NEW ORLEANS – Escherichia coli has replaced Group B Streptococcus as the most common cause of bacteremia in febrile infants in a nationally representative, retrospective study.

Among the 183 positive blood cultures drawn from 179 febrile but otherwise healthy-appearing infants, 76 were E. coli (42%), 41 Streptococcus agalactiae (22%), 10 S. pneumoniae (5%), and 10 S. aureus (5%). Roughly half of the cultures were gram negative.

Notably, no cases of Listeria bacteremia were identified, Dr. Rianna Evans said during the plenary session at the Pediatric Hospital Medicine 2013 meeting.

"There has been a definite rise in ampicillin-resistant gram-negative pathogens, which is going to be affecting our cephalosporin resistance," she said. "There is not a national database yet on this that we know of, but we identified ampicillin-resistance patterns at several of our sites, and it’s been noted in other cities as well."

Results from the current study, conducted at six sites across the United States, are supported by prior regional studies indicating a shift in bacterial epidemiology from Group B strep toward gram-negative pathogens in young infants. A change in the types of pathogens being seen is expected given the introduction of universal Group B strep prophylaxis in pregnant mothers and several new vaccines in the past decade, said Dr. Evans, a hospitalist at Children’s Hospital of The King’s Daughters, Norfolk, Va.

The investigators analyzed positive blood cultures from infants, age up to 90 days, drawn from January 2006 through December 2012 at six sites: Illinois; Rochester and Albany, N.Y.; Virginia; Minneapolis; and Los Angeles. Cultures were excluded if they were drawn in the intensive care unit, from indwelling vascular catheters, or from infants with a history of major surgery.

In all, 79% of the infants were classified as non–low risk according to Modified Rochester Criteria, 71% were febrile at the time the culture was drawn, 49% had a concurrent urinary tract infection, and 13% had concurrent meningitis. Their average age was 39 days.

Infants with E. coli were significantly more likely to be non–low risk than low risk (P = .001), while 56% of those with S. aureus bacteremia had evidence of a skin and soft tissue infection, Dr. Evans said.

She recommended single coverage with third-generation cephalosporin in febrile, non–toxic infants and that clinicians consider S. aureus and Enterococcus spp. coverage in ill-appearing infants with evidence of skin and soft-tissue infections.

During a discussion of the results, an audience member said it was dangerous to propose practice change based on a retrospective study, even if it was multicentered. Dr. Evans agreed and said there was a lot of discussion about their recommendations.

"It’s a bold statement, and I know it’s not going to be taken universally, which is why we need a prospective study identifying this," she said. "But with our work and all of the other studies that have been done on a regional level, I think there’s good enough data to at least suggest this and take it forward with a prospective study to see if there are true differences."

Dr. Evans said they could expand the retrospective analysis to bring in other sites, possibly through coauthor Dr. Eric Biondi’s ongoing study evaluating the time to blood culture positivity.

As a result of increasing antibiotic resistance to several types of gram-negative bacilli, the Centers for Disease Control and Prevention has also set up the Multi-Site Resistant Gram-Negative Bacilli Surveillance Initiative (MuGSI) to track infections due to carbapenem-resistant organisms including E. coli, Enterobacter cloacae, Enterobacter aerogenes, Klebsiella pneumoniae, K. oxytoca, and multidrug-resistant Acinetobacter baumannii. Data from the initiative are expected this year.

In one of the regional studies from Kaiser Permanente Northern California, E. coli was identified in 56% of positive samples collected from 2005 to 2009, and again there were no cases of Listeria bacteremia (Pediatrics 2012;129:e590-6).

Clinicians should consider Listeria if the maternal history or cerebrospinal fluid is suggestive of meningitis, but may want to revise their thinking when teaching residents about the bacterial causes of serious bacterial infections, Dr. Evans suggested.

Finally, another audience member expressed concern about whether cephalosporin alone would be a safe choice, particularly by residents, given that there was a fair amount of enterococcus in the sample and residents often miss this and forget about ampicillin. Dr. Evans observed that this has been the most provocative point when discussing the results with other institutions, but she said enterococcus was present in just 4% of the study’s positive samples and prior studies have shown that only 2% of well-appearing children presenting with fever are expected to have bacteremia. "It’s still a fairly low number, and you’re seeing a higher number of kids with ampicillin-resistant gram-negative rods," she added.

The meeting was sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association. Dr. Evans and her coauthors reported having no financial disclosures.

pwendling@frontlinemedcom.com

Bacterial epidemiology shifting nationally in febrile infants