New findings from the NeoIPC Colonisation Surveillance study, published in JAMA Network Open, show that the more babies on a neonatal unit are already carrying resistant bacteria, the more likely other babies are to pick them up too, strengthening the case for infection prevention and control (IPC) strategies aimed at the whole unit, not just the highest-risk infants.
Premature or critically ill babies often spend weeks in hospital, where they can pick up bacteria from their surroundings, including bacteria resistant to common antibiotics. Before making a baby sick, resistant bacteria usually colonise the gut or skin silently. This silent colonisation is a common precursor to infection, and a potential source of spread to other babies on the unit.
Doctors have long suspected that “colonisation pressure”, meaning how many babies on a unit already carry resistant bacteria at any given time, raises this risk further, but solid evidence from newborn units has been limited, particularly for bacteria carrying extended-spectrum beta-lactamase (ESBL) resistance genes, now one of the most common forms of antibiotic resistance seen in hospitals.
To examine this, the NeoIPC Colonisation Surveillance study (or feasibility study), sponsored by Penta Foundation and conducted in collaboration with City St George’s, University of London, was conducted in 24 neonatal units across eight European countries between January 2022 and June 2024. Across 943 babies, the research team tracked which previously uncolonised babies went on to acquire ESBL-harbouring bacteria, and compared this against how many other babies on the same unit were already colonised at the prior survey.
ESBL genes turned out to be the most commonly detected resistance genes, found in 14% of all stool samples, though this varied widely between units, from 0% to 39% of babies colonised depending on the hospital. Among babies being followed for new acquisition, 17% picked up ESBL-harbouring bacteria during the study.
Critically, the more ESBL-colonised babies there were on a unit, the higher the chance that other babies would go on to acquire ESBL bacteria themselves: for every 10-percentage-point increase in the proportion of colonised babies on the unit, an individual baby’s odds of acquiring ESBL bacteria rose by roughly 70-90%, independent of that baby’s own prematurity, length of hospital stay, or antibiotic use.
This effect also held regardless of gestational age: colonisation occurred in babies both above and below 32 weeks, showing that even babies who are not in the highest-risk group for severe infection still contribute to unit-wide risk. Together, these findings suggest that how many colonised babies are on a unit at any given time is, in itself, an important driver of acquisition risk.
In practical terms, this means IPC shouldn’t be targeted only at babies who seem most vulnerable to infection. Colonisation pressure gives units a practical, monitorable signal to act on, for example by intensifying IPC practices when unit-wide colonisation is running high.
Importantly, the results suggest IPC doesn’t need to wait for infection. Because colonised babies raise the risk for everyone around them, even when they show no symptoms themselves, reducing colonisation may protect both the individual baby and the wider unit. This reframes infection prevention as a collective effort, rather than one narrowly targeted at the infants judged individually highest-risk.