Abstract
Introduction:
Telemedicine use expanded rapidly during the COVID-19 pandemic, but publications analyzing patient perspectives on telemedicine are few. We aimed to study whether patient perspectives offer insights into how best to utilize telemedicine in the future for hematology and cancer care.
Methods:
A modified Telemedicine Satisfaction and Usefulness Questionnaire (TSUQ) was sent to adult hematology/oncology outpatients at the University of Minnesota Masonic Cancer Clinic who had ≥1 prior phone and/or video visit between March 15, 2020, and March 31, 2021. Two focus groups were subsequently conducted with volunteers who completed the survey. We evaluated dichotomized TSUQ items using logistic regression, and focus group data were analyzed qualitatively using constant comparison analysis.
Results:
Of 7,848 invitations, 588 surveys were completed. Focus groups included 16 survey respondents. Most respondents found telemedicine satisfactory, easy to use, and convenient, with the majority preferring a hybrid approach going forward. Oncology patients, females, and higher income earners endorsed decreased telemedicine satisfaction. Concerns were voiced about fewer in-person interactions, communication gaps, and provider style variability.
Discussion:
Adult hematology/oncology patients had varied perspectives on telemedicine utilization success based on gender, income, and disease burden, suggesting that a one-size-fits-all approach, as was implemented nearly universally during the COVID-19 pandemic, is not an ideal approach for the long term. Given that telemedicine use is likely to remain in some form in most centers, our findings suggest that a nuanced and tailored approach for some patient subgroups and using feedback from patients will make implementation more effective.
Introduction
The onset of the COVID-19 pandemic drove an abrupt but necessary worldwide transition to telemedicine across all aspects of medicine whenever possible to pragmatically minimize infectious exposures. However, the COVID-19 pandemic continues in most of the world, and its effects on the practice of medicine, in particular for hematology/oncology care, still resonate over 2 years after it began. Early studies into the effects of telemedicine on medical practice primarily focused on implementation and the effects on medical staff, or alternatively were patient-focused, but could only address immediate changes at the beginning of the pandemic. 1 –3 Over the ensuing months, health care systems everywhere developed hybrid models of in-person and virtual care as the new norm, but published patient perspectives on these long-term changes remain limited. 4 –6
We conducted a mixed-methods study including a validated patient survey about telemedicine along with focus groups, with the aim to gain insight into patient perspectives and views on the long-term role of telemedicine in hematology/oncology care, ultimately with the goal of optimally addressing patient needs with hybrid care models going forward.
Methods
STUDY DESIGN AND POPULATION
This cross-sectional study was approved by the University of Minnesota Institutional Review Board. All volunteers were provided detailed information about the purpose of the study and confidentiality of results of the survey in which no identifiable information would be linked to responses. Participants of the focus groups, after voluntarily providing their contact e-mails, underwent a full verbal consent discussion with associated signed e-consent to adhere to institutional COVID restrictions regarding limited in-person interactions. The study was conducted involving individuals who had had at least one scheduled and completed telemedicine visit (telephone and/or video) at the University of Minnesota/M Health Fairview's Masonic Cancer Center (MCC) between March 15, 2020, and March 31, 2021. Telemedicine visits included appointments with physicians, advanced practice providers, and/or genetic counselors, and involve hematology, oncology, bone marrow transplantation (BMT), palliative care, and/or genetic counseling subspecialties.
Individuals were excluded from the study if they received some care within the MCC, but only had telemedicine visits with specialties other than that listed previously, if they did not have active access to electronic communication through MyChart (Epic Systems Corporation, Verona, WI) at the time of the survey, if they could not read English to complete the questionnaire, and/or if they did not have any completed telemedicine visits during the time frame.
