Differences in US antibiotic prescription use by facility and patient characteristics: evidence from the National Ambulatory Medical Care Survey.

BACKGROUND: Antibiotic resistance is increasing, largely due to the overuse of antibiotics. Patient demographic characteristics can influence rates of antibiotic prescription, but less research has assessed the role of facility-level characteristics.

OBJECTIVE: The objective of this study was to examine the prevalence of antibiotic prescriptions for viral-like illness, as well as patient and provider factors that influence antibiotic prescription practices.

METHODS: We conducted an observational cohort study using data from the 2012 National Ambulatory Medical Care Survey. We compared the prevalence of antibiotic prescription for all aetiologies and for viral-like illnesses between community health centres, non-community health centre clinics and emergency departments. Then, we used logistic regression models to compare the odds of antibiotic prescription use by facility and patient characteristics.

RESULTS: Data came from 630 community health centre visits, 857 non-community health centre outpatient clinic visits and 627 emergency department visits. Compared to patients visiting non-community health centre clinics, patients visiting community health centres and emergency departments for any aetiology were more likely to receive antibiotic prescriptions. Patients with viral-like illnesses were less likely to receive antibiotics at community health centres and exhibited similar odds of receiving antibiotics at emergency departments. Certain patient demographics (age, race/ethnicity and payment source) were associated with variation in overall antibiotic prescription, but these factors were mostly unassociated with antibiotic prescription for viral-like illnesses.

CONCLUSIONS: The care setting that patients visit may influence their odds of receiving antibiotics. Initiatives addressing overuse of antibiotics should be mindful of facility- and patient-based characteristics when designing interventions.