Establishment of Logistic Regression Prediction Model for Epidemiological Characteristics and Influencing Factors of Pulmonary Tuberculosis in Xuzhou from 2019 to 2021
Objective:To investigate the epidemiological characteristics of pulmonary tuberculosis in Xuzhou from 2019 to 2021,and construct a logistic regression prediction model for influencing factors.Methods:788 cases of pulmonary tuberculosis registered in Xuzhou from January 1,2019 to December 30,2021 were selected as the study subjects.Observed the epidemiological characteristics of pulmonary tuberculosis in Xuzhou from 2019 to 2021,and used univariate and multivariate logistic regression analysis to analyze the influencing factors of pulmonary tuberculosis incidence.Constructed a Log P regression prediction model and analyzed its predictive performance.Result:Epidemiological statistics showed that from 2019 to 2021,the majority of pulmonary tuberculosis patients in Xuzhou were male,aged ≥ 60 years old,residing in rural areas,and married,the number of patients in the disease course and education level classification group was relatively close.Univariate analysis showed that the incidence of pulmonary tuberculosis was related to contact history of pulmonary tuberculosis,drinking history,tuberculosis health education,gender,average monthly income of per person in family,age,occupation,smoking history,BCG vaccination history,and residence(P＜0.05).The results of multivariate logistic regression analysis showed that residing in rural areas,smoking history,gender as male,drinking history,average monthly income of per person in family＜3000 yuan,age≥60 years old,occupation as a farmer or worker,and contact history of pulmonary tuberculosis were risk factors for pulmonary tuberculosis,while BCG vaccination history was a protective factor(P＜0.05).The corresponding Log P model,after ROC analysis,it showed that this prediction model had good predictive performance for the incidence of pulmonary tuberculosis,with an ROC-AUC(0.95C1)of 0.881(0.817-0.928)and an accuracy of 86.5％.Conclusion:Xuzhou should focus on population affected by pulmonary tuberculosis of male,elderly,residing in rural areas,smoking history,drinking history,low average monthly income of per person in family,farmers or workers,and contact history of pulmonary tuberculosis,and should actively carry out the vaccination of BCG vaccine.In addition,the prediction model constructed through influencing factors has good predictive performance on the incidence of pulmonary tuberculosis.
XuzhouPulmonary tuberculosisEpidemiological characteristicsInfluencing factorsPrediction model
NETL
NSTL
万方数据
