Introduction
Leading up to the COVID-19 pandemic, direct-to-consumer (DTC) telemedicine for unscheduled acute care was one of the fastest growing areas within telehealth (1,2). Demand grew exponentially during the COVID-19 pandemic as the public sought alternatives to face-to-face care, with national telemedicine companies like Teladoc fielding twice as many calls in March 2020 as the same time last year (3). As policymakers weigh options for permanent changes to telemedicine regulations at the end of the public health emergency, the impact of telemedicine expansion on healthcare quality, utilization, and spending is a central question (4).
There are several ways in which telemedicine can be provided, and it is important to distinguish between these types when considering changes to telemedicine regulation and policy. First, DTC telemedicine, as defined in this paper, refers to a live video-based encounter that is initiated by the patient on-demand. DTC telemedicine visits are most commonly performed by clinicians who work for national for-profit companies and with whom a patient has no existing relationship, though many health systems and some practices now also offer on-demand visits for patients under their care. This is in contrast to scheduled telemedicine visits provided by the patient’s own primary care provider. These clinicians have an established relationship with the patient, access to their entire medical file, and can have the patient follow up in-person for additional tests or a physical exam if necessary.
DTC telemedicine is also different from other types of telemedicine such as teleconsults (initiated or facilitated by one provider to another provider such as in telestroke) and asynchronous telemedicine (not occurring in real-time), which we will not address in this paper. In many cases, DTC telemedicine services may be distinguished from other routine telemedicine visits in insurance claims data via billing codes, however this distinction varies by payer.
Proponents of DTC telemedicine believe it is a cost-efficient, convenient alternative to emergency room and urgent care visits, and a way to improve access to care. However, because DTC telemedicine is traditionally provided by independent for-profit companies separate from existing health systems, there is concern that promotion of DTC telemedicine may further fragment care and contribute to increased health care spending. Additionally, evidence on the quality of care rendered by DTC telemedicine providers has been mixed, with some studies finding similar rates of antibiotic prescribing for acute respiratory conditions, while others found higher rates, as well as decreased use of diagnostic testing for conditions such as strep throat (5).
It also remains unclear whether DTC telemedicine saves health care costs when considering entire episodes of care (6,7). On one hand, there may be cost savings if the telemedicine encounter resolves the issue; however, a potentially inadequate telemedicine evaluation may lead to more downstream visits, such as secondary office or urgent care visits, adding additional costs. To better understand this issue, we evaluated DTC telemedicine claims for acute respiratory infections from a large commercial payor and assessed whether these encounters led to more downstream care when compared to matched controls.
Methods
Overview
We chose to examine follow-up visits after DTC telemedicine and in-person encounters for acute respiratory infections for several reasons. First, acute respiratory infections are one of the most common reasons patients seek unscheduled care, either virtually or in-person through the primary care physician office, urgent care center, or emergency department (8,9). They are also self-limited in most cases and typically do not require additional testing or specific treatment after an initial evaluation.
Data source
We used Blue Cross Blue Shield of Michigan insurance claims data to identify patients who had a DTC telemedicine visit (Current Procedural Terminology [CPT] code 99444) or in-person visit (CPT codes 99201–99205, 99211–99215) for acute respiratory infections from 2016 through 2019. Of note, the CPT code used for this analysis was specific for DTC telemedicine services provided by a national telemedicine company beginning in 2016 (10). Prior studies have primarily used International Classification of Diseases, Ninth Revision (ICD-9) codes to identify acute respiratory infections (11,12). We similarly included the diagnoses of otitis media, bronchitis, bronchiolitis, pneumonia, sinusitis, pharyngitis, tonsillitis, laryngitis, tracheitis, nasopharyngitis, and mastoiditis (H65–70, J00–06, J09–22, J40, A221, A481, A3791, B250, and B440), using published crosswalks to translate from ICD-9 to ICD-10 diagnoses (13).
Identifying episodes of care
We identified all claims for each patient within a seven-day window (i.e., episode of care) following an index visit for acute respiratory infection. Index visits were identified as those with no other visit for acute respiratory infection within the preceding seven days. In cases where we identified multiple visits on Day 0, we prioritized DTC telemedicine, urgent care, then office visits for the index visit definition. This is because the progression of care that patients seek out is likely to be virtual care followed by in-person urgent care evaluation, then a primary care visit, and unlikely the reverse. Additional records by the same provider on the same day were counted as duplicates and removed.