STUDY PROCEDURES
An invitation to the survey was sent through MyChart to all the eligible MCC patients. The survey was open for 28 days; data collection was conducted through REDCap. 7 While no contact information was required to be shared for survey completion, interested participants could volunteer for a follow-up focus group for a $20 stipend by sharing an e-mail address upon survey completion. This contact information allowed for contact to discuss focus group details over the phone before e-consent. Potential focus group volunteers were screened before outreach to ensure that a demographic mix representative of the MCC patient population was involved. Screening factors included race and/or ethnicity, age, place of residence to ensure a mix of rural, urban, and suburban perspectives, and whether patients primarily saw classical or malignant hematology, oncology, BMT, palliative care, and/or genetic counseling. Of note, while age and place of residence of survey respondents were omitted from the survey to maintain anonymity, these data were made available if a participant volunteered for the focus group to conduct the screening.
Within 6 weeks of survey closure, two focus groups were held, each lasting 75–90 min. These focus groups were conducted virtually due to the ongoing COVID-19 restrictions, after prior e-consent for recording and aggregate data sharing. The focus groups were conducted with audio/visual recording over Zoom videoconferencing software (San Jose, CA) with secure invitation-only access through the University of Minnesota. The sessions were led and securely recorded in an HIPAA-compliant manner by independent nonclinical professionals within the University of Minnesota's Office of Measurement Services (OMS), with author A.A.B. present for both groups solely for question clarifications.
The focus group questions were based largely on survey feedback and were standardized ahead of time with OMS staff and author A.A.B. to ensure accurate comparisons between the two sessions. The same two OMS staff members were present for both groups, with one leading the discussion while the other recorded real-time thematic observations and nonverbal communication to augment the transcribed recordings.
MEASURES
In the first phase of the study, we conducted a cross-sectional survey using a modified version of the 21-question Telemedicine Satisfaction and Usefulness Questionnaire (TSUQ). 8 Answer scores on each TSUQ item range from 1 to 5, with 1 representing strong disagreement with the statement and 5 showing strong agreement, which we dichotomized as 1–2 (disagreement with statement) versus 3–5 (neutral/agreement with statement). Subscores for Satisfaction and Usefulness were also collected, also on a 1–5 scale. TSUQ questions were amended to include telephone visits in questions that only addressed video visits in the original version. One question was removed because it referred to in-person clinic visits, which were not within the scope of our study due to the pandemic. Three questions were added to address telemedicine access and three more to ask about effects of the pandemic on patient care in the past and anticipated in the future.
Self-reported demographic information included gender (female, male, other), race and/or ethnicity (American Indian, Alaskan Native, black/African American, Hispanic/Latinx, Native Hawaiian/Pacific Islander, multiple races, white, other), annual household income in U.S. dollars (less than $20,000, $20,000–49,999, $50,000–74,999, $75,000–99,999, $100,000–149,999, ≥$150,000, prefer not to answer), maximum education level (no high school degree or high school degree, vocational or business school, some college, college or university graduate [bachelor's or equivalent], graduate, or professional training [graduate degree]), marital status (married or partnered, divorced or separated, widowed, single, prefer not to answer), and diagnosis-related questions. All responses were anonymous and compiled in aggregate for analyses.
ANALYSES
Frequencies of survey responses and demographics were summarized and compared using chi-squared tests. We present mean, standard deviation (SD), median with interquartile ranges as appropriate, and range and 95% confidence intervals (CIs), overall and by disease groups, of the TSUQ responses. We used one-way analysis of variance and two-sample t tests to compare response scores between different demographic and clinical subgroups. Adjustments for multiple comparisons were carried out using the Bonferroni correction. We used multiple logistic regression, adjusted for age (<65 years, ≥65 years based on self-reported enrollment in Medicare vs. not), gender, race (non-Hispanic white, other), income (<$50,000, $50,000–99,000, ≥$100,000, prefer not to say), education (no college degree, at least college degree), and having cancer (yes, no). Analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC).
Constant comparison analysis was used to evaluate focus group data. Using the recordings in combination with the real-time documentation, responses were coded in descriptive units that were subsequently grouped into categories. Larger themes were delineated from these categories. A comparison of the two focus group discussions allowed for an analysis of generalizability in experience and feedback. University of Minnesota IRB Registration number STUDY00012805.
