When is telemedicine appropriate in the management of head and neck cancer? A mixed-methods assessment among patients and physicians

Abstract

Background:

Evidence suggests oncology patients are satisfied with and sometimes prefer telemedicine compared to in-person visits; however, data are scarce on when telemedicine is appropriate for specific cancer populations. In this study, we aim to identify factors that influence patient experience and appropriateness of telemedicine use among a head and neck cancer (HNC) population.

Methods:

We performed a mixed-methods study at a multi-site cancer center. First, we surveyed HNC patients and analyzed factors that may influence their telemedicine experience using multivariate regression. We then conducted focus groups among HNC oncologists (n=15) to evaluate their perception on appropriate use of telemedicine.

Results:

From Jan-Dec 2020, we collected 1,071 completed surveys (response rate 24%), of which 551 first-unique surveys were analyzed. About half of all patients (56%) reported telemedicine as “same or better” compared to in-person visits, whereas the other half (44%) reported “not as good or unsure”. In multivariate analyses, patients with thyroid cancer were more likely to find telemedicine “same or better” (aOR:2.08, CI95% 1.35–3.25) compared with other HNC populations (mucosal/salivary HNC). Consistently, physician focus group noted that patients with thyroid cancer were particularly suited for telemedicine due to less emphasis on in-person exams. Physicians also underscored factors that influence telemedicine use, including clinical suitability (treatment status, visit purpose, exam necessity), patient benefits (travel time, access) and barriers (technology, rapport-building).

Conclusion:

Patient experience with telemedicine is diverse among the HNC population. Notably, patients with thyroid cancer had overall better experience and were identified to be more appropriate for telemedicine compared to other patients with HNC. Future research that optimizes patient experience and selection is needed to ensure successful integration of telemedicine into routine oncology practice.

INTRODUCTION

Telemedicine is reshaping how patients experience cancer care.^1–5 The COVID-19 pandemic has accelerated adoption of telemedicine across health systems. In oncology, use of telemedicine, and synchronous virtual visits in particular, was imperative to ensure care continuity during the pandemic. Furthermore, it has been recognized to provide a feasible and effective care delivery model for various cancer care services, such as virtual consultation, pretreatment screening, remote patient-monitoring, symptom management, and survivorship care.^6–8

Despite the option to return to in-person visits post-pandemic, telemedicine use remained generally above historic norms, and in some settings up to 38 times higher, than pre-COVID-19 use.⁹ Sustained telemedicine utilization is due in part to enhanced patient and physician acceptance of telemedicine.^{7, 10–13} Further, studies have suggested that patients are not only satisfied with, but sometimes even prefer, telemedicine over in-person visits, as virtual visits potentially offer convenience, flexibility, cost saving, and for certain populations, access to care.^{2, 10} Physicians and health systems have also accustomed to hybrid practice models, facilitated by changes in telemedicine regulations and reimbursement, and largely continue to offer patients options for both in-person and virtual visits when appropriate.^14–17

Naturally, questions arise as to whether, when, and how best to incorporate telemedicine into cancer practices. In other words, when is telemedicine appropriate in replacing an in-person visit? While telemedicine has reported many benefits, concerns have also been raised. In response, the American Society of Clinical Oncology (ASCO) published standards and practice recommendation for telemedicine implementation.¹⁸ These guideline outlined general guidance; however, they also acknowledged significant research gaps. In particular, prospective data remained scarce on defining best practices for specific cancer populations, where research is needed to evaluate the appropriateness of telemedicine in cancer care.

In this study, we assessed the telemedicine experience among a group of subspeciality physicians and patients with head and neck cancers (HNC). HNC encompasses a diverse pathologies (head and neck mucosal, cutaneous, and thyroid cancers) and management modalities (surgery, chemotherapy, radiation).^{19, 20} Importantly, the complexity of managing patients with HNC brings challenges to the integration of telemedicine.^{21, 22} Our study aims are 1) to identify factors that influence patient experience on telemedicine and 2) to better understand providers’ perspective on the appropriateness, barriers and facilitators of telemedicine use among the HNC population.

