Characteristics of primary medical team disagreements with antimicrobial stewardship program interventions

Background: Antimicrobial stewardship programs (ASP) have been shown to reduce antibiotic use, development of antibiotic-resistant bacteria, and hospital costs. Previous research on prospective-audit-with-feedback (PAF) ASPs has characterized which antibiotics trigger recommendations, what type of recommendations are made, and how often primary medical teams agree with recommendations. However, factors involved with disagreements are poorly understood.

Objective: To identify characteristics when PAF ASP interventions are not agreed upon by the primary medical care team.

Methods: Using the ASP data repository at Children’s Mercy Hospital-Kansas City, data was extracted on ASP interventions in which disagreement occurred from 3/3/08 to 3/2/15. Demographic information, medical specialty team, area of the hospital, presence of complex medical conditions, antimicrobial triggering review, and length of stay were examined. Percent disagreement was compared using Pearson's chi-square and the Mann-Whitney was used to compare length of stay.

Results: Over the study period, a total of 16,347 patients were reviewed by the ASP; 3,144 (19.2%) had recommendations made; 761 (24%) interventions were not agreed upon. Patients with interventions not agreed upon had a significantly longer length of stay (14 v 9 days; P<0.001) and a higher prevalence of complex chronic conditions (27% v 18%; P<0.001). Disagreement was higher for patients seen in the NICU (32%; P<0.001), PICU (27%; P<0.001) and Heme/Onc (26%; P<0.001) floors when compared with General Pediatrics (17%). Disagreement was also higher among patients receiving amoxicillin/clavulanate (40%; P<0.001) and meropenem (31%; P<0.001) compared to cefotaxime (20%). An intervention to stop therapy had the highest level of disagreement, specifically when a viral infection had been identified (45%; P<0.001) or when no indication for antibiotics had been identified (40%; P<0.001), compared to broadening coverage (14%).

Conclusion: Complex medical care represents an area of significant disagreement with ASP recommendations. Subspecialty involvement in stewardship activities, tailoring interventions specifically for complex patients, and more research on appropriate antibiotic use in these settings could produce more effective stewardship strategies.