Abstract
Introduction:
We hosted a national consensus conference with a diverse group of stakeholders to develop a patient-centered research agenda focused on reducing disparities in telehealth use.
Methods:
Attendees were purposively invited to participate in a 2-day virtual conference. The group developed a prioritized research agenda focused on reducing disparities in telehealth uptake, with discussion informed by findings from a scoping review. All work was conducted in partnership with a Steering Committee of national experts in telehealth and patient-centered care (n = 5) and a community-based Telehealth Advisory Board with experience with telehealth use and barriers (n = 8).
Results:
Sixty individuals participated in the conference and discussion resulted in a final list of 20 questions. Fifty-two attendees voted on the final prioritization of these questions. Results were aggregated for all voters (n = 52) and patient-only voters (n = 8). The top question identified by both groups focused on patient and family perspectives on important barriers to telehealth use. The entire group voting identified telehealth's impact on patient outcomes as the next most important questions, while the patient-only group identified trust-related considerations and cultural factors impacting telehealth use as next priorities.
Conclusions:
This project involved extensive patient and stakeholder engagement. While voting varied between patients only and the entire group of conference attendees, top identified priorities included patient and family perspectives on important barriers to telehealth, trust and cultural barriers and facilitators to telehealth, and assessment of telehealth's impact on patient outcomes. This research agenda can inform design of future research focused on addressing disparities in telehealth use.
Introduction
The COVID-19 pandemic exacerbated health and social inequities. Patients were more likely to avoid seeking necessary health care, and providers were forced to triage treatment by urgency due to resource constraints. 1 Telehealth use surged during the pandemic to facilitate patient access to needed health care while minimizing risk of COVID-19 exposure.2,3 Telehealth connects individuals to their health care providers when in-person care is not available, necessary, or possible. 4 Modes of telehealth may include talking to a provider over the phone or video chat, sending and receiving messages using a secure portal, or using a remote monitoring device to share health data with a provider. Services are typically categorized as synchronous (allowing for direct, real-time engagement between patient and provider) or asynchronous (nondirect communication, like storing and forwarding information).
While this growth in telehealth use facilitated access to needed care for many patients, it also threatened to worsen an already significant disparity in outcomes for vulnerable populations. Factors that impact patient access to and use of telehealth are complex. The American Telemedicine Association describes 10 related components or elements that are needed to eliminate health disparities using telehealth, ranging from structural anti-racism to connectivity. 5
Digital readiness is a term that encompasses the digital skills, trust, and ability to use tools needed to carry out online tasks. 6 Approximately 29% of Americans have low levels of digital readiness, which transcends lack of access to digital devices; for example, lack of readiness may be driven by low trust in digital health platforms or lack of skills to navigate digital health platforms. 7 Digital readiness barriers can lead to disparities in engagement with telehealth-based clinical care. 8 Left unaddressed, disparities in telehealth use and uptake will exacerbate gaps in access to evidence-based tools and resources and lead to increased health disparities. 5
To that end, our team undertook a two-step process to develop a patient-centered research agenda to reduce disparities in telehealth uptake. This process was conducted in close partnership with a Philadelphia-based patient advisory board and a national Steering Committee comprised of experts in patient communication, patient and family centered care, telehealth, and health care disparities. In the first step, we performed a scoping review to understand the current state of literature regarding both barriers and disparities in telehealth uptake. In the second step, we held a national consensus conference in which we convened a broad range of stakeholders to review the results of the scoping review and to use findings to develop a prioritized research agenda. In the following article, we report on these activities, including a list of the final prioritized research agenda.
Methods
Conference planning was initiated with the establishment of two stakeholder advisory groups, one national and one local. The national Steering Committee was convened in August 2021 to support both the scoping review and consensus conference. Committee members had expertise in patient-centered care, patient and family advocacy, health care equity and disparities, and telehealth practice, policy, and research. Five members participated in meetings over a 17-month period to inform the development of scoping review research questions and data extraction tools and to support the planning, implementation, and reporting of the consensus conference. As this was not human subjects research, IRB approval was not required.