METHODS:

Study Design

We conducted a mixed-methods study among patients and physicians from Memorial Sloan Kettering Cancer Center (MSK). MSK is a NCI-designated comprehensive cancer center that encompasses 7 clinical sites in New York and New Jersey. MSK is a high-volume cancer center with over 700,000 patient visits annually. From January 2020 to December 2020, we surveyed all MSK patients via an electronic patient portal about their telemedicine experience after each visit as part of routine quality improvement. Surveys were then filtered to include only patients with HNC and each patient’s first completed survey in analysis. Patient sociodemographic and clinical data were extracted and de-identified from electronic health record. Then, we conducted four 60-minute semi-structured focus groups comprised of 3–4 physicians (n=15), who specialized in radiation, medical, and surgical head and neck oncology. No monetary incentives were provided to study participants. This study was reviewed and approved by MSK Institutional Review Board. We report results using GRAMMS framework.²³

Quantitative Survey Methods and Analysis

The survey included questions about patient experience of telemedicine versus in-person visits (primary outcome) and patient satisfaction with telemedicine (secondary outcome). Table 2 outlined our survey questions, which were modified from prior telemedicine survey-based research.^24–26 Responses on patient experience of telemedicine versus in-person visits were dichotomized to facilitate ease of comparison, where ‘Not as Good’ and ‘Unsure’ were combined to represent low desirability and ‘Same’ and ‘Better” were combined to represent satisfactory desirability. Responses on satisfaction with telemedicine were collected on a Likert-type scale, and trichotomized as “Disagree” (“Strongly Disagree,” “Disagree,”), “Neutral” (“Neither Agree or Disagree,”), and “Agree” (“Agree” or “Strongly Agree”).

Table 2.

Telemedicine visit survey items and summary of patients’ responses (N=551).

	Responses
	“Not as Good or Unsure”				“Same or Better”
Survey Questions	N		%		N		%
How did having a televisit compare with having an in-person visit with your healthcare provider?	244		44%		307		56%
	“Disagree”			“Neutral”				“Agree”
	N	%		N		%		N	%
I would recommend televisit to other patients like me.	25	5%		72		13%		454	82%
I would have another televisit with my healthcare provider.a	25	5%		50		9%		473	86%
I was satisfied with the time it took to connect to my provider.b	31	6%		49		9%		466	85%
I was satisfied with the instruction I received about how to connect with my televisit.c	22	5%		25		5%		481	91%

missing data:

^a.n = 3

^b.n = 5

^c.n = 2

Analytic Approach (Primary Outcome):

We performed multivariate logistic regressions to assess the association of patient experience with telemedicine against covariates. Covariates included patient factors (age, sex, race, marital status, primary language, home state, home distance from center, current location at home or clinic), clinical factors (diagnosis, performance status, receiving chemotherapy or radiation), physician factors (profession, current location at main or regional campus), visit factors (type, length, billing level, month, first telehealth visit or not), and a technological factor (device type). First, univariate logistic regressions were performed on each covariate. Significant univariable covariates with p<0.05 were selected in multivariate model. Final model reported adjusted odds ratio with statistical significance set at p<0.05. Analyses were performed using JMP PRO (version 14.0).

Qualitative Focus Group Methods and Analysis

We conducted four 60-minute focus groups with HNC physicians in medical, radiation, and surgical oncology using Zoom videoconference (ZM, San Jose, CA). To ease recruitment, we identified a convenience sample of HNC physicians by their specialty, practice location and availability to participate. We ensured approximate equal representation of each specialty by having at least one medical, surgical, and radiation oncologist in each group. All participants gave verbal consent for audio-recording. A qualitative methods specialist conducted each focus group according to established methodologic guidelines.²⁷ The number and size of focus groups were selected to help achieve data saturation, defined as point at which all themes were fully explored, with no new information arising from additional questioning. Each group was audio recorded and recordings transcribed verbatim for analysis.

Physicians were asked to address semi-structured questions relating to their telemedicine experience with a focus on appropriateness, barriers, and facilitators of telemedicine in four domains: multidisciplinary collaboration, care coordination, symptom management, and continuity of care (see Supplement for Focus Group Guide). To develop the focus group guide, the lead study investigators (E.G.) first reviewed the preliminary patient survey data to identify domains of interest. These domains were reviewed and modified by additional physician stakeholders (T.H, J.C, D.P.). A qualitative methods specialist then converted these domains into conversational focus group questions.

All notes and transcripts were analyzed using thematic content analysis.^27–29 Two study members with qualitative methods expertise (J.G., S.C.) and a trained researcher (N.V.) created an initial codebook with a priori codes derived from interview guide. Two coding members (J.G., N.V.) independently reviewed and coded transcript in NVivo Pro, Version 12.0, highlighting significant statements within each domain and recording analytic memos (i.e., reflections about the segment of text). During coding process , inductive codes were developed on novel concepts that emerged from data and the codebook was updated. Coding team met regularly to reach consensus regarding code definitions and application, discrepancies, and primary themes for transcript. Any inductive codes added to codebook was then applied to previously analyzed transcripts. We analyzed code reports to identify key themes observed across transcripts. Themes included prominent sentiments within each category, and any significant divergences between subgroups (e.g., unique experiences of surgical vs. medical oncologists). Each theme was iteratively revisited to identify instances of theoretical saturation.

