Risk factors for the development of hospital-acquired pediatric venous thromboembolism—Dealing with potentially causal and confounding risk factors using a directed acyclic graph (DAG) analysis

Abstract

Introduction Hospital-acquired venous thromboembolism (HA-VTE) in children comprises multiple risk factors that should not be evaluated separately due to collinearity and multiple cause and effect relationships. This is one of the first case-control study of pediatric HA-VTE risk factors using a Directed Acyclic Graph (DAG) analysis. Material and methods Retrospective, case-control study with 22 cases of objectively confirmed HA-VTE and 76 controls matched by age, sex, unit of admission, and period of hospitalization. Descriptive statistics were used to define distributions of continuous variables, frequencies, and proportions of categorical variables, comparing cases and controls. Due to many potential risk factors of HA-VTE, a directed acyclic graph (DAG) model was created to identify confounding, reduce bias, and increase precision on the analysis. The final model consisted of a DAG-informed conditional logistic regression. Results In the initial conventional univariable model, the following variables were selected as potential risk factors for HA-VTE: length of stay (LOS, days), immobility, ICU admission in the last 30 days, LOS in ICU, infection, central venous catheter (CVC), number of CVCs placed, L-asparaginase, heart failure, liver failure, and nephrotic syndrome. The final model using the set of variables selected by DAG analysis revealed LOS (OR = 1.106, 95%CI = 1.021–1.198, p = 0.013), L-asparaginase (OR = 26.463, 95%CI = 1.609–435.342, p = 0.022), and nephrotic syndrome (OR = 29.127, 95%CI = 1.044–812.508, p = 0.004) as independent risk factors for HA-VTE. Conclusion The DAG-based approach was useful to clarify the influence of confounders and multiple causalities of HA-VTE. Interestingly, CVC placement—a known thrombotic risk factor highlighted in several studies—was considered a confounder, while LOS, L-asparaginase use and nephrotic syndrome were confirmed as risk factors to HA-VTE. Large confidence intervals are related to the sample size; however, the results were significant.