968. Effect of Outpatient Antibiotic Ordering Restrictions on Antibiotic Prescribing Patterns at a State-wide VA Healthcare System

Abstract

BackgroundApproximately 30% of antibiotics prescribed in the outpatient setting are inappropriate, mostly due to unnecessary prescriptions (Rx) for upper respiratory infections. Ordering restrictions is one approach to curtail inappropriate use. However, this approach may cause unintended consequences, such as increases in Rx of higher level antibiotics. This study evaluated the downstream effect of an azithromycin (AZM) ordering restriction.MethodsThis was a pre–post evaluation of the impact of an AZM removal (October 2017) on prescribing patterns of common outpatient antibiotics at the VA Maryland Healthcare System. AZM restriction was placed >10 years ago for concerns of emerging AZM resistance and overuse. During the study period, fluoroquinolone (FQ) use was scrutinized due to increasing toxicity risk. The proportion of several outpatient antibiotic Rx were compared between October 2017 and September 2018 (FY17) and October 2018 and September 30, 2018 (FY18) using χ 2 and logistic regression. FQ and AZM Rx were also stratified by location of prescribing clinic (urban vs. rural) and duration (≤14 days vs. >14 days).ResultsThere were 15,972 and 14,451 prescriptions in FY17 and FY18, respectively. AZM Rx increased from 1,247 (7%) Rx in FY17 to 1,734 (11%) in FY18 (P < 0.0001) with an OR of 1.8 (95% CI 1.65–1.94). There was a greater effect on shorter than longer duration (OR 1.9 vs. 1.3, P < 0.0001), but no significant effect difference for urban and rural clinics (OR 1.8 vs. 1.9, P = 0.6). Conversely, FQ Rx decreased from 2,414 (15%) in FY17 to 1,731 (11%) in FY18 (P < 0.0001) with an OR of 0.7 (95% CI 0.66–0.76). There was a greater effect on shorter than longer duration (0.6 vs. 1.2, P < 0.0001) and also a greater effect on urban than rural clinics (OR 0.6 vs. 0.97, P < 0.0001). Doxycycline, amoxicillin–clavulanate and trimethoprim–sulfamethoxazole did not change significantly.ConclusionRemoval of AZM restriction led to a significant decrease in FQ Rx, with greater effect in shorter duration and urban clinics, and an increase in AZM Rx, with greater effect in shorter duration, but no difference in clinic setting. Disparity of rural prescribers needs further exploration, as do other interventions outside of restrictive ordering, which needs periodic evaluation of risk and benefit if implemented. Disclosures All Authors: No reported Disclosures.