RESULTS:

Patient and Physician Characteristics

From January 2020 to December 2020, 4,456 surveys were administered for telemedicine visits, of which 1,071 were completed (24.0% response rate). Of those completed (without missing response), 374 were excluded for ineligible diagnoses, and146 for duplicate surveys (same patient having multiple visits) (See Supplement for CONSORT diagram). 551 completed, first-unique surveys were analyzed (Table 1). Surveyed patients were predominantly ≥ 60 years-old (64%), white (88%), married (74%), English-speaking (98%), lived in New York (58%) and within 50 miles from nearest MSK campus (76%). Patients had diagnoses of mucosal HNC (66%), cutaneous HNC (11%), or thyroid cancer (24%). Majority of patients had performance status with ECOG 0 or KPS ≥ 90 (69%). Half were either undergoing chemotherapy (47%) or radiation (59%). Patients often had telemedicine visits at home (95%), with a medical oncologist (49%), at MSK main campus (59%), and used an Apple device (68%). Majority of visits were for follow-up (85%) and commonly a patient’s first-ever telemedicine visit (57%). Visits usually lasted <20 minutes (61%), and with a billing level of 4 to 5 (moderate to high complexity of medical decision-making) (61%). Telemedicine visits occurred in three sequential time periods in 2020 - Apr to Jun (49%), Jul to Sep (44%), and Oct to Nov (8%).

Table 1.

Baseline of characteristics of telemedicine visits

Characteristic			N = 551	%a
Patient
	Age (years)
		Younger than 60	198	36%
		60 or older	353	64%
	Sex
		Male	268	49%
		Female	283	51%
	Race
		White	462	84%
		Non-White	61	11%
		NA	28	5%
	Marital Status
		Married	402	73%
		Not Married	143	26%
		NA	6	1%
	Primary Language
		English	539	98%
		Non-English	11	2%
	Distance (miles)
		Less than 50	360	65%
		50 or more	113	21%
		NA	78	14%
	Patient Location
		Home	491	89%
		On Campus	24	4%
		NA	36	7%
Clinical
	Diagnosis (cancer)
		Thyroid	131	24%
		Mucosal/Salivary HNC	362	66%
		Skin	58	11%
	Performance Status
		ECOG 0 (KPS 90–100)	156	28%
		ECOG 1 (KPS 70–80)	63	11%
		ECOG 2+ (KPS 60 or below)	6	1%
		NA	326	59%
	Treatment Timing (Chemotherapy/Radiation)
		Completed Treatment	227	41%
		Ongoing Treatment	186	34%
		NA	138	25%
Physician
	Physician Location
		Main Campus	325	59%
		Regional Sites	226	41%
	Physician Department
		Medical	271	49%
		Radiation	140	25%
Visit		Surgery	140	25%
	Visit Type
		New Visit	85	15%
		Follow-up	466	85%
	Time Period (2020)
		Apr. – Jun.	286	49%
		Jul. – Sept.	241	44%
		Oct. – Nov.	42	8%
	Visit Length
		Less than 20 minutes	334	61%
		20 minutes or more	217	39%
Telemedicine
	First Ever Televisit
		Yes	305	55%
		No	233	42%
		NA	13	2%
	User Device
		Handheld	376	68%
		Computer/Laptop	167	32%
		NA	8	1%

^a:Percentages do not necessarily add up to 100% due to missing data.

NA: missing data

Four focus groups were conducted, each represented by at least one surgical oncologist, one medical oncologist, and one radiation oncologist, with 3–4 participants in each group (n=15, totaling 5 medical, 6 radiation, and 4 surgical oncologists).

Patient Experience with Telemedicine (Survey)

Most patients were satisfied with the time (85%) and instructions given to connect with their providers over telemedicine (91%) and would recommend (82%) and have another telemedicine visit with their providers (86%) (Table 2). Approximately half reported telemedicine was the “Same or Better” compared to in-person visits (56%), whereas half reported “Not as Good or Unsure” (44%). In multivariate analyses (Table 3), patients who found telemedicine the same or better than in-person visits were more likely to have thyroid cancer compared with other HNC populations (mucosal/salivary HNC) (adjusted odds ratios [aOR]: 2.08, CI 95% 1.35 – 3.25) and have visits in the time periods Jul to Sept (aOR: 1.49, CI 95% 1.04 – 2.14) and Oct to Nov (aOR: 2.32, CI 95% 1.17 – 4.83) compared with Apr to Jun (Figure 1).

Table 3.