For each patient with an episode of care initiated by a DTC telemedicine visit for acute respiratory infection, we identified two matched controls with an in-person (office or urgent care) index visit. Controls were matched on age, gender, rural or urban location, and comorbidities (congestive heart failure, chronic obstructive pulmonary disease, diabetes, chronic kidney disease, renal failure, and/or cancer). Patients with less than 6 months of insurance coverage prior to the index visit were excluded to ensure adequate identification of comorbidities.
Statistical analysis
We first described the trend in DTC telemedicine use overall and for acute respiratory infections over the study period. We then described patient characteristics of the DTC telemedicine group and of unmatched and matched in-person visits for acute respiratory infection. Our primary outcome was the percentage of episodes that included a subsequent office, urgent care, emergency department, or DTC telemedicine visit for a related principal diagnosis. We used a chi-squared test to compare this outcome between intervention (telemedicine) and control (in-person) groups.
Sensitivity analyses
A more inclusive or exclusive definition of a “related visit” may impact measures of downstream utilization. For example, strictly requiring acute respiratory infection as the primary diagnosis in a subsequent visit may undercount the true number of related visits. Therefore, we performed additional sensitivity analyses by varying the definition of a “related visit” within the 7-day episode to determine how follow-up rates changed. Our primary analysis identified subsequent visits with an associated principal diagnosis of acute respiratory infection. A second more broad definition included all visits involving claim lines with procedure codes potentially related to acute respiratory infection regardless of the diagnosis associated with that specific claim. These included CPT codes for chest/sinus imaging, streptococcus, influenza, or other viral testing, and routine blood work (e.g. complete blood count, basic metabolic panel). Finally, we examined overlapping and inclusive combinations of these two definitions (i.e, definition #1 AND #2 (most specific) vs definition #1 OR #2 (most sensitive)).
Limitations
Our study has several limitations. First, we used administrative claims data from a commercial insurer in one state. Utilization patterns by uninsured patients or those on Medicaid or Medicare, or in other states, may differ. Claims data also may not fully capture phone calls and other forms of unbilled healthcare utilization, though this likely affects both the DTC telemedicine and in-person groups equally. Additionally, while we have matched episodes for in-person and DTC telemedicine care on a number of patient and clinical characteristics, there may be important confounders that were not accounted for.
Due to the limitations of claims data, it is not possible to determine whether related visits following DTC telemedicine resulted from inadequate evaluation or appropriate triaging of patients, or whether they were medically indicated. A DTC telemedicine patient with a downstream office visit may have avoided the emergency department. Alternatively, unfamiliarity with or distrust of telemedicine may lead patients to seek out additional in-person care even if the initial telemedicine evaluation was sufficient. For example, in a recent national poll of older adults, two-thirds of respondents reported concerns about the quality of care in a telemedicine visit compared to in-person care (14).
Finally, our findings may not be generalizable to other telemedicine encounters such as those performed by the patient’s own primary care physician. A pre-existing patient-provider relationship likely changes the likelihood for a downstream visit. Future research addressing these issues will be important to inform policymakers, providers, and payers about the value of DTC telemedicine.
Results
Among approximately 10 million beneficiaries, 80,585 total DTC telemedicine visits were conducted, with volume increasing annually (Exhibit 1). We examined 85,963 index visits for acute respiratory infection (28,716 via DTC telemedicine and 57,427 matched in-person controls). Exhibit 2 lists the demographic and clinical characteristics of patients with in-person versus DTC telemedicine visits for acute respiratory infections before and after matching. Compared to unmatched visits, individuals using DTC telemedicine for acute respiratory infection were more likely to be female, be a non-elderly adult, live in an urban area, and have fewer comorbidities.
EXHIBIT 1.
Direct-to-consumer telemedicine episodes from 2016–2019
Note: Quarterly variation represents seasonality
Source: Analysis of Blue Cross Blue Shield of Michigan insurance claims, 2016–20
EXHIBIT 2.
Demographic and clinical characteristics of patients seeking care for acute respiratory infections
| Non-telemedicine group before matching, % (n=3,402,889) |
Direct-to-consumer telemedicine group, % (n=28,716) |
||
|---|---|---|---|
| Gender | Female | 56.6 | 61.3**** |
| Age | 0 to 9 | 26.7 | 7.6**** |
| 10 to 19 | 16.4 | 8.3 | |
| 20 to 29 | 10.3 | 12.5 | |
| 30 to 39 | 10.1 | 27.2 | |
| 40 to 49 | 11.4 | 24.2 | |
| 50 to 59 | 12.9 | 16.1 | |
| 60 to 69 | 8.0 | 4.1 | |
| 70 to 79 | 2.5 | 0.1 | |
| 80+ | 1.8 | 0.0 | |
| Rurality | Urban | 94.2 | 95.6**** |
| Rural | 5.8 | 4.4 | |
| Comorbidities | Cancer | 2.3 | 1.4**** |
| Diabetes | 5.8 | 4.1**** | |
| Congestive heart failure | 1.5 | 0.4**** | |
| Chronic obstructive pulmonary disease | 2.7 | 0.7**** | |
| Chronic kidney disease/renal failure | 0.7 | 0.2**** |
Note: After matching, the control and DTC telemedicine groups were identical in the above characteristics.
