 to March 2017 were retrospectively collected,recorded as group A.Data collection for each case includes gen-eral information,hematology and auxiliary examinations.General information includes:age,gender,smoking,smoking index,clinical symptoms,past medical history,family history;hematological examinations include:Leukocyte count(WBC),erythrocyte sedim-entationrate(ESR),neuron-specificenolase(NSE),cytokeratin19fragments(CYFRA 21-1),carcinoembryonic antigen(CEA),carbohydrate antigen 125(CA125),carbohydrate antigen153(CA153),carbohydrate antigen 199(CA199);auxiliary examinations include lung function,imaging,and pathology of lung cancer by surgical or non-surgical means.Lung function examinations include forced expiratory volume in one second(FEV1),forced vital capacity(FVC),FEV1/FVC and the percentage of FEV1(FEV1%pred).Imaging exami-nation is the main tumor location and the presence of pulmonary emphysema.According to the results of pulmonary function,whether COPD was combined was divided into two groups:simple lung cancer group and lung cancer combined with COPD group.Based on the above data,statistically significant factors were selected by logistic regression analysis to construct a mathematical model that can predict lung cancer with COPD.In addition,during the period from April 2017 to September 2018,a total of 50 patients with path-ological diagnosis and a diagnosis of COPD with clear lung function were identified in Qingdao Municipal Hospital Group for the first time,recorded as group B.The ROC curve was drawn according to the data of group B,and the area under the curve(AUC)was calculated to verify the predictive ability of the model,and the sensitivity and specificity of the model for lung cancer combined with COPD were verified.ResultIn all the collected data,through single factor analysis,between the two groups,Age,gender,smoking,smoking index,clinical symptoms(cough,sputum,fever,wheezing),pulmonary function indicators,ESR,tumor location in imaging and the presence of pulmonary emphysema and the pathological types of tumor are significant differences.Multivariate analysis showed that the age,smoking and clinical symptoms(wheezing)of patients with lung cancer and lung cancer combined with COPD were statistically significant.These factors can independently affect lung cancer and lung cancer with COPD.Using the above influencing factors as data support for the study,the mathematical model for predicting lung cancer with COPD was calculated as=ex/(1+ex),x=-8.2+0.1×ages+1.4×smoking+1.5×wheezing,e is the natural logarithm,e≈2.72.The sensitivity of the mathe-matical model for predicting lung cancer with COPD was 65.9%(27/41).Using Group B data,specificity was 66.7%(6/9),positive predictive value was 90.0%(27/30),and negative predictive value was 30.0%(6/20).The area under the mathematical prediction model curve obtained in this study is 0.855±0.027.ConclusionCompared with the simple lung cancer group,lung cancer combined with COPD is mainly older men with an average age of more than 60 years old.They almost have long-term and large smoking history.The clinical symptoms often include cough,sputum,chest tightness,wheezing,fever,chest pain,hoarseness,weight loss,blood in the sputum.The pathological type is mainly squamous cell carcinoma,and the tumor site is more common in the left upper lobe and the right lower lobe.According to the GOLD classification[1],the lung function of patients with lung cancer complicated with COPD is mostly grade II.The age of patients with simple lung cancer was 8 years younger than that of lung cancer combined with COPD.The pathological type was mainly adenocarcinoma,and the lesion occurred in the right upper lobe.By establishing a lung cancer combined with COPD prediction formula,it is of great significance to help guide the early screening of lung cancer and lung cancer with COPD.
Keywords/Search Tags:lung cancer, lung cancer combined with COPD mathematical prediction model, Logistic regression analysis
PDF Full Text Request
Related items