115 - Can a Clinical Care Bundle Protect Against NEC in the NICU?

Publication Number: 4113.115

Heather E. Croy, Sidney Kimmel Medical College at Thomas Jefferson University/duPont Hospital for Children, Philadelphia, PA, United States; Amy B. Mackley, Christiana Care Health System, Newark, DE, United States; Kelley Kovatis, ChristianaCare, Newark, DE, United States

Background: Necrotizing Enterocolitis (NEC) affects 5-7% of premature infants. The etiology of NEC is multifactorial but recent studies suggest that the microbiota may play a role. A Clinical Care Bundle (CCB), designed to reduce the risk of NEC, was implemented after the incidence of NEC in our unit increased from 0.1 cases per 100 VLBW infant days to 0.6 cases per 100 VLBW days. Understanding the impact of a CCB on the intestinal microbiota of infants admitted to the Neonatal Intensive Care Unit (NICU) may provide insight into how practice changes impact the microbiota.

Objective: The primary aim is to compare the intestinal microbiota of infants born pre and post CCB implementation. We hypothesize, infants with stool collected after the CCB will have increased diversity.

Design/Methods: The study was approved by the Christiana Care Institutional Review Board. Stool samples collected from infants born at ≤32 weeks born before the CCB implementation (Oct 2023) were compared to stool samples collected from infants born after implementation. Stool samples were collected at 3 time points: 0-2 days, 2 weeks, and 1 month. The stool was then transferred to a -80 ºC freezer until processing. Demographic and clinical characteristics were summarized using descriptive statistics. DNA extraction and sequencing were completed using PCR-based16s ribosomal targeted metagenomic sequencing (16s rRNA amplicon-seq). Advanced negative-binomial generalized linear model statistical analysis was performed.
Definitions: Alpha diversity measures the total number of microbial species. Beta diversity describes the microbiota composition shifts between groups.

Results: There was no difference in baseline demographic and clinical data before and after the CCB (Table 1). Alpha diversity increased over time in both the pre and post CCB groups. No difference was found in the alpha diversity when comparing the pre and post groups at equivalent time points (Fig 1). Beta diversity shifted significantly between the Pre and Post CCB groups. Clostridium, a bacteria genus believed to be protective against NEC, was observed with more frequent abundance in the post CCB group (Fig 2). A more normal and balanced distribution of bacterial abundance that included Enterobacteriaceae, a species that represents normal gut flora, was observed more frequently in the post CCB group (Fig 2).

Conclusion(s): Implementation of a CCB favorably affects the composition of the microbiota in premature infants admitted to a level 3 NICU. This correlated with decreased NEC rates in our unit, suggesting that CCBs that target the microbiome may improve neonatal outcomes.

Table 1


Figure 1
Figure 1. Alpha Diversity comparison between all three timepoints (1, 2, and 3) using Chao-1 index in the Pre- and Post-CCB implementation groups. Chao-1 index measures the estimated total number of taxa in the given microbiota. **: p <= 0.01, ***: p <= 0.001, ****: p<= 0.0001; Tube-group: matched sample group IDs using various clinical factors.

Figure 2
Figure 2: (A) Microbial compositions of 7 different types of microbiota communities (cm1 ~ cm7) using a Dirichlet-Multinomial Mixture model (DMM) based on the most abundant microbial taxa. The estimated relative abundance of individual microbial taxon was based on the fitted values in the DMM model. Only the aggregated compositions at taxonomic family level is shown for visualization purpose. (B) Heat map of predicted enterotype across each time point Pre and Post CCB implementation. The enterotype is represented by the predicted, most likely, microbial community using the DMM model. Matched patients from the same groups using various clinical factors are organized in the same rows.