14. Outcomes from the AHRQ Safety Program for Improving Antibiotic Use Across 439 Long-term Care Facilities

BACKGROUND: Implementing effective antibiotic stewardship programs (ASPs) in long-term care (LTC) settings is challenging. We present the results of an intervention intended to change the culture of antibiotic prescribing in 439 United States LTC facilities (LTCF). METHODS: The LTC Safety Program assisted LTCFs with establishing and implementing ASPs from 12/2018 to 11/2019. Through webinars held 1–2 times per month and other educational content, the Safety Program emphasized 1) the science of safety to improve teamwork and identify antibiotic-associated harm and 2) clinical best practices in making antibiotic treatment decisions. Content was organized using the Four Moments of Antibiotic Decision Making Framework (Figure 1). All staff (e.g., physicians, nurses, nurse assistants) were encouraged to participate. LTCFs submitted monthly antibiotic days of therapy (DOT), numbers of new antibiotic starts, urine cultures (UCX) ordered, Clostridioides difficile LabID events, and census data. Generalized linear mixed effects models were used to calculate pre-post intervention changes at bi-monthly intervals for antibiotic DOT, antibiotic starts and UCX, each per 1,000 resident-days (RD), and C. difficile LabID events per 10,000 RD, comparing the beginning (1/2019 and 2/2019) and end (11/2019 and 12/2019) of the Safety Program. Figure 1. Four Moments of Antibiotic Decision Making in the Long-Term Care Setting RESULTS: Of 439 LTCFs who completed the Safety program, the majority were mid-sized (75–149 beds; 229, 52.2%), most were non-hospital based and owned by a larger system (246, 56.0%), with similar distributions between urban and rural settings. Of these, 348 (79%) submitted both baseline and end-of-intervention data. Antibiotic starts decreased from 7.89 to 7.48 starts/1000 RD; P = 0.02). Days of therapy for all antibiotics decreased from 64.1 to 61.0 DOT/1,000 RD; P = 0.068) and for fluoroquinolones (an antibiotic targeted in the Safety Program) from 1.49 to 1.28 DOT/1,000RD; P=0.002. UCX decreased from 3.01 to 2.63 orders/1000 RD; P = 0.001). There were no significant differences in C. difficile LabID events Table 1. [Image: see text] Table 1. Changes from baseline (Jan-Feb, 2019) to the end (Nov-Dec, 2019) of the AHRQ Safety Program [Image: see text] CONCLUSION: By targeting both antibiotic prescribing culture and knowledge of best practices, the AHRQ Safety Program led to significant reductions in antibiotic use across a large cohort of LTCFs. DISCLOSURES: Morgan Katz, MD, MHS, AHRQ (Research Grant or Support)FutureCare Health Systems (Consultant)Roche (Advisor or Review Panel member) Robin Jump, MD, PhD, Accelerate (Grant/Research Support)Merck (Grant/Research Support)Pfizer (Grant/Research Support, Advisor or Review Panel member)Roche (Advisor or Review Panel member)