Association of Telehealth with Short-Interval Follow-Up

Discussion

Newly expanded telehealth visits were associated with significantly higher rates of total short-interval follow-up visits and sooner first occurrence of short-interval follow-up visits compared with in-person visits in the year after the onset of the COVID-19 pandemic. These findings are similar to prepandemic insurance claims data showing higher subsequent visit rates after video visits compared with in-person visits.¹¹ However, a study of a large primary care practice with well-established telehealth services showed rates of short-interval follow-up between telehealth and in-person index visits were similar,¹³ suggesting telehealth can be implemented in a manner that does not increase downstream use in certain contexts.

Although the rate of telehealth-associated short-interval follow-up in the pandemic period was similar to the baseline prepandemic rate of short-interval follow-up, it was significantly higher than the rate of pandemic period in-person-associated short-interval follow-up. This suggests an important difference in rate of short-interval follow-up visits between in-person and telehealth index visits in the pandemic period. There are multiple potential reasons for this meaningful difference.

We found notably higher rates of short-interval follow-ups occurring within 2 weeks of a telehealth visit. This finding may be due, in part, to the sudden, rapid expansion of telehealth in our practice. We accelerated our implementation and provider training and did not have adequate time for patient education on telehealth. Therefore, providers and patients may have felt unprepared to navigate an encounter without the ability to fully gauge body language or perform a physical examination. This may have led both providers or patients to request an in-person follow-up visit in a few days to assure a thorough assessment had been completed and an accurate diagnosis reached. If initial unfamiliarity and discomfort with telehealth contributed to our findings, we expect rates of telehealth-associated short-interval follow-ups will decrease over time as patients and providers become accustomed to this visit modality.

Higher rates of telemedicine short-interval follow-ups may have been driven by the characteristics of our scheduling practices. Since the number of appointments on provider schedule template was fixed, increasing telehealth appointments decreased availability of in-person appointments. It is possible that the earliest available appointment was telehealth rather than in person. The patient or provider may have chosen the earliest available appointment regardless of perception of the appropriateness of visit type for the complaint, which may have increased the need for a short-interval follow-up.

Similarly, in the year following the onset of the pandemic, patients with COVID symptoms could only be seen via telehealth. While some evidence suggests patients initially seen for COVID symptoms do not require higher rates of follow-up if seen via telehealth,¹² and many patients with acute illnesses will self-select out of telehealth when given the choice,¹⁴ prepandemic data of telehealth visits conducted for upper respiratory symptoms demonstrated significantly higher follow-up rates than in-person visits.¹⁰ As safety concerns related to the COVID-19 pandemic subside, identifying patients presenting with acute physical symptoms and directing them to in-person visits may reduce telehealth-associated short-interval follow-ups.

Some patients with nonacute reasons for visit may also be less suitable for telehealth. Patients who rate their health as poor are less likely to be satisfied with telehealth than those who rate their health as good.⁵ Refined telehealth scheduling algorithms should consider reason for visit and patient health status to improve patient and provider satisfaction with telehealth and potentially reduce rates of short-interval follow-ups. However, development of such algorithms in primary care is challenging. Most primary care visits address multiple issues,¹⁵ and many of these may not be shared with the scheduler when the appointment is made. In addition, as patient self-scheduling becomes more common in primary care, accurate and complete understanding of reason for visit may become even less predictable and appropriate assignment of visit type less reliable.

Our findings have significant implications for both primary care practices and patients. While some acute illnesses or disease exacerbations warrant close follow-up, unnecessary use of 2 visits because the initial visit type was inappropriate reduces overall clinic access. To accommodate the increased need for post-telehealth follow-up visits, especially those between 0 and 15 days, practices may have to reserve unusually high numbers of open appointment slots or risk further delaying appropriate evaluation, reducing patient satisfaction and potentially leading to unnecessary urgent care or emergency department visits. The negative impact on clinic access is particularly troublesome in primary care, where practices are often excessively busy and access is already limited.

Unnecessary short-interval follow-up may reduce patient satisfaction with their care and cause undue financial burden on patients who must pay for 2 visits when 1 visit of the optimal type could have sufficed. This would disproportionately affect self-pay patients and those with high deductible insurance plans and exacerbate waste in our health care system. If telehealth visits are perceived as contributors to low-value care due to need for higher rates of follow-up, insurance carriers and lawmakers may reduce reimbursement and legislative support for telehealth.

Future research is needed to identify strategies for optimization of telehealth scheduling processes to minimize need for short-interval follow-up, thereby improving practice efficiency, improving patient-centered outcomes including cost and patient satisfaction, and promoting sustainability of telehealth in primary care.