Abstract: Objective To analyze the influencing factors of peripherally inserted central catheter (PICC)-related thrombosis during induction chemotherapy in children with acute leukemia by Logistic regression analysis. 
Methods A total of 184 children with acute leukemia were retrospective selected. According to the presence or absence of thrombosis at the catheter site during induction chemotherapy period(at 15 d after catheter placement), they were divided into thrombus group(n=38) and the non-thrombosis group(n=146). The clinical data of the two groups were collected, and univariate and multivariate Logistic regression analyses were used to analyze the influencing factors of PICC-related thrombosis during induction chemotherapy in children with acute leukemia. The receiver operating characteristic(ROC) curve was used to evaluate the predictive value of multivariate Logistic regression model in PICC-related thrombosis in children with acute leukemia during induction chemotherapy. 
Results There was no significant difference between two groups in age, sex, disease type, disease risk, vein catheterization, as well as white blood cell(WBC), platelet(PLT), D-dimer(D-D), prothrombin time(PT), and activated partial thromboplastin time(APTT) at the time of catheterization, and WBC, PLT, PT and APTT at 15 d after catheterization(P＞0.05). There were significant differences in catheter placement, catheter-related infection, use of hemostatic drugs and D-D level at 15 d after catheterization between two groups(P＜0.05). Multivariate Logistic regression analysis showed that right-sided catheterization, catheter-related infection, use of hemostatic drugs, and high D-D level at 15 d after catheterization were independent risk factors for PICC-related thrombosis in children with acute leukemia during induction chemotherapy(P＜0.05). ROC curve analysis showed that the area under the ROC curve(AUC) of multivariate Logistic regression model in predicting PICC-related thrombosis in children with acute leukemia during induction chemotherapy was 0.917, 95%CI was 0.866 to 0.954, sensitivity was 84.21%, and specificity was 84.44%. 
Conclusion There are a number of independent risk factors for PICC-related thrombosis during induction chemotherapy in children with acute leukemia. Multivariate Logistic regression model has a reliable predictive value, which is conductive to reduction of the risk of PICC-related thrombosis.

Key words: leukemia, biphenotypic, acute, induction chemotherapy, peripherally inserted central catheter-related thrombosis