Optimizing antibiotic therapy to the shortest effective duration is an important antimicrobial stewardship (AS) intervention. However, measuring and tracking duration of therapy for hospitalized patients is labor intensive because an antibiotic course often spans the inpatient and outpatient settings, each with their own unique data sources. The National Healthcare Safety Network (NHSN) antimicrobial use (AU) option facilitates automated, inpatient AU collection, but is unable to capture outpatient antibiotic use, connect AU to an indication, or calculate duration of therapy. An automated, reproducible tool capable of calculating duration would be valuable to identify and track AS interventions.
We developed a visual analytic tool that analyzes inpatient & outpatient, patient-level AU for pneumonia (PNA), urinary tract infection (UTI), and skin and soft tissue infection (SSTI) and is able to calculate duration. Inpatient AU is calculated based on barcode administration data in accordance with the NHSN AU option. Outpatient AU is calculated based on discharge prescription data. The tool is currently being tested by AS teams at 8 VA hospitals. Participating AS teams discuss application to AS monthly as a part of a learning collaborative. We report the calculated 2015 and 2016 durations for PNA, UTI, and SSTI.
The median duration in days across the 8 sites for SSTI was 13.6 (interquartile range [IQR] 12.1-15.0, range [R] 11.3-16.7) in 2015 and 13.0 (IQR 11.7-15.6; R 10.7-16.7) in 2016. The median duration for PNA was 11.2 (IQR 9.0-14.0, R 8.7-15.6) in 2015 and 10.6 (IQR 10.1-11.7, R 9.2-14.8) in 2016. The median duration for UTI was 10.6 (IQR 9.3-12.8, R 8.8-14.7) in 2015 and 12.0 (IQR 8.7-13.9, R 8.0-15.8).
There was inter-hospital variability in treatment durations for each infection. Durations were overall in excess of evidence-based recommendations across all hospitals. The tool provided an automated method to analyze syndrome-specific treatment durations allowing our learning collaborative to identify duration as a target for AS intervention. Additionally, the tool will allow us to efficiently and reproducibly track the impact of AS interventions on duration over time.
M. Goetz, None
K. Madaras-Kelly, None
M. Jones, None
P. Glassman, None
E. Spivak, None