Harvesting the Low Hanging Fruit: A Benchmarking Tool for Implementation of Intravenous to Oral Antibiotic Switch Programs in 14 Southeastern Community Hospitals

Background: Intravenous to oral (IV-PO) switch policies are a simple antimicrobial stewardship intervention designed to improve patient care and save money. Comparison of measurements of oral and total utilization of targeted, high bioavailability antimicrobials in individual hospitals may help improve implementation of IV-PO switch policies.

Methods: We performed a retrospective cohort analysis of 14 hospitals enrolled in the Duke Antibiotic Stewardship Outreach Network to evaluate IV-PO switch efficiency for targeted drugs: azithromycin, ciprofloxacin, levofloxacin, moxifloxacin, doxycycline, linezolid, metronidazole, and voriconazole.  Electronic medication administration records from inpatient units for calendar year 2014 were used to calculate the proportion of oral (PO)/total days of therapy (DOT) by targeted antimicrobial treatment course. Descriptive statistics were used to evaluate PO/total DOT by hospital and antimicrobial. Negative binomial regression was used to evaluate the effect of IV-PO switch policy and other possible predictors.

Results: Seven of 14 hospitals had an IV-PO switch policy. A total of 57,413 courses of targeted antimicrobials were evaluated. PO route was used for roughly a third of antibiotic treatment days and ranged widely among hospitals (median 0.33, range 0.22-0.49). Distributions of PO/total DOT also varied by targeted antimicrobial. Doxycycline was most given by PO route (PO/total DOT 0.77); levofloxacin was least given by PO route (PO/total DOT 0.24). Treatment courses in hospitals with existing IV-PO switch policies had significantly greater PO/total DOT than those without a policy, although the effect was small (0.34 vs. 0.33, rate ratio 1.03, 1.00-1.06, p=0.04). Other significant predictors of PO/total DOT included hospital, antimicrobial, intensive care unit, and length of stay.

Conclusion: Overall, PO antimicrobial administration accounted for a third of DOT in this community hospital cohort.  Hospitals with existing IV-PO switch policies more often used PO antimicrobials.  While comparison studies such as this can help community hospitals assess use of IV-PO switches, hospital-level measures will require additional risk adjustment before interpretation.