Drew Schwartz (Washington University) presented a longitudinal study of roughly 2,500 fecal samples from 250 preterm neonates across three NICUs, showing frequent dominance of a single organism in many infants and a link between gut abundance spikes and subsequent bloodstream infections. "We can do this on a six to 12-hour time frame" using rapid Oxford Nanopore sequencing and computational analysis, Schwartz said, describing the potential to detect taxa and antibiotic-resistance genes quickly enough to influence empiric antibiotic selection.
Schwartz described nested case-control analyses where infants who later developed bacteremia showed elevated abundance of the causative species in the two weeks before infection and noted that in some cases the identical bloodstream isolate matched gut sequences. He proposed weekly or on-demand gut sequencing integrated into clinical records to provide individualized empiric antibiotic guidance and to improve stewardship by identifying infants unlikely to need gram-negative coverage.