Comparison of the value of the Swensen model,Brock model and PKUPH model in predicting the properties of solitary pulmonary nodules

SU Shuyan; SHI Jiaxin

doi:10.3969/j.issn.1674-4136.2024.06.013

Abstract：Objective To compare the value of Swensen model,Brock model and PKUPH model in predicting the properties of solitary pulmonary nodules.Methods One hundred and twenty seven patients with solitary pulmonary nodules confirmed by the Affiliated Lianyungang Hospital of Xuzhou Medical University from January 2020 to December 2021 were taken as the research objects,the clinical data were collected,and parameters related to the Swensen model,Brock model and PKUPH model were substituted into the modles for calculation,in order to predict the properties of the solitary pulmonary nodules,and the ROC curves were used to compare the predictive efficacy of these 3 models.Results Among the 127 patients with solitary pulmonary nodules,there were 52 benign nodules,accounting for 25.12％,and 80 malignant nodules,accounting for 74.88％.The parameters related to the Swensen model,Brock model and PKUPH model of the study subjects were substituted into the models and the ROC curves were calculated,and the results showed that the area under the curve (AUC) of the Swensen model,Brock model and PKUPH model were 0.631 (95％CI:0.533～0.730,P<0.05),0.509 (95％CI:0.409～0.608,P>0.05),0.812 (95％CI:0.742～0.881,P<0.05);the AUC values of Swensen model,Brock model,PKUPH model were 0.631,0.509,and 0.821,with the sensitivities of 72.9％,61.3％,70.3％,and the specificities of 57.7％,48.1％,82.7％,respectively.The positive likelihood ratios were 1.641,1.074,1.808,and the negative likelihood ratios were 0.402,0.813,0.155,respectively,with the positive predictive values of 90.97％,60.65％,70.97％,and the negative predictive values of 28.85％,46.15％,82.69％.Conclusions Swensen model,Brock model and PKUPH model can be applied in the prediction of the properties of solitary pulmonary nodules,among which PKUPH model has the best predictive efficacy.