786 - Extreme Prematurity: Survival, Outcomes, and Procedure Burden using National Database

Active interventions near the margins of viability are now shifting more towards 22-23 weeks through guidelines and improved care. However, survival data is limited to urban academic institutions and varies between 5 to 21% depending on the interventions. Knowing survival and associated morbidities at this margin of viability in community hospitals is essential in counseling families and shared decision-making. We hypothesize that using a national database of discharges from U.S. community hospitals will show increased survival of < 23 weeks gestation (GA) and comparable morbidities to other gestational ages at the margins of viability.

Objective:

To determine the overall survival, major complications, and procedures at each increasing week at the margins of viability using a national database.

Design/Methods: This is a retrospective cohort study of hospital discharge data of extremely preterm infants using the 2016 -2020 Healthcare Cost and Utilization Project National (Nationwide) Inpatient Sample (HCUP-NIS). Weighted analyses were used to produce nationally representative estimates. We divided the data into four groups: < 23-, 23-, 24-,
and 25 GA. The Cochran-Armitage test was used for yearly trends. Categorical variables such as demographics, hospital characteristics, neonatal complications, and procedures were compared using the Chi-square test. The p-value of < 0.05 was considered significant. 

Results:

Out of 36,335 ≤25 GA discharge records, 26,210 (72%) survived. Survival rates among < 23 GA increased by 51% (0.43 to 0.65/10,000 births) with no significant change in 23, 24, and 25 GA (Fig 1). There was an increase in survival by almost 20% with each additional week from 23 to 25 GA (Table 1).  < 23 GA had lower survival (5.8%), birth weights, and cesarean births as well as more births in urban nonteaching and medium-sized hospitals compared to the other GAs. Blacks were higher in < 23 and 23 GA compared to 24 and 25 GA. For survivors, neonatal complications and surgical procedures were decreased with each increasing week from 23 to 25 weeks GA (Table 2). Outcomes such as BPD, ROP, sepsis, procedures such as gastrostomy, and inguinal hernia repair were lesser for < 23 GA compared to 23 GA.

Conclusion(s):

In this national database, we found an increasing yearly trend with overall lower survival for < 23 GA. However, among survivors, we did not find poorer outcomes for < 23 GA compared to 23 GA.  We speculate that overall low survival is because of non-initiation of active management at < 23 GA and can improve by implementing institution-specific treatment guidelines for this population.