Antibiotic overuse in intensive care units is one of the most consequential drivers of antimicrobial resistance, and pediatric patients are not exempt. When critically ill children on ventilators develop airway infections, the instinct to treat aggressively and repeatedly culture is understandable — but this quality improvement study suggests that instinct may be precisely wrong, and that structured restraint is both safe and effective.
At a tertiary-care pediatric ICU, a clinical decision-support tool was deployed to standardize when endotracheal aspirate cultures should be collected, how ventilator-associated tracheitis (VAT) should be microbiologically confirmed, and — critically — how long antibiotics should run. Across 1,072 culture samples and 213 confirmed VAT episodes, the intervention produced a 23% drop in monthly culture rates (from 9.2 to 7.1 per 100 ventilator days) and a 53% reduction in serial repeat cultures. Antibiotic days of therapy fell by 53% (from 10.8 to 5.1 per 100 ventilator days), and the proportion of cases treated with a 3-day course surged 13-fold — from just 5% to 65% of episodes. Crucially, no increase in mortality, ventilator-associated events, ventilator days, or length of stay was detected across balancing measures.
VAT sits in a diagnostically ambiguous space between colonization and frank pneumonia, which historically encouraged overprescribing. This study joins a small but growing body of evidence — largely from adult critical care — suggesting that short-course antibiotic strategies in respiratory infections do not compromise outcomes. The pediatric-specific data here are particularly valuable given that children's ICUs have been underrepresented in stewardship literature. Limitations include the single-center design and quality-improvement methodology, which lacks a randomized control arm and cannot fully isolate the intervention's causal contribution from secular trends. Nonetheless, the magnitude of antibiotic reduction achieved without measurable harm is notable. For institutions still defaulting to open-ended VAT treatment, this protocol offers a replicable, nurse-championed model worth examining closely.