In addition, patients and community members in the Philadelphia region were invited in October 2021 to join a Telehealth Advisory Board (TAB) to inform patient-centered consensus conference planning and implementation. Members were primarily recruited through local community organizations and existing patient advisory councils, with the goal of convening a diverse set of stakeholders who represent various communities impacted by telehealth disparities in the greater Philadelphia region. Eight members participated in monthly meetings to inform conference invitations, agenda, materials, and patient engagement, and all TAB members were invited to the consensus conference. Two TAB members cochaired the group and were primarily responsible for meeting facilitation.
A major first step of conference planning included a scoping review to answer the question: What barriers and disparities in telehealth uptake and use have been documented in the literature? The primary purpose of the scoping review was to inform conference discussion and consensus building. The review was undertaken in accordance with the Joanna Briggs Institute (JBI) methodological guide for scoping reviews.
9
The scoping review process was developed collaboratively between the project team, a medical librarian, and the steering committee. For the purpose of the review, the team used the following operational definitions:
- Telehealth: Bidirectional communication between a patient and a provider. Includes both synchronous (communication occurring between provider and patient at the same time, such as a video visit) and asynchronous (communication occurring at different times, such as text messages or e-mails). - Barrier: A factor (experiences or perceptions) that negatively impacts patients' telehealth use. - Disparity: A difference in telehealth use between two groups that is found to be statistically significant (p < 0.05).
While the scoping review included literature published globally between 2011 and 2021, data presented at the consensus conference narrowed in on sources from the United States published between 2017 and 2021 to help focus discussions. The scoping review protocol was submitted with Open Science Framework registries (https://osf.io/df6aw/). Comprehensive scoping review results will be published in a forthcoming manuscript.
The national consensus conference was held from September 28–29, 2022, via Zoom. The purpose was to develop a patient-centered prioritized research agenda focused on reducing disparities in telehealth uptake and use. The consensus conference convened a broad group of stakeholders (n = 60) from across the country with experience and expertise in telehealth infrastructure, policy, and patient care. Attendees were selected for their experience and engagement with telehealth. Both the TAB and steering committee provided input on conference invitations. In alignment with recommendations from the JBI Scoping Review Methodology Group, attendees also represented knowledge users relevant for the interpretation of scoping review results. 10 A full list of organizations represented by attendees can be found in Table 1.
Conference Attendees
The conference began with an introduction by project lead Dr. Kristin Rising, in which an overview of the scoping review process and conference purpose were shared. Next, Dr. Amy Leader presented the scoping review findings relevant to “barriers.” Attendees were then divided across six breakout groups to discuss gaps in the research and draft potential research questions related to “barriers.” After small group discussion, attendees came back together into the full group and a member of each breakout group presented key discussion points and the list of generated questions to the full group.
During a break, the project team compiled draft research questions across all breakout groups and entered them into Qualtrics online survey platform. 11 Each attendee then voted on the level of importance of the entire list of drafted questions using the online voting platform. The survey for each question read, “As an area of research, the following question is: 1) not important, 2) somewhat important, 3) neutral, 4) important, or 5) very important?” Attendees were also asked to self-identify as an academic, clinician, community representative, industry representative, or patient to allow for sorting of voting results by different groups. All 60 external attendees as well as the three project leads (K.L.R., A.E.L., B.W.) were invited to vote.
Upon completion of discussion and voting on “barriers,” Dr. Brooke Worster presented the scoping review findings and draft research questions regarding “disparities” to the entire conference group. The same processes for breakout groups and voting were then repeated for discussion related to disparities in telehealth uptake. At the end of the 1st day, there were 58 newly drafted and voted on research questions, 35 of which came from the “barriers” discussion and 23 from the “disparities” discussion. A total of 56 attendees submitted a voting survey on day 1, including the three project leads. A full list of questions, with the original language maintained, can be found in Table 2.
Attendee Generated Draft Research Questions
SES = socioeconomic status; TH = telehealth.
In preparation for day 2, the research team compiled voting results across all 58 questions to identify the top-rated questions for discussion. Questions were sorted by mean score and the top 20 questions from the full group voting results as well as patient-only results were extracted. Any questions that were included in either (1) the top 10 rating from either group or (2) the top 20 rating in both groups were kept in the list for discussion on day 2. This resulted in 17 questions for group discussion in day 2 (Table 3).
