Analysis Of Risk Factors In Patients With Pulmonary Nodules And Establishment Of Clinical Prediction Model

Objectives: Through the analysis of the general clinical data,laboratory indicators and imaging characteristics of patients with pulmonary nodules,the risk factors of pulmonary nodules were summarized,and the clinical prediction model of pulmonary nodules was established,which provided objective basis for clinicians to evaluate the benign and malignant pulmonary nodules,improved the accuracy of diagnosis of pulmonary nodules,and reduced the impact of human factors on the judgment of benign and malignant pulmonary nodules.Methods:A total of 203 patients with pulmonary nodules who were evaluated by the multidisciplinary consultation team of the First Affiliated Hospital of Kunming Medical University from 2020 to 2021 and confirmed the pathological nature of pulmonary nodules were retrospectively analyzed as the modeling group.Pathological specimens were obtained through thoracic surgery,percutaneous lung biopsy,and transbronchial biopsy.According to the pathological results,they were divided into benign group(n = 52)and malignant group(n = 151).The general clinical data(gender,age,smoking history,respiratory symptoms,personal history of malignant tumor,family history of malignant tumor)of the enrolled patients were collected.Laboratory indexes(CEA,NSE,NLR,PLR,LMR);The imaging features of nodules(location,maximum diameter,shape,margin,presence or absence of cavity,air bronchogram sign,vascular convergence sign,pleural traction,and calcification)were analyzed.Univariate analysis was performed on each variable,and variables with significant statistical results were screened for binary logistic regression analysis.Finally,the independent risk factors of malignant nodules were summarized and the prediction model was established.In addition,70 patients with pulmonary nodules who underwent thoracic surgery and confirmed pathology in the First Affiliated Hospital of Kunming Medical University in 2022 were randomly selected as the validation group to verify the predictive value of the model.Results: 1.Univariate analysis showed that the patient’s age,nodule margin,GGO component,maximum diameter of nodule,vascular convergence sign,pleural traction sign,and NLR were statistically different between benign and malignant pulmonary nodules.2.Multivariate logistic regression analysis showed that: Age(OR = 1.04,P = 0.04),maximum diameter(OR = 1.11,P =0.01),margin(OR = 1.88,P < 0.001),GGO component(OR = 2.01,P =0.01),NLR(OR = 0.33,P < 0.001),PLR(OR = 1.01,P=0.01),and air bronchial sign(OR = 0.28,P =0.01)were statistically different between benign and malignant pulmonary nodules,and were independent predictors of benign and malignant pulmonary nodules.3.To establish the prediction model of malignancy of pulmonary nodules,malignancy predictive value(P)= ex/(1+ex),X = 3.03 +(0.04×age)+(0.10×maximum diameter)+(0.63×edge)+(0.7×GGO composition)+(-1.29×air bronchogram)+(-1.12×NLR)+(0.001×PLR).Where e is the natural logarithm.4.Malignant prediction rate of each patient was calculated,taking the pathological results as the gold standard,ROC curve was drawn,and the area under ROC curve was calculated to be 0.768(0.691-0.846).Appropriate critical value X=0.514 was selected as the standard value for the diagnosis of benign and malignant pulmonary nodules.The sensitivity and specificity of prediction of this model were 0.841 and0.673.5.The ROC curve was drawn by putting the data of validation group into the prediction model and Mayo model respectively.The results showed that the AUC of validation group was 0.753(95%CI: 0.632-0.874),and the AUC of Mayo model group was 0.669(95%CI: 0.539-0.799).The current model is more accurate than the Mayo model.Conclusions: 1.This study showed that age,maximum diameter,margin,GGO component,NLR,PLR and air bronchogram were independent predictors of benign and malignant pulmonary nodules.NLR and air bronchogram were protective factors.2.The regression equation of the prediction model established in this study is as follows: Malignant predictive value(P)= ex/(1+ex),X = 3.03 +(0.04×age)+(0.10×maximum diameter)+(0.63×edge)+(0.7×GGO composition)+(-1.29×air bronchogram)+(-1.12×NLR)+(0.001×PLR).Where e is the natural logarithm.The unit of age was year,the unit of diameter was mm,the unit of smooth edge was 0,the unit of irregular edge was 1,the unit of lobulation was 2,and the unit of burr was3.No GGO component was 0,some GGO component was 1,all GGO component was2.No air bronchogram was 0,air bronchogram was 1.3.The prediction model has the advantages of stability,accuracy and reliability,which is worthy of clinical promotion and assisting clinicians to develop individualized management plans for patients with pulmonary nodules.