番茄社区

TORONTO -- Physicians should be cautious about prolonged use of antibiotics in premature babies who don't have proven sepsis, researchers said here.

Antibiotics are appropriate for culture-proven sepsis but are also widely used when no pathogen can be isolated, according to Karen Puopolo, MD, PhD, of the Children's Hospital of Philadelphia (CHOP).

But late-onset sepsis and late onset culture-negative infection -- LOS and LOCNI -- are not the same thing, Puopolo reported at the Pediatric Academic Societies (PAS) annual meeting.

An analysis of 3,940 infants with either condition, or neither, showed markedly different risks of death and neurodevelopmental impairment, suggesting that treating LOS and LOCNI in the same way is wrong, she stated.

"We should be increasingly thoughtful about the way we treat these infants," she said. "They might just need more from us than empiric administration of antibiotics."

A related study of infants with LOCNI by the same authors, found wide variation in how they were treated, but outcomes were not different regardless of whether they were given antibiotics.

In 5,807 infants born prematurely from 2006 to 2014 in 24 centers affiliated with the , deaths and neurodevelopmental impairment rates in the wake of LOCNI were the same regardless of antibiotic use, according to Sagori Mukhopadhyay, MD, also of CHOP.

But "in the absence of benefit," there is nothing to counterbalance the "potential risks of antibiotic exposure," she said.

The studies underline the knowledge gap around culture-negative disease, commented PAS session co-moderator Paul Spearman, MD, of Cincinnati Children's Hospital Medical Center.

"The main take-away is that we need to understand better what they're calling culture-negative infections because the outcome is very, very different -- it's much better than proven sepsis," he told 番茄社区. "It's very likely that in most of those cases there's an alternate cause that's not bacterial sepsis. "It's showing that we need better markers ... and to understand what looks like an infection when it's not bacterial sepsis."

The second study, he noted, adds to the need for better understanding, if only because it suggests the clinical practice varies widely from place to place. "A long duration of antibiotics that differs from center to another is not necessarily a good thing," he said, "but they didn't show it was a bad thing."

It might be that LOS and LOCNI are the same thing but for some reason a pathogen simply can't be found in patients with the latter condition. In that case, serious outcomes ought to be the same, Puopolo's group hypothesized.

They analyzed records of the 24 centers affiliated with the Neonatal Research Network, and found data on 3,940 infants born prematurely between 2006 and 2014 who had LOS, LOCNI, or neither. They specifically excluded many children with both conditions or with other serious infections such as necrotizing enterocolitis.

The primary endpoint was the composite of death and neurodevelopmental impairment at 18 to 26 months of corrected age. The investigators also looked at those endpoints separately.

For the primary endpoint, they found statistically significant differences for:

• Adjusted relative risk for babies with LOS vs those with neither condition: 1.29
• Babies with LOCNI vs neither condition: 1.13
• LOS vs LOCNI: 1.14

When they looked at the components of the endpoint separately, they found that infants with LOS had a significantly higher risk of death but not of neurological impairment compared with babies who had neither condition. The pattern was reversed for LOCNI compared with neither -- a significant risk of impairment but no difference in the risk of death.

And finally, babies with LOS had a greater risk of death than those with LOCNI, but there was no difference in the risk of impairment.

The findings suggest that "LOCNI is not simply LOS in which cultures failed to identify an infecting organism." Puopolo said. Yet her group estimated that about 70% of the antibiotics used in the study cohort were given to babies with LOCNI.

Possible etiologies for the outcomes need more study, Puopolo said, noting that "dysbiosis" caused by the antibiotics themselves might be responsible or the drugs might be replacing a more appropriate therapy.

Comment