Ecologicassociationbetween ambient air pollutantsand NICU admissions in Floridabetween 2012 and 2018

Abstract

Background/ Aim Neonatal intensive care units (NICU) in hospitals care for newborns with medical conditions such as preterm or low birthweight. While the association between ambient air pollution and adverse birth outcomes are well-studied, associations with NICU admission are under-studied. Our study investigates the ecological associations between short-term ambient PM 2.5 and NICU admissions and will assess if associations varies by place-based social stressors. Methods We conducted an ecological time-series analysis of electronic medical records (EMRs; n=1,019,253) for the state of Florida from January 1, 2012 to December 31, 2018. EMRs of births along with NICU admission were obtained from the OneFlorida Data Trust. Daily air pollutant average concentrations were obtained from the US EPA air monitoring database. We assigned daily Zip Code-level air pollution concentrations using inverse distance weighting. Socioeconomic indicators, greenspace, and maximum daily temperatures were used to control for time-invariant and time-variant spatial confounders. All variables were analyzed at the 5-digit zip code level. We used a multivariable zero-inflated Poisson regression model to test the ecologic association between PM 2.5 and NICU admissions with a random effect for metropolitan statistical area. Results Of the 1,019,253 neonates in our study, 4.8% (n=48,616) were admitted to a NICU. Controlling for daily maximum ambient temperature, socioeconomic status, and greenspace, random effects regression models showed each standard deviation increase of PM2.5 was significantly associated with a higher incidence rate ratio (IRR) for NICU admissions of IRRPM2.5=1.01 (95%CI=1.00, 1.02; p=0.004). Conclusion NICU admissions were significantly associated with short-term concentrations of PM2.5. This ongoing analysis will explore multiple day-lags of air pollutant levels as well as interaction between each air pollutant and place-based social stressors.