National Trends in Oral Antibiotic Prescribing in United States Physician Offices from 2009 to 2016.

Background: Prior studies have found that outpatient antibiotics are commonly prescribed for non-bacterial conditions. It is unclear if national prescribing has changed in recent years given recent public health and antimicrobial stewardship initiatives. This study aimed to describe antibiotic prescribing in United States (U.S.) physician offices.

Materials/methods: This was a cross-sectional study of all sampled patient visits in the Centers for Disease Control and Prevention's National Ambulatory Medical Care Survey from 2009 to 2016. Antibiotic use was defined as at least one oral antibiotic prescription during the visit as identified by Multum code(s). Patient visits were categorized by U.S. geographic region and season. ICD-9-CM and ICD-10 codes were used to assess diagnoses and categorize antibiotic use as appropriate, possibly appropriate, or inappropriate.

Results: Seven billion visits were included for analysis, with 793,415,182 (11.3%) including an antibiotic. Prescribing rates were relatively stable over the study period (102.9-124.9 prescriptions per 1000 visits); however, 2016 had one of the lowest prescribing rates (107.7 per 1000 visits). The most commonly prescribed antibiotic class was macrolides (25 per 1000 visits). The South region and winter season had the highest antibiotic prescribing (118.2 and 129.7 per 1000 visits, respectively). Of patients who received an antibiotic, 55.9%, 35.7%, and 8.4% were classified as inappropriate, possibly appropriate, and appropriate, respectively. The most common conditions in which antibiotics were prescribed inappropriately included those with no indication in any of the predefined diagnosis codes (40.1%), other skin conditions (17.3%), and viral upper respiratory conditions (13.3%).

Conclusions: There was no significant reduction in outpatient antibiotic prescribing rates among U.S. outpatients from 2009 to 2016 and prescribing varied by region and season. These data suggest that more than half of antibiotics were prescribed inappropriately, with the majority of antibiotics prescribed with no indication. However, these findings need to be confirmed with robust prospective studies.