p<0.001
Source: Analysis of Blue Cross Blue Shield of Michigan insurance claims, 2016–2019
Compared to matched controls, episodes initiated via DTC telemedicine were more likely to lead to downstream related visits (10.3% vs 5.9%, p <0.01). However, the distribution of downstream visits was different. DTC telemedicine encounters resulted in more office and urgent care visits, while in-person visits had slightly more downstream emergency department visits—though this number was very low in both groups (Exhibit 3).
EXHIBIT 3.
Rate and site of related visits within 7 days after an index visit for acute respiratory infection
| Index visit was through DTC telemedicine, % (n=28,716) |
Index visit was through office or urgent care, % (n=57,427) |
|
|---|---|---|
| Any site | 10.3 | 5.9**** |
| Office | 6.0 | 4.5**** |
| Emergency room | 0.5 | 0.6*** |
| Urgent care | 1.7 | 1.0**** |
| Telemedicine | 2.5 | 0.0**** |
Note: Rate refers to percentage of acute respiratory infection 7-day episodes that had a downstream related visit.
p<0.01
p<0.001
Source: Analysis of Blue Cross Blue Shield of Michigan insurance claims, 2016–2019
Sensitivity analyses
Using the expanded definition of a related visit identified a higher rate of follow-up visits (Exhibit 4), though the overall distribution of subsequent visits by site of care was similar. Definitions 3 and 4 yielded nearly identical results to the primary analysis and second definition, respectively, and are therefore not reported separately.
EXHIBIT 4.
Rate and site of related visits within 7 days after an index visit for acute respiratory infection using expanded CPT codes to identify related visits
| Index visit was through DTC telemedicine, % (n=28,716) |
Index visit was through office or urgent care, % (n=57,427) |
|
|---|---|---|
| Any site | 18.1 | 13.6**** |
| Office | 12.6 | 11.5**** |
| Emergency room | 1.1 | 1.5**** |
| Urgent care | 2.3 | 1.3**** |
| Telemedicine | 3.4 | 0.0**** |
Note: Rate refers to percentage of acute respiratory infection 7-day episodes that had a downstream related visit.
p<0.001
Source: Analysis of Blue Cross Blue Shield of Michigan insurance claims, 2016–2019
Discussion
Our results show that patients using DTC telemedicine for acute respiratory infections are slightly more likely to have a repeat related visit within seven days than similar patients who initially use in-person care. Our findings are comparable to prior work which examined follow-up visit rates for acute respiratory infection at 2 and 21 days (11). Another study by Ashwood et al, which reported on average follow-up costs rather than specific rates of follow-up, also found slightly higher evaluation and management costs in the two days following a DTC telemedicine visit for acute respiratory infection as compared to an office visit, suggesting there were more downstream visits (6). Both studies support the notion that DTC telemedicine visits can lead to higher rates of downstream care. This could be due to a number of reasons: patients may use DTC telemedicine out of convenience but may then seek further evaluation either because they were directed to or because their symptoms worsened. Concerns about the inability to conduct a physical exam or the quality of care provided by a telemedicine visit, as suggested in a recent report by the National Poll on Healthy Aging, may also be reasons some patients seek additional care (14). It is not possible to distinguish the precise cause for follow-up visits using claims data. Future research such as chart review for follow-up recommendations and antibiotic prescriptions or more qualitative patient and provider interviews on this topic will be important for identifying why these additional visits occur.
DTC telemedicine visits for acute respiratory infection also had a different pattern of follow-up compared to in-person visits. Patients who used DTC telemedicine were more likely to have an office visit, urgent care visit, or repeat telemedicine visit than the in-person group, but they were slightly less likely to go to the emergency department. This may be due to patient self-selection, where patients with milder symptoms choose to use DTC telemedicine. Conversely, since the difference in emergency department visits between the two groups was much smaller than differences in downstream office and urgent care visits, these visits may be directly related to telemedicine use for a variety of reasons, as discussed above.
