Uscher-Pines L, Mulcahy A, Cowling D, Hunter G, Burns R, Mehrotra A. Antibiotic Prescribing for Acute Respiratory Infections in Direct-to-Consumer Telemedicine Visits. JAMA Intern Med. 2015;175(7):1234–1235. doi:10.1001/jamainternmed.2015.2024
Direct-to-consumer (DTC) telemedicine companies provide consumers with around-the-clock access to care for common nonemergent conditions through telephone and live video visits via personal computers and mobile phone apps. Approximately 1 million DTC telemedicine visits between patients and physicians serving these companies, without an established relationship, were delivered in 2014.1
DTC telemedicine is often more convenient and less expensive than in-person visits. However, concerns about the quality of these services have been expressed2,3: lack of a physician-patient relationship and access to medical records; limitations of the physical examination; and barriers to testing could lead to overuse of antibiotics.
There have been few evaluations of DTC telemedicine quality. Using health plan claims, we compared antibiotic prescribing rates for acute respiratory infection (ARI) between Teladoc, a large DTC telemedicine company, and physician offices.
In April 2012, the California Public Employees’ Retirement System first offered Teladoc as a covered benefit. We limited the study population to members aged 18 to 64 years, who were continuously enrolled from April 2012 to October 2013 who had 1 or more ARI visits. This study was approved by the institional review board for RAND Corporation.
We identified ARI visits using International Classification of Diseases, Ninth Revision diagnosis codes based on prior methods.4 We eliminated follow-up visits at any site within 21 days and visits with competing diagnoses that may have required antibiotics. We identified any oral antibiotic prescription within 3 days of the visit and defined broad-spectrum antibiotics as macrolides and flouroquinolones.
We compared antibiotic and broad-spectrum antibiotic prescribing rates for Teladoc and physician offices. In multivariate models, we adjusted for sex, age, chronic illness (using the Charlson Comorbidity Index), site of care, and ARI diagnoses. Using the predictive margin method, we report predicted prescribing rates, adjusting for covariates.5
Teladoc users were less likely to be 51 years of age or older or have 1 or more chronic illnesses (Table 1). In both unadjusted and adjusted analyses, the fraction of ARI visits at which an antibiotic was prescribed was similar for Teladoc and physician offices (Table 2). The adjusted antibiotic prescribing rate for all ARI visits was 58% for Teladoc vs 55% at physician offices (P = .07). This pattern varied by specific diagnosis, with Teladoc more likely to prescribe antibiotics for pharyngitis and bronchitis and less likely for upper respiratory infection and nasopharyngitis.
The most common antibiotics prescribed were similar across sites: azithromycin (Teladoc, 58% vs physician offices, 45%); amoxicillin (Teladoc, 27% vs physician offices, 29%); and levofloxacin (Teladoc, 3% vs physician offices, 5%). For cases in which an antibiotic was prescribed, the adjusted broad-spectrum antibiotic prescribing rate for all ARI visits was 86% for Teladoc vs 56% at physician offices (P < .01).
Antibiotic prescribing rates for ARIs overall were similar for Teladoc and physician offices. However, both settings had high rates of inappropriate prescribing for conditions such as bronchitis; consistent with prior research that half of outpatient antibiotic prescriptions are not clinically indicated.6
When antibiotics were prescribed, Teladoc used more broad-spectrum antibiotics. This is concerning because overuse increases costs and contributes to antibiotic resistance. Greater use of broad-spectrum antibiotics may be driven by the tendency for physicians serving DTC companies to practice conservatively, with limited diagnostic information. DTC companies can work to lower rates through targeted quality-improvement initiatives to change physician behavior (eg, timely feedback), as well as direct education to patients to influence demand.
Our study had several limitations. Differences in patients and the complexity of conditions seen across settings may not be accounted fully in multivariate models. Furthermore, there is subjectivity in assigning individual ARI diagnoses, and in some cases, the diagnosis may be selected to align with the chosen treatment.
Use of DTC telemedicine is increasing rapidly despite lack of research about whether it is offering care of comparable quality. Our research suggests overuse of broad-spectrum antibiotics for ARI visits. Studies such as this can inform the coverage decisions of insurers and employers and ongoing policy debates in the United States about the appropriateness of different models of telemedicine.
Corresponding Author: Lori Uscher-Pines, PhD, MSc, RAND Corporation, 1200 S Hayes St, Arlington, VA 22202 (email@example.com).
Published Online: May 26, 2015. doi:10.1001/jamainternmed.2015.2024.
Conflict of Interest Disclosures: None reported.
Funding/Support: Support for this work comes from a grant from the California Health Care Foundation. Dr Mehrotra also reported receipt of support from a National Institutes of Health (NIH) grant (R21 AI097759-01).
Role of the Funder/Sponsor: The California Health Care Foundation and NIH had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.