Top-Rated Research Questions
Discussion on the 2nd day started with a review of the top 17 questions identified based on day 1 voting. Upon presentation of the questions, conference attendees identified challenges in comparing the individual questions due to variable wording and level of focus across the questions. The group collectively decided that it would be more informative to identify priority research themes that represented the questions, and then develop a final set of focused questions later. The remainder of the discussion on day 2 was spent identifying 15 research themes that were organized into 4 categories, with some overlap in content: barriers, trust, outcomes, and interventions/infrastructure (Table 4).
Priority Research Themes and Categories
After conclusion of the conference, the research team developed a set of research questions to represent the key categories and associated research themes identified during the conference. These questions were iteratively refined by the TAB and the Steering Committee members to ensure that the content adequately reflected the conference findings, and that the question wording was clear and balanced across questions. A final list of 20 questions was sent out to all attendees electronically for a final round of voting, which focused on attendees developing a rank order of questions from most to least important. Voting results were discussed with both the TAB and the Steering Committee, with a focus on exploring potential causes or meaning of the most significant voting discrepancies.
Results
Scoping review
The database searches yielded 11,156 citations after deduplication. Title and abstract screening resulted in the exclusion of 9,913 citations that did not meet identified inclusion criteria. A total of 1,243 full-text citations were assessed for eligibility. Upon full-text review, an additional 623 citations were excluded for various reasons. There were 618 references ultimately included for data extraction.
Sixty-four percent (n = 394) of all of the sources were based in the United States, while the remaining studies took place in numerous other countries. A total of 369 sources (60%) were published between 2020 and 2021, representing a significant surge in telehealth-related publications during the years of COVID-19. A majority of sources looked at synchronous modalities of telehealth (69%, n = 430) and collected data from the perspective of patients and their families (85%, n = 526). More studies utilized quantitative methods (57%, n = 354) than qualitative methods (28%, n = 171) or mixed methods (15%, n = 91). A total of 428 sources to 395 sources (64%) included data on barriers to telehealth uptake and use and 293 sources (47%) pertained to telehealth disparities.
Conference attendees
During conference registration, attendees were asked to indicate which professional group they most identified with. Sixteen individuals identified as clinicians (27%), 12 as nonclinician academics (20%), 8 as patients (13%), 8 as industry representatives (13%), 7 as community organization representatives (12%), and 9 as other (15%).
Conference breakout group discussions
Across the six barrier-focused breakout rooms, groups commonly discussed the relative importance of access and outcomes. They worried that removing barriers might not, as assumed, lead to increased telehealth uptake. Many participants also noted that barriers were framed through patient differences (e.g., race, age) rather than the underlying structural causes. They thought about whether barriers transcended individual factors and might better be considered at the system level. All groups were in favor of using digital navigators to reduce barriers. Several groups considered the implicit assumption that video telehealth is preferred to audio-only as well as remarking upon provider education and training as impediments to patient uptake of telehealth.
Across the disparity-focused breakout groups, the underlying premise that disparities arose due to deeper societal flaws rather than demographic characteristics was ubiquitous. A common thread between groups was how telehealth must be tailored to address the needs of different populations—both to drive access and to deliver care equitably. Most groups identified that measures of success in reducing disparities should be patient-centered and patient-identified, rather than assumed by researchers.
Voting results
A total of 52 conference attendees voted on the final list of research questions (86.7% response rate). As the focus of this conference was to establish a patient-centered prioritized research agenda, voting results were assessed for the full group of attendees as well as for the patient-only group. While the top question was the same across both groups, differences arose among lower-priority items. Table 5 shows the 20 questions and voting results.
Final List of Research Questions as Prioritized By All Voters (n = 52) and Patient-Only Voters (n = 8)
Questions with a difference >5 in rating between the two groups are shaded in grey.
Both TAB and Steering Committee members identified the fact that the questions ranked most highly by the full group were more focused on assessing outcomes related to telehealth use, while the questions ranked more highly by the patient-only group were more focused on facilitating telehealth use. One patient in the TAB identified the full group votes as “putting the cart before the horse,” suggesting that more attention was needed to understanding and addressing barriers to uptake before looking at outcomes. There was also discussion about the wording used to describe various populations in need, with members of the TAB identifying “marginalized” (as used in question 2) as a word that may be poorly understood or not relatable by patients. The TAB suggested that “disadvantaged” (as used in question 16) is a term that may resonate better with patients. In addition, rewording was suggested to improve question 19, as follows: “How do we use stories to best support telehealth and its impact?”
