854. The Impact of the CMS SEP-1 Core Measure on Antimicrobial Utilization: a Multicenter Interrupted Time-Series (ITS) Analysis

Abstract Background Hospitals began reporting the SEP-1 Core Measure to CMS in October 1, 2015, to promote the use of best practices for patients with sepsis. The impact of SEP-1 on overall antimicrobial utilization (AU), a potential unintended consequence, is unclear. Methods We performed an ITS analysis to evaluate changes in antimicrobial utilization after SEP-1 implementation. AU was measured as days of therapy (DOT)/1,000 days present (dp) for all adult inpatients who spent more than 24 hours in 18 hospitals in the southeastern United States. The 12-month period from October 1, 2014 to September 30, 2015 was defined as the “pre” period. After a 1-month wash-in, the 12-month period from November 1, 2015 to October 31, 2016 was defined as the “post” period. AU was aggregated by hospital by month for inpatient units. Total AU and NHSN AU categories were analyzed separately. ITS was modeled using a segmented regression analysis through a GEE model with negative binomial distribution and log link. Results A total of 362,460 patients had 688,583 DOT pre-SEP1 (mean 1.9 DOT/admission), and 291,884 patients had 530,382 DOT post-SEP1 (mean 1.8 DOT/admission). The diagnosis of sepsis (3.1%) and median length of stay (3, IQR 2–4) were unchanged after SEP-1. Utilization of combined vancomycin and piperacillin–tazobactam (P-T) increased 17% at SEP-1 implementation but this increase was not statistically significant (Table). Overall AU, anti-MRSA agents, and anti-pseudomonal agents were unchanged after SEP-1 (figure, table). Conclusion Implementation of the CMS SEP-1 measure did not lead to higher rates of AU in our cohort of hospitals, although this study did not assess adherence to SEP-1. Further research is needed to improve the use of antimicrobial therapy in hospitalized patients with suspected sepsis. Disclosures All authors: No reported disclosures.