J. C. Apfeld1, Z. J. Kastenberg1,5, K. G. Sylvester1,2,3,4, H. C. Lee3,4 1Stanford University School Of Medicine,Department Of Surgery,Palo Alto, CA, USA 2Stanford University School Of Medicine,Center For Maternal And Fetal Health, Lucile Packard Children’s Hospital,Palo Alto, CA, USA 3Stanford University School Of Medicine,Department Of Pediatrics,Palo Alto, CA, USA 4Stanford University,California Perinatal Quality Care Collaborative (CPQCC),Palo Alto, CA, USA 5Stanford University School Of Medicine,Center For Health Policy/Center For Primary Care And Outcomes Research,Stanford, CA, USA
Introduction:
There has been significant expansion in low-level (AAP 2007 designation IIA,IIB) and midlevel (IIIA,B) neonatal intensive care units (NICUs) in recent decades. Previous literature has established the relationship between case-specific surgical volume and outcomes. It is also widely recognized that very-low-birthweight (VLBW) infants experience better outcomes at higher-volume/level-of-care NICU’s. The relationship between expertise of the center and neonatal surgical volumes for specific anomalies is less well established. We sought to determine the relationship between outcomes for infants born with gastroschisis and volume/level-of-expertise of the treating center.
Methods:
A retrospective cohort study was conducted of NICUs in California. We used data from the California Perinatal Quality Care Collaborative from 2008-2014 to assess outcomes among a population-based sample of 1588 gastroschisis infants, according to levels of NICU care. Birth at each respective AAP Level was examined, alongside different birth-hospital volume measurements. SAS 9.3 was used for univariate and multivariate analysis to examine mortality, total days on ventilation, and total length of stay.
Results:
1588 total infants were born with gastroschisis between 2008-2014, 146 at NICUS with AAP Level IIA or B, 133 at Level IIIA, 633 at IIIB NICUs, and 666 at IIIC NICUs. Infants born at Level IIIA NICUs had higher adjusted odds of acute transfer to a higher NICU level at 14.4 (95% CI, 8.8-23.5) and adjusted odds of death at 2.2 (CI, 0.8-6.2), while ORs for mortality at level IIIB were 1.5 (CI, 0.7-3.0) and at level IIIC were 0.5 (CI, 0.2-1.1). In multivariate linear regression, infants born at Level IIIB NICUs had longer time on a ventilator by 1 days (p=0.04) and longer total length of stay by 6 days (p=0.01). Births at Level IIA/B had a significantly longer time on a ventilator at 5.2 days (p<0.001) and longer length of stay at 5.5 days (p=0.16). IIIC NICUs had shorter time on a ventilator by 2.1 days (p<0.0001) and a shorter length of stay by 6.6 days (p=0.005). Lower hospital NICU volume and VLBW volume trended towards higher mortality, more days on ventilation, and longer length of stay, but were not statistically significant predictor variables. Hospital volumes for both gastroschisis births and repairs were not associated with the outcome variables.
Conclusions:
Outcomes for gastroschisis in the state of California varied by NICU AAP Level designation and were less dependent on annual hospital gastroschisis births and repair volume. We also find that AAP NICU designation is more important then either overall NICU volume, or VLBW volume as a surrogate for experience with complex newborn care. Taken together, these data suggest that neonatal outcomes are sensitive to the level of comprehensive care provided in the NICU for congenital surgical anomalies of moderate technical complexity like gastroschisis.