Download Full Text (1.1 MB)
Background: A broader understanding of antibiotic variation can inform stewardship efforts. Diversity indices are frequently calculated in ecologic studies to measure the variety of species in a given habitat, however they have not yet been used to study variation in antibiotic use in the healthcare setting.
Objective: To develop a novel assessment of antibiotic variation, based on the breadth of an antibiotic’s associated conditions as well as its use, measured by days of therapy (DOT).
Design/Methods: We assessed antibiotic use in children aged 60 days-18 years hospitalized at a US Children’s hospital included in the Pediatric Health Information System (PHIS) database. We identified common antibiotics, defined as those that account for >1% of total antibiotic DOT at >80% of hospitals. An antibiotic diversity index (ADI) was calculated for each antibiotic at each hospital using the number of conditions and DOT to mimic the Shannon-Weiner entropy index, an index used in ecology to measure the complexity of an ecosystem. Hospital-level variation in antibiotic use was measured by the spread of ADI scores across hospitals relative to the mean ADI (i.e. coefficient of variation [CV]).
Results: Table 1 shows commonly used antibiotics at 49 hospitals. Cefazolin has the highest mean ADI indicating it is used in the highest number of conditions. However, it also has the smallest CV indicating consistent use across hospitals. . Conversely, meropenem has a relatively small mean ADI but has the highest CV. This indicates that meropenem is used in a relatively small number of conditions but is used more variably across hospitals. Figure 1 shows the utility of the ADI to identify variation in distribution of both an antibiotic’s DOT and associated conditions. In most hospitals, use of cefazolin is focused consistently in 7-9 conditions while meropenem is used consistently across hospitals in 2-3 conditions.
Conclusion(s): ADI is a novel tool to measure variability in antibiotic use across hospitals. Comparisons of ADIs highlight antibiotics associated with unequal patterns of use and their associated conditions.
Presented at the 2021 PAS Virtual Conference
DePorre, Adrienne G.; Richardson, Troy; Newland, Jason; and Neuman, Mark, "A novel way to describe variation in antibiotic use" (2021). Posters. 188.