186 - Prescriber-Led Antibiotic Timeout Alert Responses: A Retrospective Multicenter Study

Background: An antibiotic time out is when a clinician is prompted to reevaluate antibiotic therapy for appropriateness after 48-72 hrs. The purpose of this study is to evaluate the impact of a prescriber-led electronic antibiotic timeout alert.

Methods: This was a retrospective cohort study conducted at seven hospitals of a health system. The alert prompts prescribers to reevaluate antibiotic therapy after > 72 hours upon opening the electronic chart. Prescribers can then choose to “Continue”, “Renew”, “Discontinue” or “Defer” for each antibiotic. The alert fires for every time the chart is opened, for every prescriber, every time any antibiotic has been active for > 72 hours and no action response (i.e., “Continue”, “Renew”, “Discontinue”) has been selected. This was a 6 month study from November 1, 2018 – April 30, 2019. The primary outcome measure was the percentage of patients having > 1 antibiotic discontinued via the electronic antibiotic timeout alert. Secondary outcomes included frequency of timeout alerts per patient, and percentage of discontinuations per response, by drug, by indication and by patient age.

Results: A total of 213950 alerts fired for 13263 patients during the study period. 94.2% and 0.4% of responses were “Defer” and “Discontinue”, respectively. The average number of alerts per patient was 16.1.  The percentage of patients who had at least one discontinuation was 5.3%. 38.2% of patients had no response other than “Defer”. There were a total of 10306 action responses, 4.8% of which were “Discontinue”. Among action responses, discontinue rates for the indications of “COPD exacerbation”, “Prophylaxis, Surgical/Procedural”, “Prophylaxis, Immunosuppression”, “Osteomyelitis” were 9.6%, 12%, 1.3% and 1%, respectively. Azithromycin and trimethoprim/sulfamethoxazole had discontinuation rates of 9.3% and 0.9%, respectively. Patient age had a significant impact on discontinuation rates with rates of 4.7%, 3.5%, and 10.9% for patients > 18, 2-17 and < 2 years of age (p< 0.001).

Conclusion: An electronic antibiotic timeout alert implemented at seven hospitals was largely ignored. This alert may be refined by limiting it to antibiotic orders with specific features associated with higher rates of discontinuation.