Neonatal-Perinatal Medicine Fellow University of Michigan Medical School West Bloomfield, Michigan, United States
Background: The neonatal period is a critical time for neurological development while simultaneously being a time of high seizure susceptibility. Seizure management in this population continues to be variable, with goals of improving neurodevelopmental outcomes, minimizing adverse effects of antiseizure medications (ASMs), and avoidance of secondary injury. There are limited systematic data about the incidence and severity of adverse cardiorespiratory effects of ASM initiation during acute neonatal seizure management. Objective: Determine the relationship between ASM loading doses and subsequent escalation in respiratory and/or vasopressor support requirements for neonates in the University of Michigan Newborn Intensive Care Unit within the first 72 hours after ASM initiation. Design/Methods: NICU patients born between January 2016 and December 2022 with a postmenstrual age of < 44 weeks at the time of their first seizure were included. Infants were excluded if they received loading doses of ASMs at another institution, had major medical or surgical problems affecting their cardiorespiratory status, or died within the first 72 hours of seizure management. Respiratory and cardiovascular requirements and cumulative ASM doses within the first 72 hours of medication initiation were obtained. Data analysis will be done across the following time periods: 0-6 hours, 6-12 hours, 12-24 hours, and 24-72 hours after the first ASM was started. Spearman correlation will be performed for ASMs and change in respiratory support and vasoactive inotropic score during each time period, followed by multivariate analysis taking into account concurrent sedative/analgesic administration, underlying neurologic diagnosis, and other confounding variables. Multiple linear regression will be done to evaluate the effect of ASM dose received on change in respiratory support and vasoactive inotropic score after adjusting for confounding variables.
