(370010-48) Mechanical Ventilation and Venous Thromboembolism Risk: A Multicenter, Retrospective Cohort Study

Introduction: Hospital-acquired venous thromboembolism (HA-VTE) is a leading cause of in-hospital morbidity and mortality. Invasive mechanical ventilation (IMV) may be a risk factor for HA-VTE development among critically ill children. We aim to estimate the cumulative incidence and odds of HA-VTE among critically ill children with and without IMV exposure.

Methods: We performed a retrospective, observational cohort study of children < 18 years of age hospitalized in a pediatric intensive care unit within 142 pediatric intensive care units affiliated with the Virtual Pediatric Systems registry from Jan 2016 - Dec 2022. We excluded post-operative cases, neonatal admissions, encounters with critical congenital heart disease, IMV exposure < 24 hours, preexisting tracheostomy, or if HA-VTE was present on admission. The primary outcome was cumulative incidence of HA-VTE (including deep venous thrombosis [DVT], pulmonary embolism, and organ-specific venous thrombosis). Descriptive features and putative HA-VTE risk factors were compared using chi-squared or Wilcoxon rank-sum testing. Multivariate logistic regression was employed to assess IMV exposure (and IMV duration in a IMV-only subgroup analysis) as a putative HA-VTE risk factor.

Results: A total of 691,118 eligible encounters were studied of which 86,922 (12.4%) underwent IMV. Those with IMV exposure had a greater HA-VTE rate (1.9%) compared to those without IMV exposure (0.1%). Most HA-VTE were DVT (93.3%). The median time to HA-VTE development after IMV exposure was 6 days (interquartile range: 3,14). After accounting for prothrombotic risk factors (e.g., central venous catheterization, patient age, and infectious disease), IMV exposure was independently associated with HA-VTE (odds ratio [OR] 2.83, 95% confidence interval [CI] 2.46 to 3.25, P< 0.001). In the IMV-only subgroup, a greater duration of IMV exposure was independently associated with HA-VTE (OR 1.029, 95% CI 1.027-1.032, P< 0.001).

Conclusions: Among critically ill children, we observed HA-VTE to be independently associated with IMV exposure and duration thereof. While these findings require prospective validation, they suggest critically ill children undergoing IMV are an important target population for future risk-stratified pediatric thromboprophylaxis trials