Evaluating the antibiotic spectrum index in a stewardship-focused clinical trial for childhood pneumonia

Résumé fourni par la source

Abstract Objective: The antibiotic spectrum index (ASI) outcome quantifies antibiotic exposure based on spectrum of activity. Our objective was to examine ASI as an exploratory outcome in the context of a recent stewardship-focused, clinical trial in childhood pneumonia that originally used a binary guideline-concordant outcome. Design: Secondary analysis of a randomized clinical trial. Setting: Two tertiary pediatric hospitals. Methods: Encounters were randomly assigned to clinical decision support (CDS) or usual care treatment arm. The ASI was calculated by summing daily ASI scores for each unique antibiotic administered. It was evaluated as a continuous and ordinal measure: No Antibiotics (ASI = 0), Narrow (1-2), Intermediate (3-4), Broad (5-7), and Very Broad (≥8). Proportional odds regression modeled the ordinal ASI outcome in the first 24 hours by treatment arm and compared to the guideline-concordance outcome. Results were stratified by emergency department (ED) disposition. We also conducted a longitudinal, descriptive analysis of day-to-day ASI for those with in-hospital dispositions. Results: We included 1027 encounters, 549 (53%) were randomized to CDS and 478 (47%) usual care respectively. ASI Category did not differ by treatment arm overall (Odds Ratio: 0.88[95% Confidence Interval: 0.70,1.09]), which mirrored binary guideline-concordance. Mean ASI was lower for concordant encounters (2.1 vs 8.4, P < 0.001) and across all ED dispositions. In the longitudinal analysis, there were 1137 day-to-day ASI comparisons, with only 7% representing spectrum escalations. Conclusions: The ASI outcome yielded similar results to a dichotomous concordance outcome. However, ASI provided more granular insights into antibiotic prescribing, suggesting ASI may be a useful outcome measure in future stewardship-focused trials.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.