Results
PARTICIPANTS
A total of 7,848 survey invitations were sent out, of which 588 surveys were completed (7.5% response rate). Full demographics are shown in Table 1. Respondents' characteristics generally mirrored the overall group of recipients, with 72.1% females (63.4% females invited) and 93.8% self-reported non-Hispanic white, compared with 87% among those invited. Cancer was the primary diagnosis for most recipients (74.4%) and respondents (74.7%). Among all respondents, 36.5% were off therapy and another 17.9% receiving long-term maintenance therapy. Of the 30.2% of respondents without a cancer diagnosis (not including BMT), 9.6% had seen BMT specialists, 7.9% received classical hematology care, and 8.2% had other visits, including genetic predisposition visits and palliative care. Most (70.8%) of those invited were ≥50 years old, and most (73.5%) were from the Minneapolis/Saint Paul metropolitan area, while 21.8% lived in rural areas.
Survey Participant Characteristics
TSUQ RESULTS
Participants felt that telemedicine implementation during the first year of the pandemic was overall successful. Results for the TSUQ are shown in Table 2. The mean satisfaction subscore was 3.8/5.0 (SD 0.92), and the usefulness mean was 3.4/5.0 (SD 0.92). Telemedicine was most often delivered by video (64.2%), with 20.9% having only telephone visits and 14.9% having used both visit methods. The convenience of telemedicine was appreciated by 87.4% of respondents, and 92.1% felt that it saved time. Notably, a large majority of patients, 78.6%, would prefer a mix of telemedicine visits in the postpandemic era (Table 3).
Telemedicine Satisfaction and Usefulness Questionnaire Results
The total number of respondents was 588. The TSUQ scale (range 1–5) was grouped into 1–2 ratings as “Disagree,” while 3–5 were categorized as “Neutral/Agree.”
IQR, interquartile range; SD, standard deviation; TSUQ, Telemedicine Satisfaction and Usefulness Questionnaire.
Telemedicine Visit Characteristics and Preferences
Data include 584/588 (99.3%) of respondents.
However, telemedicine did not work for all patients equally well. Access to telemedicine options was a concern for some respondents, with 2.3% of them saying that telephone access for telemedicine visits was an issue and 6.7% for video. Functionality was also problematic at times; 18.5% reporting difficulty using telemedicine modalities, and 29.9% saying that the software sometimes malfunctioned. One-third of respondents felt that the lack of physical interaction during the visit was a concern and 37.3% found that verbal communication through telemedicine was less satisfactory than when done in person. While 80.2% said their providers had adequate understanding of the patient's problem during a telemedicine visit, only 62.7% said that talking to a provider through telemedicine was as satisfying as talking in person.
New concerns were less than half as likely to be addressed during a telemedicine visit versus ongoing concerns (18.2% vs. 40.8%); 41.1% said that no acute concerns were addressed over telemedicine (Table 3).
In the adjusted analyses, there were distinct differences of perspective between some patient subgroups. Those with reported incomes less than $50,000 annually favored telemedicine compared with those with incomes ≥$100,000 (adjusted odds ratio [OR] 2.47, 95% CI 1.16–5.28). Those with a cancer diagnosis were less likely to prefer telemedicine in the future compared with other populations (OR 0.52, 95% CI 0.34–0.81). Males preferred telemedicine compared with females (OR 1.68, 95% CI 1.0–2.83).
FOCUS GROUP RESULTS
Out of 120 individuals who volunteered to participate in focus groups, 16 were selected. Their median age was 63 years (range 22–81 years), with 67% females. Most (87.5%) were non-Hispanic white with a nearly even mix of rural, suburban, and urban residents. Eleven participants had solid tumor diagnoses, two had classical hematology diagnoses, two others had telemedicine cancer predisposition screening visits, and the final participant had a hematologic malignancy. Their overall duration of time that focus group participants had received care within the MCC spanned between 8 months and 30 years. They had between 1 and 30 telemedicine visits during the study time frame, not including electronic communications with care providers. Recurrent themes from the focus groups that are highlighted below are shown in Table 4 along with sample quotations.
