PHILADELPHIA – A weeklong course of empiric antibiotic treatment in neonates may result in a higher subsequent rate of chronic lung disease, compared with infants treated for just 2 days, based on an association seen in a review of more than 900 very low birth weight newborns.
But this finding is not yet ready to definitively guide practice, as it came from a nonrandomized, retrospective study that may have failed to control for all possible confounding variables, Dr. Alexandra Novitsky said at the annual meeting of the Eastern Society for Pediatric Research.
Dr. Alexandra Novitsky
In her adjusted analysis, which controlled for several baseline variables, the 159 very low birth weight (VLBW) neonates who received a "long" course of empiric antibiotic treatment, usually for 7 days, had a statistically significant, twofold higher rate of also having chronic lung disease during follow-up, compared with the 747 neonates who received a "short" antibiotic course, usually for 2 days, said Dr. Novitsky, a neonatologist at Christiana Hospital in Newark, Del.
"It’s too early to draw conclusions about changing therapy," based on this finding, said Dr. David A. Paul, associate director of neonatology at Christiana Hospital, who collaborated on the analysis. "We did our best to control for possible confounders, and it still suggested that longer antibiotic treatment altered outcomes, but there may have been things [for which] we did not control," he said in an interview. The next step is to design a prospective study and determine if changing the duration of empiric antibiotic therapy changes outcomes. "But the current findings raise concern that we should be cautious about the duration of treatment," said Dr. Paul.
In the cases reviewed, each physician delivering care determined the duration of antibiotic treatment. Some may have opted for a longer course of treatment because they were concerned that "not all babies have culture-proven sepsis," Dr. Paul said in an interview. The physicians "may have feared that the babies had infections that were missed in their blood cultures. They treated presumed sepsis," he said.
Dr. Novitsky reviewed all the VLBW infants seen in the neonatal ICU at Christiana Hospital between July 2004 and June 2009. The regimen used on all neonates who received empiric antibiotic treatment consisted of ampicillin and gentamicin. The infants who received a longer antibiotic course had a significantly higher prevalence of several markers of a worse clinical profile, including lower birth weight, younger gestational age, a higher score for neonatal acute physiology (SNAP), and a lower 5-minute Apgar score. They also had higher rates of clinical chorioamnionitis, mechanical ventilation, and endotracheal tube colonization (endotracheal tubes underwent routine, weekly colonization assessments).
The infants who received a longer course of antibiotics also had a higher prevalence of antibiotic-resistant, gram-negative organisms colonizing their endotracheal tubes, a 6% rate, compared with a 2% rate among the infants who received a short course of treatment – a significant difference. The two groups of infants had roughly identical prevalence rates of colonization with antibiotic-resistant gram-positive strains.
Dr. Novitsky defined the primary outcome evaluated in the analysis, chronic lung disease, as the need for supplemental oxygen by the infant at 36 weeks postmenstrual age. This outcome occurred in 185 of the 906 (20%) neonates in the study: 17% of the infants who received a short antibiotic course, and 36% of those who received a long course.
The multivariable analysis adjusted for differences in gestational age, SNAP score, Apgar score, maternal antibiotic treatment, chorioamnionitis, pre-eclampsia, cesarean delivery, prolonged rupture of membranes, and need for mechanical ventilation. After adjustment, the two patient groups failed to show a significant difference in their rates of necrotizing enterocolitis or sepsis.
To further examine the relationship between duration of antibiotic treatment and chronic lung disease, Dr. Novitsky also presented the results of a subgroup analysis that focused on the 418 high-risk neonates in her study group, because of their delivery at 28 weeks’ gestation or younger and their SNAP score of 8 or greater. Within this subgroup, the adjusted rate for developing chronic lung disease ran 70% higher in the 108 infants who received a long course of antibiotics, compared with the 310 who received a short course, also a significant difference.
Dr. Novitsky and Dr. Paul said they had no relevant financial disclosures.