Pandemic Telemedicine Adoption Trends in a Predominantly Rural Integrated Health System

Abstract

Between March 2020 and February 2022, use of telemedicine services in the U.S. shifted dramatically in response to the evolving SARS-CoV2 pandemic. The initial wave caused many non-emergent clinical services to be postponed, including specialty care clinic visits, which were rapidly converted to telemedicine encounters. Telemedicine use ebbed and flowed with subsequent pandemic waves. This paper describes trends in telemedicine use from March 2020–February 2022 at Geisinger, a predominantly rural integrated health system. It highlights characteristics of 5,390 virtual vs. 15,740 in-person clinic visits to neurosurgery and gastroenterology specialists in December 2021 and January 2022. Differences in ordering of diagnostic testing and prescription medications, as well as post-clinic-visit utilization, varied by specialty. Virtual visits in these specialties saved patients from traveling over 174,700 miles/month to attend appointments. Analyzing telemedicine use patterns can inform future resource allocation and determine when virtual encounters can complement or replace in-person specialty care visits.

Introduction

Connecting patients with their healthcare providers via video and telephone to conduct virtual visits has been happening since well before the SARS-CoV2 pandemic [1-2]. Notable work to assess the impact of home monitoring and video encounters for patients with diabetes was conducted in the 1990s in the IdeaTel project [3-4]. Similarly, an extensive telehealth program exists at the US Department of Veterans Affairs that dates back 20 years and contributes to the healthcare of more than 900,000 veterans per year [5-6].

Telemedicine use was growing slowly in the years leading up to 2020 [7-9]. However, it was the international pandemic’s stay-at-home orders, public and private payors’ waiving place-of-service and rurality restrictions, and the limiting of nonurgent in-person appointments that spurred an unprecedented, almost overnight adoption of virtual clinic visits on a large scale. Rapid adoption was facilitated in the United States by the January 31, 2020 declaration from the U.S. Secretary of Health and Human Services that the novel coronavirus constituted a public health emergency [10], which enabled healthcare providers to bill Medicare for telehealth services regardless of where the patient or provider was physically located [11].

A tremendous amount of information has already been published related to telemedicine during the global pandemic— a PubMed search on “Telemedicine” AND “COVID” in February 2022 yielded over 7,600 results—but a number of important questions have not been extensively investigated, particularly with respect to outpatient specialty care. Among these:

• What types of clinical care scenarios can and cannot be effectively managed virtually?

• How many patients completing virtual visits are subsequently seen in the clinic within a short time window?

• How many virtual visits result in orders for laboratory testing and medication prescriptions?

• How much can virtual visits reduce the need for transportation to the clinic?

To help answer these questions, this paper describes trends in the use of telemedicine at Geisinger and specifically examines outpatient clinic visits in two specialties that had considerable use of telemedicine throughout the pandemic.

Methods

Geisinger is an integrated health system with nine hospitals and over 100 ambulatory care sites located primarily in central and northeast Pennsylvania. Geisinger implemented the Epic electronic health record (Epic Systems, Verona WI) beginning in 1996. Geisinger began conducting telemedicine video visits in 2008 in outpatient, inpatient, and emergency room settings. Prior to the outbreak of COVID, telemedicine capabilities (leveraging technology from Teladoc, Purchase NY) were well established at Geisinger; however, these tools were used in less than 1% of specialty care clinic visits in 2019. We considered a telemedicine visit one where the patient and the provider were in physically distinct locations (typically with the provider in a clinic and the patient at home), and the visit was conducted by video or telephone.

Prior to the pandemic, Geisinger had developed a series of dashboards to track telemedicine adoption. The dashboards displayed trends including total visit volume completed through in-person, telephone, and video encounters. Filters could be applied to facilitate exploration of other characteristics related to telemedicine use, such as subspecialty, provider type, visit type (telephony, video to clinic, or video to home), and primary payor.

Another dashboard presented information about the geographic distance between the home addresses and clinic addresses of telemedicine visits. This distance was calculated using straight-line (Euclidean) distance using the Haversine formula [12]. The dashboard presented a histogram of distance categories, and filters could be applied to explore by date range, service line, county, and more. The intent of this dashboard was to assess whether telemedicine encounters were making it easier and more convenient for patients to access clinical services. Particularly for rural, geographically distant patients, we wished to understand how much potential there was to reduce travel time—i.e., “windshield time”—required for each encounter.