Focus Group Themes
Telemedicine was generally received favorably among focus group participants. They shared an increasing comfort level with digital technology and communication given the in-person restrictions. For some, particularly those living in rural settings or those who had genetic counseling visits, virtual care was well-received as it allowed for reduced time away from work and family. It also helped mitigate transportation and parking challenges, particularly in winter. Since many participants had active or past malignancies, telemedicine was favored from an infectious disease prevention standpoint, both for COVID-19 and other infections, allowing for more peace of mind. However, most participants, including some of the same individuals who valued the improved safety, struggled with the lack of in-person engagement.
The paucity of physical interactions with care teams reduced morale and left some participants feeling like they were alone in their care journey at times. None of the focus group participants with cancer favored a predominantly telemedicine model of care going forward.
Similar to survey results, focus group members shared mixed views on the video software functionality. For those with cancer, there was a general sentiment that the adjustment to telemedicine in the midst of therapy was quite difficult. Discussions around the topic of telemedicine communications, both in terms of virtual clinic visits and other contacts such as phone calls from team members and messaging through the electronic medical record, produced the most disparate answers in the two focus groups. On the positive side, participants stated that both patients and providers were more vigilant about monitoring inboxes for communication, allowing for more rapid contact, particularly for minor questions or medication requests. Some participants reported that scheduling became easier too.
However, concerns were raised about test results being shared early and at inopportune times, before the medical provider being able to discuss them in context. Furthermore, some participants transitioned between physicians during the pandemic, making the development of therapeutic relationships complicated. Finally, many felt that there was a wide variation in providers' ability to personally connect and engage virtually compared with the prepandemic relationships. This was most prominent for patients with solid tumors compared with those with nonmalignant disorders. Despite the concerns raised, most emphasized they would prefer a hybrid approach in the future, with minor visits being done virtually but with a prioritization for in-person visits. They also preferred some control over what visit method to use.
Discussion
The COVID-19 pandemic dramatically altered patient–provider engagement. Early telemedicine implementation research primarily focused on initial ripple effects and their effects on health care systems, providers, and patients. 3,9 –12 More recent research has shed light on the mixed successes and stresses that telemedicine places on medical systems and providers, particularly in hematology/oncology and related fields. 13 –16 As vaccines have become widely available and generally less virulent COVID-19 variants have emerged, a gradual transition back to in-person care has been adopted, but telemedicine will likely remain a care option for some. Therefore, systems must understand how to best serve patients with hybrid approaches. In this study, we show patient perspectives that illuminate specific aspects of telemedicine that may not be appreciated by research prioritizing system-based implementation.
Overall, our patients reported that telemedicine was successfully executed as the pandemic evolved and they see value in it for the future, although in-person visits were preferred. This general sense echoes previous published literature. 15,17,18 Yet the details reveal important patterns that may open opportunities to better practice individualized medicine. It is not surprising that individuals who live far from their care center, particularly in rural communities, see value in the improved convenience of telemedicine. 19 However, this must be balanced with the loss of significant personal interactions and nonverbal communication. Those with cancer may feel this gap most palpably given the intensity of therapeutic relationships. Our focus groups' feedback enhanced the survey data findings, emphasizing that while information could be reliably shared through telemedicine platforms, patient satisfaction with regard to personal connections is more mixed.
Previous small interview-based studies on the subject, some of which took place prepandemic, reflect this sentiment. 18 The isolation inherent in COVID-19 lockdowns may have amplified the sense of isolation. Survey respondents also reported that only a minority of telemedicine visits addressed new concerns, with a large percentage of visits not having active concerns addressed at all. This may contribute to the perception that telemedicine visits were not as fulfilling as in-person appointments.