Discussion
In this work, we conducted a scoping review to identify barriers to and disparities in telehealth uptake among various populations. We then convened a broad range of patients and stakeholders to develop a list of patient-centered research questions focused on understanding and reducing disparities in telehealth use. The selected questions were prioritized by all conference attendees, with voting results presented both for all attendees and exclusively patient attendees. All this work was conducted in close partnership with patient and community advisors to ensure patient centeredness of study conduct and findings.
The primary question as identified by all attendees and the patient attendee subgroup was “What facilitators and barriers to telehealth use are most important to address from the patient and family perspective?” Consideration of this question is vital to inform allocation of funding to support digital health equity. For instance, considerable funding has been allocated over the past few years to address limited technology and internet access, such as the COVID-19 Telehealth Program administered by the Federal Communications Commission. While these programs are intended to support the provision of devices to patients to increase engagement in telehealth, we lack data to confirm that increasing access to devices is in fact a primary priority of patients and families. Though all individuals may not have access to a device, this lack of access may not be seen as a primary issue to telehealth use until higher-order barriers, such as trust in use of telehealth or having privacy to engage in telehealth visits, are addressed. The voting results from this conference suggest the importance of engaging directly with patients and families in various communities to determine most effective allocation of future funding.
While there was agreement on the top priority question among all attendees and the patient subgroup, there were also differences in prioritization. Patients placed more emphasis on trust and cultural factors that affect telehealth uptake and use, whereas professionals prioritized health system and insurance operational barriers. Patients' priority on trust emphasized the importance of expanding conversations regarding telehealth disparities to consider barriers beyond solely “digital literacy.” When considering telehealth-specific uptake, our team has adopted the term “digital health readiness” as its overarching concept, which includes skills, trust, and use of digital tools. 6 This dichotomy between patient and professional priorities is not only interesting but also vital to focus future research funding. Balancing both the systemic function and the human factors promoting and preventing telehealth utilization are the critical “next steps” in telehealth research.
This work provides valuable findings, yet it does have limitations. Regarding the scoping review, we purposefully kept the review very broad to ensure that all relevant texts were included. This meant, however, that the depth of findings presented at the conference was still very high level, as there was too much content to summarize in depth considering the bandwidth of the team and the time allotted to this task. Further work is needed to explore subtopics in greater depth within the scoping review articles. For the conference, while we strove to engage a diverse group, it was not possible to identify and include all potentially relevant populations among our conference attendees, and thus some voices were undoubtedly missed. In addition, while we had the opportunity to explore differences in voting outcomes between the patients and entire group of attendees with both our Steering Committee and TAB members, we lacked opportunity to elicit perspective on these outcomes from many of the conference attendees.
Conclusion
The results from this work provide an important framework for use by both researchers and policymakers. The priorities we have established should help determine what next projects are most likely to advance our ability to understand and address disparities in telehealth uptake. Continued partnership with patients and community stakeholders is essential to ensure that the patient perspective continues to be incorporated into the design, implementation, and interpretation of future work, with the goal of building a more patient-centered care delivery system in which telehealth is a core component that all are able to access.
Footnotes
Acknowledgments
The authors thank the Steering Committee and Telehealth Advisory Board (TAB) members for their dedication and insight to conference planning and implementation—Jacqueline Bayard (TAB), Desiree Collins Bradley (TAB cochair), Kenneth Eng (TAB), Darcel Jackson (SC), Beverley Johnson (SC), Kirsten Meisinger (SC), Penny Parker (TAB), Tabassum Salam (SC), Gail Weingarten (TAB cochair), and Kathy Wibberly (SC). Thank you to Ting Ann Hsiao, Grace Amadio, Lindsey Shughart, and Kiana McIntosh for their contributions to data extraction.
Authorship Contribution Statement
K.L.R. and M.K. are the primary authors responsible for this document. K.L.R., B.W., A.E.L., and A.M.C. conceived the idea for this work. All authors participated in the scoping review process and conference implementation. K.L.R. secured funding for this project.
Author Disclosure Statement
None declared.
Funding Information
This review was funded through a Patient-Centered Outcomes Research Institute (PCORI) Eugene Washington PCORI Engagement Award (21350-TJU). The content presented in this publication is solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee.