Antibiotics are ineffective treatment for viral syndromes, including COVID-19.
Was characterized antibiotic prescribing in older adults with outpatient COVID-19 visits to identify opportunities to improve prescribing practices.
Were used 100% Medicare carrier claims and Part D event files to identify beneficiaries with a COVID-19 outpatient visit and associated antibiotic prescriptions. We included beneficiaries aged 65 years and older who had fee-for-service plus Part D coverage and a visit during April 2020 to April 2021. Were identified telehealth visits with Current Procedural Terminology codes and in-person visits (office, urgent care, and emergency department [ED]) with place-of-service codes.
To identify visits with a primary diagnosis of COVID-19, were first limited them to those with International Classification of Diseases and Related Health Problems, 10th Revision diagnosis code U07.1. Were then excluded visits with additional diagnosis codes for conditions for which antibiotics are always or sometimes appropriate based on clinical guidelines using a previously described tiered system. Visits were then linked to an antibiotic if prescribed within 7 days before or after the visit. We reported visits by setting and month, including antibiotic classes.
Were performed 2-sided χ2 tests to compare the distribution of characteristics between beneficiaries with COVID-19 who were and were not prescribed an antibiotic by age, sex, race, and prescribers’ location. This study was deemed nonhuman subjects research by the Centers for Disease Control and Prevention and did not require institutional review board review. Analyses were performed in SAS version 9.4. P < .05 defined statistical significance.Results
During April 2020 to April 2021, 346 204 (29.6%) of 1 169 120 COVID-19 outpatient visits were associated with an antibiotic prescription, which varied by month, with higher rates of antibiotic prescribing occurring during a wave of COVID-19 cases during the winter of 2020-2021 (range, 17.5% in May 2020 to 33.3% in October 2020).
Prescribing was highest in the ED (33.9%), followed by telehealth (28.4%), urgent care (25.8%), and office (23.9%) visits. Azithromycin was the most frequently prescribed antibiotic (50.7%), followed by doxycycline (13.0%), amoxicillin (9.4%), and levofloxacin (6.7%). Urgent care had the highest percentage of azithromycin prescriptions (60.1%), followed by telehealth (55.7%), office (51.5%), and ED (47.4%).
Differences were observed by age, sex, and location. Non-Hispanic White beneficiaries received antibiotics for COVID-19 more frequently (30.6%) than other racial and ethnic groups: American Indian/Alaska Native (24.1%), Asian/Pacific Islander (26.5%), Black or African American (23.2%), and Hispanic (28.8%).
During the first year of the COVID-19 pandemic, 30% of outpatient visits for COVID-19 among Medicare beneficiaries were linked to an antibiotic prescription, 50.7% of which were for azithromycin. Randomized clinical trials demonstrated no benefit of azithromycin in treating COVID-19, and its use for the disease has been linked to antimicrobial resistance.
The largest number of visits and highest rates of antibiotic prescribing were observed in the ED, perhaps reflecting acuity of care, and urgent care centers had the highest rate of azithromycin prescribing. Telehealth visits had the second highest antibiotic prescribing rate and were close in volume to office visits, emphasizing the importance of optimizing antibiotic prescribing practices in this setting. Antibiotic prescribing occurred at a higher rate for non-Hispanic White beneficiaries than for those from other racial and ethnic groups. Although described in pediatrics, this racial difference has not been well characterized in older adults and warrants further evaluation because it may indicate more services are being provided to White beneficiaries, even if not indicated.
Study limitations include that, although only visits for which COVID-19 was the primary diagnosis were included and visits with a diagnosis code that putatively justified an antibiotic were excluded, misclassification was possible. Underlying chronic medical conditions or severity of illness and subsequent hospital admission were not controlled for. These data may not be representative of the entire US population nor adults aged 65 years and older without Medicare prescription drug coverage. Data since April 2021 were not available, and strong evidence against azithromycin use was not published until the end of the observation period.
These observations reinforce the importance of improving appropriate antibiotic prescribing in outpatient settings and avoiding unnecessary antibiotic use for viral infections such as COVID-19 in older adult populations.