In addition to reviewing the dashboards to assess general and specific trends in telemedicine adoption, we examined the characteristics of specialty clinic visits in neurosurgery and gastroenterology clinics between December 2021 and January 2022. These specialties were selected because they (along with psychiatry) were among the highest utilizers of telemedicine across the Geisinger system. While the use of telemedicine for clinical encounters with psychiatrists and other behavioral health professionals has been reported on extensively [13-14], fewer analyses have been conducted examining specialty care in domains such as neurosurgery and gastroenterology.

In December 2021 and January 2022, the volume of COVID-19 cases treated in Geisinger facilities was the highest it had been to that point in the pandemic. Many non-urgent surgical cases were being postponed, as hospital beds and staff were allocated to care for patients who were admitted for complications of COVID-19. Nevertheless, the overall specialty clinic visit volume (combined in-person and virtual) was comparable to pre-pandemic levels.

Comparing in-person with telemedicine visits, we assessed differences in ordering of laboratory testing and prescription medications. We also measured differences in post-clinic-visit utilization, specifically investigating whether patients completing a neurosurgery or gastroenterology telemedicine visit had a higher rate of subsequently completing an in-person visit in the same specialty or an emergency department visit in the next 1–3, 4–7, or 8–14 days. Chi-square tests were performed to compare numbers of initial in-person versus telemedicine encounters with subsequent orders or post-visit utilization.

Typical reasons for specialty care visits with neurosurgeons included chronic pain, epilepsy, stroke, tumors in the brain and spinal cord, and peripheral nerve issues. Neurosurgeons frequently order imaging studies such as x-rays, computerized tomography scans, and magnetic resonance imaging) scans, in addition to procedures such as lumbar punctures. Typical reasons for specialty care visits with gastroenterologists included diagnosis and treatment of digestive disorders such as liver disease, diverticulitis, ulcers, irritable bowel syndrome, and cancers. Gastroenterologists commonly order blood and stool tests, and often send patients for abdominal imaging studies and endoscopic procedures.

Results

Figure 1 shows the dashboard Geisinger developed to track telemedicine visit volume. Between March 2020–February 2022, more than 850,000 telemedicine visits occurred, with approximately 62% of visits by video and 38% by telephone. Adoption of telemedicine varied considerably by service line. Over 70% of psychiatry visits were conducted virtually, while many specialties—including cardiology, cancer, dermatology, and women’s health had telemedicine use rates of less than 10%. Aside from psychiatry, two of the specialties with the highest percentage of telemedicine visits were neurosurgery (36%) and gastroenterology (35%).

Figure 1.

Dashboard for tracking completed telemedicine appointments. Filters could be applied to facilitate exploration of characteristics related to telemedicine use, such as subspecialty, provider type, visit type (telephony, video to clinic, or video to home), and primary payor.

Table 1 presents characteristics of in-person and telemedicine visits in neurosurgery and gastroenterology. Between December 2021 through January 2022 there were 1,110 virtual vs. 2,360 in-person clinic visits to neurosurgery specialists and 4,280 virtual vs. 13,380 in-person clinic visits to gastroenterologists.

Table 1.

Characteristics of in-person and telemedicine visits in neurosurgery and gastroenterology in December 2021 and January 2022.

	Neurosurgery		Gastroenterology
	In-person	Telemedicine	In-person	Telemedicine
Total visits	2,360	1,110	13,380	4,280
Visits per person	1.18	1.08	1.39	1.16
≥1 order for lab tests	13.3%	2.3%	24.1%	22.0%
≥1 order for medications	29.5%	2.7%	34.0%	28.8%
≥1 order for radiology	34.9%	23.6%	11.3%	6.5%
≥1 subsequent in-person visit to same specialty clinic within 14 days of index visit	4">3%	2">6%	10">6%	4">7%
within 3 days of index visit	0.3%	0.6%	3.0%	1.7%
within 4-7 days of index visit	1.1%	0.9%	3.3%	1.5%
within 8-14 days of index visit	2.9%	1.1%	4.3%	1.5%
Emergency room visits within 14 days of index visit	2.9%	1.7%	1.7%	1.8%
within 3 days of index visit	0.7%	0.5%	0.6%	0.7%
within 4-7 days of index visit	0.9%	0.7%	0.5%	0.4%
within 8-14 days of index visit	1.3%	0.5%	0.7%	0.6%
Median one-way distance between clinic and patient’s home address (miles)	16	20	14	17

Compared with in-person visits, fewer telemedicine visits in neurosurgery resulted in orders for laboratory tests (13% vs. 2%), prescription medications (29% vs. 3%), and radiology studies (35% vs. 24%). In gastroenterology, the differences between in-person visits and telemedicine visits in terms of ordering were not as dramatic: for laboratory tests (24% vs. 22%), prescription medications (34% vs. 29%), and radiology studies (11% vs 7%). In each case, the difference between ordering patterns comparing in-person to telemedicine visits were statistically significant (p<0.001).

