ASSESSMENT OF ORAL ANTIBIOTIC PRESCRIBING PRACTICES IN WMED OUTPATIENT CLINICS
BACKGROUND: Inappropriate antibiotic use in the outpatient setting is a growing concern that can lead to microbial resistance, increased treatment costs, poor health outcomes, and increased hospital admission and readmission rates. National antibiotic prescribing data typically has a considerable time lag. Recent estimates of national oral antibiotic use based on data from 2010-11 reveal a rate of 506 antibiotics per 1,000 patients, demonstrating the need for a stewardship program to ensure appropriate antibiotic use. This study sought to establish the oral antibiotic prescribing rate within the WMed clinic outpatient setting using a novel method developed by our group, as well as identify facets of patient care that correlate with physician prescribing practices.
METHODS: Data were collected retrospectively from WMed’s electronic medical record (EMR) for encounters resulting in any prescription, dated January 1, 2012, to August 31, 2014. Global oral antibiotic use rates were calculated per 1,000 patients, based on WMed prescribing rates, to provide a metric comparable to published estimates. Antibiotic prescribing patterns were examined by comparing choice of agent and dosing regimen to published treatment guidelines. Simple Poisson regression was used on patient-level data to determine patient demographics predictive of count of antibiotics prescribed offset by number of patient clinic visits. Generalized logistic regression with repeated measures for patient encounters was used to analyze differences in antibiotic prescribing rates for various provider roles, as well as trends in prescribing across time.
RESULTS: Our sample included 57,125 clinic encounters for 8,329 patients. These encounters included 1,854 (3.25%) visits where at least one antibiotic was prescribed. Global oral antibiotic use was estimated at 380 antibiotics per 1,000 patients (95% CI: 370, 391).
When offset by number of clinic visits per patient, patient age, gender, race, and primary payer were found to be predictive of antibiotic utilization (all P<0.0001). Specifically, average antibiotic use was significantly greater for individuals between the ages of 18 and 53, females, of black race, or on Medicaid.
Physician Assistants (PA) and Nurse Practitioners (NP) were approximately 10.5 times more likely to prescribe an antibiotic than Faculty providers; residents were roughly 2.6 times more likely compared to faculty, and PA/NP about 4.1 time more likely than residents. Overall, 26.5% of antibiotics prescribed and 47.9% of dosing regimens were not aligned with published guidelines. Roughly 89.5% of discordant dosing was higher than recommended. Medical residents selected dosing regimens outside of recommended guidelines 51.4% of the time. Pediatric patients were frequently prescribed a non-recommended dose (98.2%), with most dosed too high (98.2%).
It was noted that antibiotic prescribing rates at WMed decrease over time. The odds of an oral antibiotic prescription were approximately halved by the end of our study period.
CONCLUSIONS: Our study found that the WMed antibiotic prescribing patterns are lower than those reported nationally. This research has highlighted areas in which to focus stewardship efforts. Further research is in progress that will provide a more complete picture of antibiotic prescribing practices at WMed.