Premature babies in neonatal care units are very sensitive and susceptible to potentially fatal illnesses.

To help keep these babies safe, the danger of infection must be kept as low as possible.

A particular problem is late-onset sepsis which starts from three days after birth, when bacteria get into the blood and grow.

This can be very dangerous and babies with late-onset sepsis end up staying in hospital longer, need more treatment with antibiotics and can be left with life-long effects on their health.

Bacteria from the Staphylococcus family are the most common causes of late-onset sepsis.

Most members of this large group of bacteria are harmless; they are normal colonisers of our skin, which can even protect us from harmful microbes.

However, some strains, when they end up in the wrong place and get inside the body, can cause major problems, particularly for immunocompromised individuals like neonatal babies.

Staphylococcus capitis is an example of this. This is a species that is usually content living on our scalp, face and neck; capitis means "of the head' in Latin.

Some strains of S capitis are however associated with late-onset sepsis. One particular strain, known as NRCS-A, has been identified as causing serious infections in neonates around the world.

Scientists think this strain first emerged in the 1960s and spread globally throughout the 1980s as it evolved resistance to the commonly used antibiotic vancomycin.

Strains circulating now show resistance to multiple antibiotics and reduced susceptibility to antiseptics that we use to sterilise the skin of babies.

This makes the bacteria harder to treat and control, but exactly why this NRCS-A strain has become so globally successful has remained a mystery.

To try and understand what makes this strain able to spread around the world and to develop better ways to keep it under control, Professor Mark Webber and his team from the Quadram Institute and the University of East Anglia analysed the genomes of hundreds of S capitis isolates.

They worked with two Neonatal Intensive Care Units (NICUs), one in the UK and one in Germany, obtaining samples of S capitis from the skin and gut of neonatal babies, with and without late-onset sepsis.

Their results, published in the journal Microbial Genomics, found that the NRCS-A strain was commonly carried on the skin and in the gut of uninfected neonatal babies, and that transmission between babies within NICUs was likely.

By reading the complete genome of each sample, the team was able to identify tiny genetic differences between the S capitis strains that caused disease and those that didn't.

"Studying how strains like NRCS-A have become globally successful is crucial to understanding how bacteria evolve to colonise different environments and to give us new ideas about how to reduce the risks of infection in vulnerable populations," said Professor Webber.

"We hope this work can be the starting point for more research to develop better ways to protect newborn babies from the terrible consequences of infection."