Construction and validation of a nomogram prediction model for venous thrombosis in children with non-tun-neled central venous catheters
Objective To develop and validate a nomogram prediction model of venous thrombosis(VT)in children with non-tun-neled central venous catheter(CVC).Methods We collected the clinical data of 388 children who received non-tunnel CVC in our hospital from December 2022 to May 2023,and divided them into a training set(n=260)and a testing set(n=128)in light of the ratio of about 2:1.Based on multivariate Logistics regression,we found and integrated risk factors of CVC-related VT,developed an effective nomogram tool,and conducted internal validation.Calibration capability and evaluative efficiency of nomograms were determined by calibration curves,Hosmer-Lemeshow goodness-of-fit tests,and receiver operating characteristic(ROC)curves.Results Using multivariate Logistic regression analy-sis,we obtained five risk factors,including age in months,glucocorticoid,parenteral nutrition,CVC location and lumen number(P＜0.05).We integrated all risk factors into a nomogram to predict the risk of CVC-related VT.The area under the ROC curve(AUC)of the nomogram was 0.857(95％CI,0.809～0.897)and 0.846(95％CI,0.772～0.904)in the training and testing sets,respectively.The calibration curve shows high agreement between the actual and predicted probabilities,and the Hosmer-Lemeshow goodness-of-fit test is good(P＞0.05).Conclusion The nomogram developed based on the risk factors of clinical characteristics can be used as a reliable tool to evaluate VT associated with non-tunnel CVC in children.
万方数据
维普
