Resident Physician Nationwide Children's Hospital Columbus, Ohio, United States
Background: The persistence of racial disparities in infant mortality calls neonatal intensive care units (NICUs) to evaluate disparities in palliative care. While little research exists, one study on neonatal palliative care found Black infants more likely to receive cardiopulmonary resuscitation (CPR) than White infants. Conversely, the benefits of palliative care involvement are well documented, such as symptom management, facilitation of goals of care and prognosis conversations, and psychological support. This study aims to assess potential disparities in palliative care involvement that could, in part, explain differences in the way infants die. Objective: Primary: To evaluate disparities in neonatal palliative care involvement within a large, 10-unit midwestern neonatal network by infant race, ethnicity and language preference. Secondary: To assess disparities in in-hospital mortality for neonatal palliative care patients by infant race, ethnicity and language preference. Design/Methods: We are conducting a retrospective study of infants with a postnatal palliative care consultation at any site within the 10-unit neonatal network across a 10-year period (admitted 1/1/2014-1/1/2024). Using electronic medical record abstracted data, demographic and clinical features will be summarized with median (IQR) for continuous variables and count (percent) for categorical variables. Kaplan-Meier estimates will be used to visualize time from admission to palliative care consult and from consult to death or discharge by varying demographic features (race, ethnicity, language preference). Either parametric or semi-parametric survival models (accelerated failure time models or Cox proportional hazards) will be utilized to assess the association of demographic features with mortality, code status change, and CPR occurrence. Model fit and satisfaction of key assumptions will be assessed. Relevant transformations to model coefficients alongside 95% confidence intervals will be presented. IRB granted approval in October, data will be abstracted by December, and analysis completed by March 2025.