Patients who had hematology, palliative care, and genetic screening telemedicine visits were more likely to be satisfied with telemedicine than those with cancer in our study. This point is worth noting for medical centers, including ours, in which the majority of patients have a malignancy diagnosis. System decision-makers may jump to the conclusion that what works best for cancer-related care works best for everyone given the propensity of diagnoses. This could lead to a vast reduction in telemedicine use that could be detrimental to patients without alternative diagnoses. In addition, technological access remains profoundly important among patients living in rural parts of the country, including much of Minnesota, with the pandemic magnifying these access gaps. 20 One study found a significant reduction in telemedicine access in patients with Medicare who lived in rural areas, negating any value of convenience. 21
In a separate study, our team previously reported inequitable use of telemedicine in cancer survivors, particularly for those who were older or who were rural residents. 22 Because our survey was sent using the electronic communication embedded within the medical record system, we recognize that our study's responses may have been biased because those individuals disinclined to use digital communication were also less likely to do the electronic surveys. Yet even among respondents, mostly in-person visits were preferred. Ultimately, in hematology/oncology care, telemedicine will have value for certain patient subsets, particularly outside of oncology and for brief check-ins, but in-person interactions should be prioritized.
Our mixed-methods approach offered more balanced and comprehensive insights than survey-only studies could do, but there were also limitations. The cross-sectional survey was intentionally done at the 1-year point of the pandemic to best understand patient perspectives when telemedicine use was maximal, but perspectives will not be static over time. In addition, our response rate of 7.5% to the survey increases the risk of availability bias, but the raw number (N = 588) of responses was still robust, and the patient-reported demographics mirrored the demographics of those invited. We feel that these characteristics offer some generalizability for our findings. The anonymous nature of the survey necessarily limits some degree of demographic analyses, but we feel that this approach allowed for more truthfulness by survey respondents. As this was more of a needs assessment than an intervention pilot, future studies can provide more granular guidance on telemedicine optimization for patient subgroups.
The future of medical care in hematology/oncology centers should include a mix of in-person visits and telemedicine with an eye toward disease-specific and even individualized approaches. That will require some alterations to the care patterns that can become entrenched in high-volume medical centers. In that light, we must gain patient perspectives to develop the best hybrid approaches and re-engage as often as necessary. For hematology/oncology centers, it is critical to engage patients without cancer diagnoses in these efforts to avoid a one-size-fits-all approach to systems-level changes. Through this mixed-methods approach, we gained insights into the more nuanced aspects of patients' telemedicine experiences. We hope our findings contribute to patient-centered outcome research to design appropriate telemedicine models going forward.
Footnotes
Acknowledgments
We would like to acknowledge all of the patients who provided feedback to improve our division's approach to telemedicine. We would like to recognize Mark Miazga, JD, Adam Lenczuk, and Melanie Forstie from the University of Minnesota's OMS for their critical coordination of the focus groups. We also would like to recognize Melissa Schedler and Alyssa Erickson from M Health Fairview for their assistance with survey distribution and demographic analyses, respectively.
Authors' Contributions
A.A.B.: Conceptualization (lead); project administration (lead); methodology (equal); writing—original draft (lead); formal analysis (equal); and writing—review and editing (equal). P.I.J.: Conceptualization (supporting); methodology (equal); formal analysis (equal); and writing—review and editing (equal). S.G.H.: Conceptualization (supporting); methodology (equal); and writing—review and editing (equal). B.R.L.: Conceptualization (supporting); methodology (equal); formal analysis (equal); and writing—review and editing (equal). J.Y.C.H.: Conceptualization (supporting); methodology (equal); and writing—review and editing (equal). A.H.B.: Conceptualization (supporting); methodology (equal); and writing—review and editing (equal). All the authors contributed equally to the article drafting and agree with the text in the final article.
Disclaimer
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Disclosure Statement
The authors declare that there are no conflicts of interest.
Funding Information
Funding was received by the Division of Hematology, Oncology, and Transplantation within the Department of Medicine at the University of Minnesota. Research reported in this publication was also supported by the National Center for Advancing Translational Sciences of the National Institutes of Health Award Number UL1-TR002494.