To the Editor We read with interest the study by Connelly et al1 that proposes a risk factor–based predictive model for venous thromboembolism (VTE) in pediatric trauma patients. Based on data from the National Trauma Data Bank, the authors used robust regression techniques by constructing mixed-effect models on a randomly derived training set using a “split-sample” approach with 1:1 randomization. Internal validation of the derived estimates was performed on a training subset by evaluating each model performance by computing the area under the receiver operating characteristic curves. With limited literature entailing VTE risk estimates in a pediatric population, the authors are to be commended for their efforts in developing a rigorous bedside “clinical tool” that could potentially aid in risk stratification, thereby allowing preemptive measures for the mitigation of VTE.
Kalakoti P, Notarianni C, Nanda A. Predicting Venous Thromboembolism in Pediatric Trauma Patients. JAMA Surg. 2016;151(9):881-882. doi:10.1001/jamasurg.2016.0481