Antibiotic Prescribing for Respiratory Tract Infections in Urgent Care: A Comparison of In-Person and Virtual Settings

Abstract

Abstract Background Little is known about antibiotic prescribing for respiratory tract infections (RTIs) in virtual versus in-person urgent care. Methods In this retrospective study, we used electronic health record data from Cleveland Clinic Health System. We identified RTI patients via International Classification of Diseases, Tenth Revision, Clinical Modification, codes and assessed whether the visit resulted in an antibiotic. We described differences in diagnoses and prescribing by setting (virtual versus in-person). We used mixed effects logistic regression to model the odds of antibiotic receipt by urgent care setting. We applied the model first to all physicians and second only to those who saw patients in both settings. Results There were 69 189 in-person and 19 003 virtual visits. Fifty-eight percent of virtual visits resulted in an antibiotic compared with 43% of in-person visits. Sinusitis diagnoses were more than twice as common in virtual versus in-person care (36% vs 14%) and were associated with high rates of prescribing in both settings (95% in-person, 91% virtual). Compared with in-person care, virtual urgent care was positively associated with a prescription (odds ratio, 1.64; 95% confidence interval [CI]: 1.53–1.75). Among visits conducted by 39 physicians who saw patients in both settings, odds of antibiotic prescription in virtual care were 1.71 times higher than in in-person care (95% CI: 1.53–1.90). Conclusions Antibiotic prescriptions were more common in virtual versus in-person urgent care, including among physicians who provided care in both platforms. This appears to be related to the high rate of sinusitis diagnosis in virtual urgent care.