As telemedicine use and growth is sustained, DTC telemedicine by a third-party provider may lead to more downstream visits with associated increases in cost when compared to in-person care. However, this may be offset by savings from fewer emergency department visits, as well as by opportunity cost savings such as not needing to miss work for an in-person visit. Virtual visits between established patients and their primary care providers, rare prior to the COVID-19 pandemic, may also have a different pattern of episode utilization, though data is needed to support this theory. DTC telemedicine delivered via a third-party vendor, as in our study, may also have different downstream utilization compared to DTC telemedicine provided by existing health systems, where DTC telemedicine providers may be able to access a patient’s prior medical records or better coordinate follow-up care (15). Both modes have seen substantial growth in recent months, though there are not yet studies comparing outcomes between the two.
Policy implications
Prior to COVID-19, DTC telemedicine services were covered primarily by employer-based insurance plans in addition to some Medicare Advantage plans (16). DTC telemedicine was not eligible for coverage under traditional fee-for-service Medicare (17,18). As a result of the public health emergency caused by the COVID-19 pandemic, many of these regulations have been temporarily relaxed and reimbursement policies expanded through at least the end of the year 2021 (17). This resulted in a growth of nearly 12,000% in the number of fee-for-service Medicare beneficiaries using telemedicine in the first 1.5 months of the pandemic (18,19). Several large private insurers and many state Medicaid agencies also expanded telemedicine coverage during this unprecedented time (20). For example, Blue Cross Blue Shield of Michigan waived member cost-sharing for all telemedicine services from March 16 through June 30, 2020 (21).
Given this rapid expansion and ongoing concerns for infection prevention, virtual care is here to stay. Policymakers at the state and federal level must start thinking now about how telemedicine policy can maximize access to care while minimizing redundant or excess care after the public health emergency ends. There is debate over which changes to federal telemedicine regulation to make permanent, such as expanding eligible sites of care to include the patient’s home and whether telemedicine services for new (versus established) patients are eligible for coverage (22). Given that acute respiratory infection is the most common reason for DTC telemedicine visits (6,9), our study provides additional information for stakeholders to consider when weighing the costs and benefits of regulatory changes that would expand or restrict DTC telemedicine coverage. Data on the role of DTC telemedicine as compared to primary care-based virtual care is still limited, but a delineation between these services may be worth considering.
Conclusion
Now more than ever, there is a need to thoughtfully structure health policy changes to maximize the benefits of telemedicine while minimizing unnecessary care. Direct-to-consumer telemedicine has the potential to improve care access in a lower-cost setting, but the implications for overall health care utilization and spending are less clear. Although most patients seeking care for acute respiratory infection do not require follow-up within seven days, patients who used DTC telemedicine were more likely than matched controls to have secondary visits. Looking forward, it will be important for researchers and payors to more closely examine the reasons for these additional visits and to compare outcomes following telemedicine visits with a patient’s primary care provider as opposed to a third-party DTC platform.
References
- 1.Teladoc’s visits up 80% last year to 2.6 million [Internet]. Modern Healthcare. 2019. [cited 2020 Apr 27]. Available from: https://www.modernhealthcare.com/technology/teladocs-visits-up-80-last-year-26-million [Google Scholar]
- 2.Curran J Telehealth Services in the US. IBISWorld; 2019. October. Report No.: Industry Report OD5775S. [Google Scholar]
- 3.Krouse S Teladoc’s Remote Doctor Visits Surge in Coronavirus Crisis. WSJ Online [Internet]. 2020. April 14 [cited 2020 Jun 9]; Available from: https://www.wsj.com/articles/teladocs-remote-doctor-visits-surge-in-coronavirus-crisis-11586894400 [Google Scholar]
