Indomethacin Prophylaxis in Preterm Infants: Changes over Time

Background: Neonatal brain injury is one of the main causes of neonatal mortality and morbidity. Since 1994, multiple studies have provided evidence for and against indomethacin prophylaxis in the prevention of brain injury in preterm infants. However, the change in prevalence of indomethacin prophylaxis across different centers over time is unknown. This study aims to examine changes in the use of indomethacin prophylaxis in the neonatal intensive care unit (NICU) between 2008-2018. Methods: We performed a cohort study using a clinical database of infants from NICUs managed by the Pediatrix Medical Group. We included infants who were discharged between 2008-2018, <27 weeks gestation, and cared for at just one hospital. We excluded infants with major congenital anomalies and those who died on postnatal day 0 or 1 or were transferred to another hospital prior to discharge. We used a nonparametric trend test to evaluate indomethacin prophylaxis, defined as indomethacin exposure on postnatal day 0 or 1, over time. We also evaluated change in indomethacin prophylaxis over time by center among centers who contributed at least 10 infants during 4 time periods: 2008-2010, 2011-2013, 2014-2016, and 2017-2018 (Figure 2). We compared infant demographics and the prevalence of Grade III and IV intraventricular hemorrhage (IVH) of infants who did and did not receive indomethacin prophylaxis using chi-squared tests, and stratified our analysis by gestational age. Results: We identified 19,715 infants from 213 NICUs. The median (25th-75th percentile) gestational age and birth weight were 25 weeks (24-26) and 726 g (610-850), respectively. Of the 19,715 infants, 3246 (16%) received indomethacin prophylaxis. The median birthweight was 730 g (613-850) for the group that did not receive prophylaxis and 700 g (598-820) for infants who received indomethacin prophylaxis (P<0.001). The prevalence of indomethacin prophylaxis was 16% in 2013 in and 10% in 2018, but the decrease was not significant (p=0.23). However, indomethacin prophylaxis did significantly decrease from 2014 to 2018 (p=0.046) (Figure 1). 12 centers with the greatest change in indomethacin prophylaxis incidence were extracted and indomethacin prophylaxis prevalence and grade III and IV hemorrhage were stratified by year. Of 74 centers, 15 (20%) increased use of indomethacin prophylaxis between the oldest and most recent time periods, 42 (57%) decreased use, and 17 (23%) had no change. Of the 12 centers with the largest relative change in prevalence of prophylaxis, 50% showed an inverse correlation between indomethacin prophylaxis prevalence and IVH, while 50% showed a positive correlation (Figure 2). Conclusion: In this cohort, the receipt of indomethacin prophylaxis decreased but not significantly and no change in IVH was found. Both varied by center. Future studies should evaluate factors associated with changes in practice at different centers, and whether such changes impact infant outcomes.