Background: The Centers for Medicare and Medicaid Services (CMS) began requiring hospitals participating in the Inpatient Quality Reporting Program to implement a sepsis management bundle (SEP-1), effective October 2015, to facilitate efficient, effective, and timely delivery of high quality sepsis care. This study uses data reported by hospitals to the Antimicrobial Use (AU) Option of the CDCs National Healthcare Safety Network (NHSN) to investigate whether an increase in broad-spectrum antibiotic use was observed in select adult inpatient acute care locations following implementation of SEP-1.
Methods: Rates of AU per 1,000 days present, were compared across two 12-month periods: 1) Oct.2014-Sep.2015 (pre SEP-1), 2) Oct.2015-Sep.2016 (post SEP-1). Patient care wards and intensive care units (ICUs), i.e. locations, were included if ≥10 months of data were reported to NHSN in each period. Rates were modeled using Poisson regression for two drug categories, each defined in accordance with CDCs Standardized Antimicrobial Administration Ratio (SAAR): Broad-spectrum agents used predominantly for hospital onset or multidrug resistant infections (BSHO) and anti-MRSA agents. VA hospitals were not required to implement this measure and were used as a control, where the proportion of non-VA locations with a statistically significant AU rate increase were compared to the comparable proportion of VA locations.
Results: A greater percentage of non-VA wards observed a significant increase in AU from pre SEP-1 to post SEP-1 compared with VA wards: 34% vs. 22% (BSHO), and 20% vs. 14% (anti-MRSA). Differences in these VA vs. non-VA percentages were not statistically significantly (P=0.10, 0.30). About the same percentage of VA and non-VA ICUs reported increases in AU.
Conclusion: Comparisons between VA and non-VA locations were not statistically significant; however, these data raise the question of whether the greater percentage of non-VA wards with increased antibiotic use may have been driven by SEP-1 implementation. The SEP-1 measure may have a larger impact in non-ICU locations due to the frequent use of broad-spectrum antibiotics in ICUs. A more detailed investigation and continued monitoring are needed to further explore this potential association.
K. Van Santen, None
M. Neuhauser, None
L. Hicks, None
D. Pollock, None
A. Srinivasan, None