- 4.Winter A, Tabor L. Expansion of telehealth in Medicare [Internet]. Medicare Payment Advisory Commission Public Meeting; 2020. November 9 [cited 2020 Nov 30]. Available from: http://medpac.gov/docs/default-source/meeting-materials/telehealth-medpac-nov-2020.pdf?sfvrsn=0 [Google Scholar]
- 5.Uscher-Pines L, Mulcahy A, Cowling D, Hunter G, Burns R, Mehrotra A. Access and Quality of Care in Direct-to-Consumer Telemedicine. Telemed J E Health. 2016. April;22(4):282–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ashwood JS, Scott Ashwood J, Mehrotra A, Cowling D, Uscher-Pines L. Direct-To-Consumer Telehealth May Increase Access To Care But Does Not Decrease Spending. Health Affairs 2017;36(3):485–91. [DOI] [PubMed] [Google Scholar]
- 7.Nord G, Rising KL, Band RA, Carr BG, Hollander JE. On-demand synchronous audio video telemedicine visits are cost effective. Am J Emerg Med. 2019. May 1;37(5):890–4. [DOI] [PubMed] [Google Scholar]
- 8.Hsia RY, Friedman AB, Niedzwiecki M. Urgent Care Needs Among Nonurgent Visits to the Emergency Department. JAMA Intern Med. 2016. June 1;176(6):852–4. [DOI] [PubMed] [Google Scholar]
- 9.Ray KN, Shi Z, Poon SJ, Uscher-Pines L, Mehrotra A. Use of Commercial Direct-to-Consumer Telemedicine by Children. Acad Pediatr. 2019. August;19(6):665–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Blue Cross Members Have Access to Online Doctor Visits 24/7 [Internet]. Blue Cross Blue Shield of Michigan. 2016. [cited 2020 Nov 16]. Available from: https://www.mibluesperspectives.com/2016/06/24/blue-cross-members-have-access-to-online-doctor-visits-247/ [Google Scholar]
- 11.Shi Z, Mehrotra A, Gidengil CA, Poon SJ, Uscher-Pines L, Ray KN. Quality Of Care For Acute Respiratory Infections During Direct-To-Consumer Telemedicine Visits For Adults. Health Aff. 2018. December;37(12):2014–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Donnelly JP, Baddley JW, Wang HE. Antibiotic utilization for acute respiratory tract infections in U.S. emergency departments. Antimicrob Agents Chemother. 2014;58(3):1451–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.CMS’ ICD-9-CM to and from ICD-10-CM and ICD-10-PCS Crosswalk or General Equivalence Mappings [Internet]. [cited 2020 Jun 17]. Available from: http://data.nber.org/data/icd9-icd-10-cm-and-pcs-crosswalk-general-equivalence-mapping.html
- 14.Buis L Telehealth Use Among Older Adults Before and During COVID-19 [Internet]. National Poll on Healthy Aging; 2020. August [cited 2020 Nov 3]. Available from: https://www.healthyagingpoll.org/report/telehealth-use-among-older-adults-and-during-covid-19 [Google Scholar]
- 15.Wicklund E NewYork-Presbyterian, Walgreen’s Announce Telehealth Collaboration. mHealth Intelligence [Internet]. 2017. December 11 [cited 2020 Nov 16]; Available from: https://mhealthintelligence.com/news/newyork-presbyterian-walgreens-announce-telehealth-collaboration [Google Scholar]
- 16.Mehrotra A, Uscher-Pines L, Lee MS. The Dawn of Direct-to-Consumer Telehealth. In: Rheuban K, Krupinski EA, editors. Understanding Telehealth. McGraw-Hill; 2018. [Google Scholar]
- 17.Medicare Telemedicine Health Care Provider Fact Sheet [Internet]. [cited 2020 Jun 22]. Available from: https://www.cms.gov/newsroom/fact-sheets/medicare-telemedicine-health-care-provider-fact-sheet
- 18.Verma S Early Impact Of CMS Expansion Of Medicare Telehealth During COVID-19 [Internet]. Health Affairs Blog 2020. [cited 2020 Nov 13]. Available from: https://www.healthaffairs.org/do/10.1377/hblog20200715.454789 [Google Scholar]
- 19.Pifer R Medicare members using telehealth grew 120 times in early weeks of COVID-19 as regulations eased [Internet]. Healthcare Dive. 2020. [cited 2020 Jun 22]. Available from: https://www.healthcaredive.com/news/medicare-seniors-telehealth-covid-coronavirus-cms-trump/578685/ [Google Scholar]
- 20.Telehealth Coverage Policies in the Time of COVID-19 to Date [Internet]. Center for Connected Health Policy. 2020. [cited 2020 Nov 13]. Available from: https://www.cchpca.org/sites/default/files/2020-04/CORONAVIRUS%20TELEHEALTH%20POLICY%20FACT%20SHEET%20APRIL%2030%202020%20FINAL.pdf [Google Scholar]
- 21.Telehealth: Frequently asked questions for providers [Internet]. Blue Cross Blue Shield of Michigan. 2020. [cited 2020 Nov 17]. Available from: https://www.bcbsm.com/content/dam/public/shared/documents/coronavirus/provider/alerts/covid-19-teleheatlh-provider-faq.pdf [Google Scholar]
- 22.Senate US Committee on Health, Education, Labor & Pensions. Telehealth: Lessons from the COVID-19 Pandemic [Internet]. Full Committee Hearing; 2020. June 17 [cited 2020 Nov 17]. Available from: https://www.help.senate.gov/hearings/telehealth-lessons-from-the-covid-19-pandemic [Google Scholar]