The post-clinic-visit utilization analysis found that index neurosurgery and gastroenterology visits completed via telemedicine had lower rates of a subsequent in-person visit in the same specialty within 14 days compared to index visits that were conducted in person (neurosurgery: 2.6% vs 4.3%, p=0.02; gastroenterology 4.7% vs 10.6%, p<0.001). Index neurosurgery telemedicine visits were also followed by lower rates of emergency room visits within 14 days (1.7% vs. 2.9%, p=0.04) However, there was no difference in emergency room utilization following index gastroenterology telemedicine visits (1.8% vs. 1.7%, p=0.8).

Given Geisinger’s rural service area, which covered 45 counties in Pennsylvania, there was considerable variability in the distance patients would have to travel for an in-person specialty care visit if a telemedicine option was not available (Figure 2). A considerable number of the 850,000 patients who completed telemedicine visits between March 2020–February 2022 (about 38%) lived more than 20 miles from the location of the clinic. For neurosurgery and gastroenterology visits, the average travel distance saved by conducting each visit virtually was 32.4 miles each way (median distance of 20 miles for neurosurgery and 17 miles for gastroenterology). This equated to approximately 174,736 round trip miles of travel prevented each month by using telemedicine in these two specialties.

Figure 2.

Dashboard presenting information on the distance between clinics and patients’ home addresses.

Discussion

Telemedicine has been promoted over the past two decades as a tool that could revolutionize the healthcare industry [7]. Prior to the pandemic, however, the adoption of telemedicine in the U.S. was limited, hampered by factors such as limited reimbursement, licensure hurdles, and variation in state practice laws [8]. Between March 2020 and February 2022, use of telemedicine services shifted dramatically in response to the evolving SARS-CoV2 pandemic. The initial wave caused many non-emergent clinical services to be replaced by telemedicine encounters, and telemedicine use fluctuated with subsequent pandemic waves.

Numerous publications have appeared in recent months describing the use of telemedicine in various settings [15-24]. As others have reported, we observed considerable variability at Geisinger in the use of telemedicine for clinic visits across clinical specialties. It is unknown whether this variability was attributable to patient preferences, provider preferences, clinical reasons, technical constraints, or other factors. Much has been written about social determinants of health—such as insurance type, socioeconomic status, access to broadband internet, etc.—influencing use of telemedicine. While these factors are extremely important, they do not explain the variability in telemedicine use across specialties in our clinical enterprise, assuming that social determinants factors were reasonably evenly distributed across specialties.

In telemedicine visits, the use of telephone vs. video visits also varied. While we did not investigate factors associated with the use of telephone vs. video visits, Huang and colleagues found that patients of Black or Hispanic race/ethnicity, or those living in low socioeconomic status or low internet access neighborhoods were less likely to schedule video visits in primary care settings [25]. Eberly and colleagues similarly reported that older patients, female patients, Black, Hispanic, and poorer patients had less video use [26].

Did patients completing telemedicine visits contribute to increased or decreased healthcare utilization?

A key attraction for telemedicine specialty care visits for payors is the possible savings involved in replacing physician office visits with potentially less expensive virtual visits—with savings coming from the reduction in front-desk staffing and facility costs. Others argue that the expense of the telemedicine platform, the need to check patients for telemedicine readiness, and other expenses could offset this savings, and that the costs of providing care virtually vs. in-person may be comparable. Ashwood and colleagues found that the availability of telemedicine visits “may tap into unmet demand for health care, and new utilization may increase overall health care spending” [27].

For two months of specialty care visits in neurosurgery and gastroenterology, we examined whether telemedicine visits increased the likelihood that patients would be subsequently seen in the clinic within a short time window. We found that the rates of subsequently completing an in-person clinic visit within 7 and 14 days or presenting to the emergency department within 3, 7, and 14 days were slightly lower for neurosurgery patients completing an initial telemedicine vs. an in-person visit. Similarly, we found that the rates of subsequently completing an in-person clinic visit within 3, 7, and 14 days were slightly lower for gastroenterology patients completing a telemedicine vs. an in-person visit. The reasons for these differences are unclear; however, even though the large sample size yielded statistically significant differences, the clinical difference was relatively small.

