NICU antibiotic restrictions lead to shorter treatment courses

Neonatal intensive care unit
Neonatal intensive care unit

Seaman Joseph A. Boomhower / US Navy

Shortening the duration of antibiotic therapy in a neonatal intensive care unit (NICU) led to a 27% drop in overall use and an increase in appropriate antibiotic treatment with no detriment to infant safety, according to a study yesterday in The Lancet Infectious Diseases.

A neonatology research team led by the University of Texas Southwestern Medical Center monitored antibiotic use in infants admitted to the Parkland Hospital NICU in Dallas from March through November 2012, as part of the Surveillance and Correction of Unnecessary Antibiotic Therapy (SCOUT) study.

Though antibiotic therapy is often necessary for preterm infants and those with a birth weight under 2,100 grams (approximately 4 pounds, 6 ounces), inappropriate use can lead to multidrug-resistant (MDR) infections, invasive candidiasis, bronchopulmonary dysplasia, necrotizing enterocolitis, and late-onset sepsis, the authors said.

Following the 9-month observation period, antibiotic stewardship interventions implemented between October 2013 and June 2014 targeted three indications for stopping or shortening antibiotic therapy: rule-out (ie, suspected) sepsis, culture-negative pneumonia, and culture-negative sepsis.

The three diagnoses are associated with often-inappropriate antibiotic use and represent cases in which interventions might have the most benefit. Numerous non-bacterial conditions can mimic sepsis, and in treating rule-out sepsis, "providers already viewed their [antibiotic] administration as unintentional," the authors said.

Similarly, no "diagnostic gold standard" exists for culture-negative pneumonia and sepsis in the NICU, frequently leading to antibiotic administration for ambiguous symptoms arising from viral and fungal conditions.

Shorter, improved use of common antibiotics

Researchers documented antibiotic use (ie, prescribed drug, dose, frequency, indication, duration, and discontinuation) in 2,502 infants in Parkland's NICU, 1,607 of whom were admitted during the baseline observation period and 895 during the intervention period.

The two primary reasons for NICU admission during the intervention period were respiratory distress in full-term infants (279 infants, or 31%) and pre-term birth at 34 weeks or less gestation (266, or 30%). Neither maternal nor infant characteristics differed significantly between the baseline and stewardship periods, the authors said.

Interventions sought to reduce inappropriate use of antibiotics in two ways: (1) Using the electronic medical record to automatically stop ampicillin, gentamicin, and oxicillin treatment for rule-out sepsis after 48 hours, requiring the prescriber to manually extend therapy if indicated; and (2) limiting the duration of antibiotic therapy for culture-negative pneumonia and culture-negative sepsis to five days.

Antibiotic use decreased by 70.9 days of therapy (DOT) per 1,000 patient-days after Parkland's NICU instituted the new protocol. During the 9-month intervention period, use decreased by 27%, from 343.2 DOT at baseline to 252.2 DOT, the authors said.

Reductions in drug use occurred in all targeted areas. The proportion of infants whose treatment course for rule-out sepsis was limited to 48 hours or less increased from 32% at baseline to 95%. The share of treatment courses for pneumonia or culture-negative sepsis that ended within five days increased from 36% to 72% and 31% to 62%, respectively. Only three infants during the intervention period had antibiotic courses extended within 14 days of discontinuation, the authors said, and no treatment extensions occurred for pneumonia or culture-negative sepsis.

The most significant changes in therapy were observed in the use of ampicillin, gentamicin, and oxicillin, which together accounted for 93% of antibiotic therapy administered at baseline, and decreased by 31%, 30%, and 24%, respectively.

Stewardship directs drugs to appropriate uses

Because data on antimicrobial stewardship in NICU environments are limited—and because requires care is required in balancing potential harms from overuse and underuse—"close monitoring of patient safety outcomes is a crucial adjunct to stewardship efforts," the SCOUT authors said.

The interventions had no significant effects on infant safety, the authors reported, but they noted that length of NICU stay significantly increased during the intervention period, from 7 to 8 days. A slight rise in mortality of 0.7% occurred during the intervention months (from 4 deaths during the observational period to 2 during the intervention), though colonization with MDR bacteria fell from 1.4% to 1% of infants.

Shortening and/or discontinuing treatment was also associated with an increase in appropriate use, with the proportion of antibiotics used appropriately for culture-proven bacterial infections or necrotizing enterocolitis rising from 6.9% to 10%.

"By thoroughly assessing all antibiotic use, we were able to establish which clinical scenarios were most amenable to stewardship interventions, design specific interventions to target those areas, and then track both subsequent antibiotic use and safety outcomes," the authors said But they cautioned that the results may not be entirely generalizable to other NICUs, which should gauge stewardship activities based on their own patient populations and prescribing practices.

Infection prevention should accompany stewardship

In a commentary on the SCOUT study published yesterday in The Lancet Infectious Diseases, Paolo Manzoni, MD, PhD, of Sant'Anna Hospital in Torino, Italy, and Alberto Dall'Agnola, MD, of Silvio Orlandi Hospital in Bussolengo, Verona, Italy, addressed the need for preventive action in the NICU as a necessary accompaniment to antimicrobial stewardship efforts.

SCOUT represents important progress in antibiotic use reduction for diagnoses (eg, rule-out sepsis) that may involve severe illness but are not backed up by reliable or readily available laboratory confirmation and guidelines that specify duration of prophylactic treatment, the two neonatologists said. Additionally, the ubiquity of conditions that mimic sepsis complicates the question of when and if to administer antibiotics.

Given the limited evidence for treatment of suspected infection or sepsis in the NICU, Manzoni and Dall'Agnola urge a stronger focus on preventing prophylactic antibiotic use through infection control. Practices appropriate for the NICU environment include feeding infants with maternal milk, preventing fungal infections with fluconazole, administering probiotics to prevent necrotizing enterocolitis, and using evidence-based guidelines for catheter and central-line care.

As a key part of stewardship in the NICU, "antibiotic use might be decreased by simply preventing infections," the authors said.

See also:

Jul 21 Lancet Infect Dis abstract

Jul 21 Lancet Infect Dis commentary

This week's top reads

Our underwriters