What Triggers an Antimicrobial Stewardship Program (ASP) Intervention in Pediatrics?
Background: ASPs are growing in pediatrics. Many programs function by reviewing a large number of antimicrobial prescriptions on a daily basis to identify opportunities to improve or modify prescribing, thus leading to an intervention or recommendation. Little is known about the frequency with which reviewed prescriptions lead to an intervention and the patient and clinical factors most strongly associated with an intervention. A better understanding of these factors could lead to more targeted ASP reviews and more efficient use of resources.
Objectives: Identify the antibiotics and clinical diagnoses most strongly associated with a pediatric ASP.
Methods: We reviewed the frequency and types of interventions made by a pediatric ASP across 5 years, from 3/2008 to 3/2013. Our program uses a prospective audit and feedback structure where prescriptions for any of 18 selected antibiotics are reviewed daily for potential interventions. Interventions were grouped into four categories: stop therapy, modify therapy (i.e. change antibiotic), optimize therapy (i.e. alter dosing or route of administration) and consult infectious diseases. We used a multinomial distribution model to determine the probability of each ASP intervention group, based on the specific antimicrobial agent or disease category.
Results: A total of 14,407 ASP reviews were included in our analysis. Among these, a total of 2,318 (16%) prompted an ASP intervention. The most common types of ASP recommendations were stop or modify therapy. The clinical diagnoses with the highest predicted probability of an intervention were community acquired pneumonia (CAP, 0.26), ear/nose/throat (ENT, 0.25), genitourinary (0.23), and respiratory infections (0.21) (Figure 1). The antibiotics with the highest predictive probability of an intervention were ceftriaxone (0.20), clindamycin (0.20), and gentamicin (0.19) (Figure 2).
Conclusion: We identified several clinical diagnoses and antimicrobials that are associated with higher than average likelihood of triggering an ASP intervention. This analysis will assist in enhancing our ASP to focus not only on specific antibiotics but to also target specific conditions for review and development of clinical practice guidelines.
A. L. Hersh, None
D. Yu, None
L. Stach, None
A. Myers, None
M. A. Jackson, None
J. Day, None
R. Mcculloh, None
J. Newland, Pfizer: Grant Investigator, Grant recipient