When examining how many virtual visits had orders for diagnostic testing and medication prescriptions compared to in-person visits, we saw considerable differences in neurosurgery vs. gastroenterology. Based on these findings, we expect there may be variability in other specialty clinic visits as well. In two months of neurosurgery visits, very few telemedicine visits resulted in orders for laboratory testing or prescriptions; however, relatively more visits had orders for imaging. This finding may have been the result of different reasons for telemedicine visits compared to in-person visits. In two months of gastroenterology visits, the ordering of laboratory testing, diagnostic imaging, and prescriptions was relatively high among telemedicine visits. These findings may suggest that gastroenterology visits, whether they were conducted in-person or virtually were more comparable in terms of the services provided than neurosurgery visits completed in-person vs. neurosurgery visits completed virtually.

Benefits of telemedicine visits for making care convenient

One of the most obvious benefits of telemedicine is improved convenience for patients and providers. Perhaps the earliest mention in the biomedical literature that virtual visits might be useful in clinical practice was reported in the Lancet on November 29, 1879. The writer cited the case of an anxious mother who, convinced that her baby had the croup, was connected in the middle of the night to the family doctor via “telephonic communication.” (The telephone was invented by Alexander Graham Bell in 1876.) “Lift the child to the telephone,” he instructed, “and let me hear it cough.” After listening, the doctor declared it was not a case of the croup, and “all anxiety quieted, the trio settle[d] down happily for the night” [28-29].

The Ancient Greek τῆλε (têle) means “at a distance, far off, far from.” Geisinger patients lived primarily in central and northeast Pennsylvania in 45 counties that were predominantly rural. The average distance between a patient and the nearest specialty clinic was over 32 miles in Euclidean distance. (Measuring distance on a Euclidean plane—i.e., “as the crow flies”—somewhat understates the actual driving distance based on roads; this is a limitation of our analysis.) It was not unusual for patients to spend one or two hours traveling to complete a 20-minute in-person visit. When patients had challenges related to transportation or difficulty getting time away from work to complete appointments, the healthcare access burden was even higher.

Telemedicine can overcome barriers to care associated with time and distance. It is also worth mentioning that other factors affected by telemedicine, such as space utilization and decreased infection risk can help optimize the care for both patients and providers. Telemedicine is a tool that, when used appropriately, can help deliver the right care in the most effective way. This study did not examine whether telemedicine specialty care clinic appointments were more or less likely to be missed than in-person appointments. Adepoju and colleagues previously reported that telemedicine was strongly associated with fewer missed appointments in Federally Qualified Health Center clinics in Texas during the pandemic [30].

What happens to telemedicine reimbursement when the COVID Public Health Emergency ends?

Concerns about telemedicine reimbursement are nearly as old as the concept of telemedicine itself. While acknowledging the convenience of the telephone, the editor of the Lancet cautioned in 1883: “The only fear we have is that when people can open up a conversation with us for a penny, they will be apt to abuse the privilege, and that to have a dozen telephone consultations in one day, or conversations that might be thought to supersede a consultation, would be a doubtful addition to one’s advantage or repose” [28, 31]. Prior to the pandemic, major challenges with telemedicine adoption were variability in reimbursement, including federal requirements in the U.S. that telemedicine encounters originate from clinical facilities vs. the home, in most cases.

Declaring COVID-19 a public health emergency in the United States enabled the Department of Health and Human Services (HHS) and Centers for Medicare & Medicaid Services to reimburse clinic visits using telemedicine for Medicare patients irrespective of location and at the same rate as in-person visits. While the declaration of a public health emergency (which has been renewed quarterly by the HHS Secretary) is expected to continue well into 2022, without legislative action by Congress, experts have indicated that “it’s going to pretty much return to the way it was before COVID” [32]. Pending legislation, Including the Telehealth Modernization Act of 2021, the CONNECT for Health Act of 2021, and the Telemental Health Care Access Act would formalize and cement telemedicine access and reimbursement policies beyond the pandemic. Studies regarding the efficacy and effectiveness of telemedicine can help develop an evidence base for best practices and inform sound policymaking.

Conclusion

Several factors have converged over the past several years to provide a climate that supports the use of telemedicine on a larger scale than we have seen in the past. These factors include advances in technology, a more open regulatory environment, and the rise of consumerism. Out of necessity caused by the worldwide pandemic, telemedicine in 2020 was adopted broadly. At our health system, patterns of telemedicine use varied not only in response to the pandemic, but by clinical specialty. Analyzing telemedicine use patterns can inform future resource allocation and determine when virtual encounters can complement or replace in-person specialty care visits. At least in the two specialties we examined, where more than one-third of visits were conducted via telemedicine, we did not find evidence that telemedicine required an additional in-person visit, suggesting that it can be truly substitutive rather than additive.