Analysis Of Incidence And Risk Factors Of ADLI In Southern Xinjiang And Establishment Of Nomogram Prediction Model

Objective : To explore the incidence and the risk factors of anti-tuberculosis drug-induced liver injury(ADLI)in patients with Tuberculosis in Southern Xinjiang,and establish a Nomogram prediction model.Methods:A retrospective study was conducted on 300 tuberculosis patients,who were centrally isolated and treated from September2020 to March 2022 in the Southern Xinjiang ’s Hospital.The inpatients with ADLI were selected as the case group(41 cases)and the Non-ADLI patients as the control group(259cases).The General clinical information of the study subjects before anti-tuberculosis treatment such as age,gender,treatment Status,tuberculosis condition,nutritional status,history of underlying diseases and other laboratory indicators such as Hemoglobin(Hb),Albumin(ALb)were collected.And the independent risk factors of ADLI were screened by single-factor and multi-factor logistic regression,which R software was used to construct nomogram predictive models and conduct the power validation.Results :In this study,13.67%(41/300)patients developed ADLI.Multivariate logistic regression analysis showed that Retreatment(OR=2.599,P=0.022),hypoproteinemia(OR=4.941,P < 0.001),virus hepatitis(OR=6.215,P<0.001),Other liver diseases(OR=3.501,P=0.035),extrapulmonary tuberculosis(OR=3.407,P=0.019)were independent risk factors for ADLI in patients with tuberculosis in Southern Xinjiang.The Nomogram prediction model was constructed based on the independent risk factors.The area under the receiver operating characteristic(ROC)curve was0.833(95%CI:0.761-0.905).The Bootstrap method was used for internal repeated sampling for 1 000 times,The C-index of the nomogram prediction model was 0.816,95%CI:(0.768,0.864),indicating that the model has a good discrimination ability.Both the model performance curve and the calibration curve are in good agreement with the overall shape of the ideal curve,indicating that the model has a good Degree of calibration.Conclusions:Retreatment,hypoproteinemia,virus hepatitis,Other liver diseases and extrapulmonary tuberculosis may increase the the incidence of ADLI in patients with tuberculosis in Southern Xinjiang.The Nomogram prediction model for ADLI among patients with tuberculosis in this study has good predictability,and can provide a scientific basis for clinical early identification and prevention.