LAUSR

BACKGROUND To study the distribution of the birthplaces of very-low-birth-weight (VLBW) infants and examine whether delivery at different levels of hospital affects neonatal and infant mortality. METHODS This population-based cohort study was retrieved from Taiwan Maternal and Child Health Database. Livebirth singleton VLBW infants born between 2011 and 2014, with BW between 500 and 1499 g and gestational age ≥22 weeks were enrolled. The main outcomes were risk-adjusted odds ratios (aOR) of neonatal and infant mortality by birthplace, which was categorized as medical center (MC), regional hospital (RH), district hospital (DH), and clinic (C) based on Taiwan's hospital accreditation system. RESULTS Of 4560 VLBW infants enrolled, 3005 (66%) were born in MCs, 1181 (26%) in RHs, 213 (5%) in DHs, and 161 (4%) in Cs. Neonatal mortality rates were 10%, 15%, 16%, 17%, and infant mortality rates were 13%, 17%, 18%, 21%, if born in MCs, RHs, DHs and Cs, respectively. The aORs for neonatal and infant mortality were 1.94 (95% CI 1.53-2.48) and 1.67 (1.34-2.08) for those born in RHs, 2.26 (1.38-3.70) and 1.82 (1.16-2.86) for infants born in DHs/Cs, as compared to those born in MCs. For VLBW infants born in RHs, DHs, and Cs and postnatally transferred to MCs, the aORs of neonatal and infant mortality were lower than those who were not transferred. CONCLUSION VLBW infants born outside of MCs had higher neonatal and infant mortality and a two-fold higher risk of mortality than those born in MCs. When possible, VLBW infants should be born in MCs.