Univariate and multivariate binary logistic regression analysis for baseline characteristics associated with “How did having a televisit compare with having an in-person visit with your healthcare provider?” (“Same/Better” vs. “Not as Good/Unsure”).

		Univariate Analysis			Multivariate Analysisa
Characteristic		Odds Ratio Unadjusted	95% CI	P-value Unadjusted	Odds Ratio Adjusted		95% CI	P-Value
Age
	Younger than 60 years (77/121)	1.41	0.99 – 2.01	0.057
	60 years or older (ref) (167/186)	-
Sex
	Male (117/151)	1.05	0.75 – 1.47	0.773
	Female (ref) (127/156)	-
Race
	White (210/252)	0.68	0.38 – 1.17	0.167
	Non-White (ref) (22/39)	-
Marital Status
	Married (181/221)	0.91	0.62 – 1.33	0.625
	Not-Married (ref) (61/82)	-
Primary Language
	English (239/300)	0.72	0.19 – 2.40	0.599
	Non-English (ref) (4/7)	-
Distance
	Less than 50 miles (155/205)	1.17	0.76 – 1.79	0.472
	50 miles or more (ref) (53/60)	-
Patient Location
	Home (215/275)	1.8	0.79 – 4.24	0.167
	On Campus (ref) (14/10)	-
Diagnosis (cancer)
	Thyroid (39/92)	2.23	1.47 – 3.45	<0.001	2.08		1.35 – 3.25	0.001
	Skin (29/29)	0.95	0.54 – 1.65	0.845	0.94		0.53 – 1.65	0.823
	Mucosal/Salivary HNC (ref) (176/186)	-			-		-	-
Performance Status
	ECOG 0 or KPS ≥ 80 (85/119)	1.87	0.76 – 4.76	0.178
	ECOG 1+ or KPS ≤70 (ref) (12/9)	-
Timing Treatment
	Completed Treatment (105/122)	0.8	0.54 – 1.19	0.272
	Ongoing Treatment (ref) (76/110)	-
Physician Location
	Main Campus (137/188)	1.23	0.88 – 1.74	0.228
	Regional Sites (ref) (107/119)	-
Physician Department
	Medical Oncology (116/155)	1.16	0.77 – 1.75	0.483
	Radiation (63/77)	1.06	0.66 – 1.70	0.81
	Surgery (ref) (65/75)	-
Visit Type
	New Visit (29/56)	1.65	1.03 – 2.71	0.042	1.36		0.82 – 2.26	0.236
	Follow-up (ref) (215/251)	-	-	-	-		-	-
Visit Time Period (2020)
	Apr. – Jun. (ref) (136/132)	-	-	-	-		-	-
	Jul. – Sept, (95/146)	1.58	1.11 – 2.26	0.011	1.49		1.04 – 2.14	0.028
	Oct. – Nov. (13/29)	2.3	1.17 – 4.75	0.019	2.32		1.17 – 4.83	0.019
Visit Length
	Less than 20 minutes (145/189)	1.09	0.78 – 1.54	0.61
	20 minutes or more (ref) (99/118)	-
First Ever Televisit
	Yes (136/169)	0.92	0.65 – 1.29	0.626
	No (ref) (99/134)	-
User Device
	Handheld (159/217)	1.25	0.87 – 1.81	0.224
	Computer/laptop (ref) (80/87)	-

^a.The multi-variate model only includes variables that reached p < 0.05 in the univariate analyses.

ref: reference group in multivariate analysis

Figure 1.

Multivariate binary logistic regression analysis: predicted probability of “How did having a televisit compare with having an in-person visit with your healthcare provider?” (“Same/Better” vs. “Not as Good/Unsure”) for diagnosis, visit time period, and visit type.

Physician Experience with Telemedicine (Focus Group)

We categorized physician experience into three thematic domains: 1) physician preferences for telemedicine; 2) factors influencing patient suitability for telemedicine; 3) benefits and barriers of telemedicine. Illustrative quotations from each domain is presented in Table 4.

Table 4:

Qualitative Results

Theme	Illustrative Quote
Physician Preferences for Telemedicine by Sub-Specialty
Universal Preferences The type of malignancy and treatment stage dictate when a visit is appropriate for telemedicine.	““In the acute subacute period following treatment, I think [telemedicine] is very useful in that situation. And long term follow-up for patients that are doing well and having minimal symptoms, I think there’s a role for telemedicine in that situation also.” “[For patients with thyroid cancer] there isn’t really much of an exam. I don’t even scope those patients. I don’t think that’s useful. Thyroid would be fine with just telemedicine.”
Medical Oncology Patients on active chemotherapy treatment are not suitable for telemedicine due to possible acute side effects.	“If it’s on treatment, for example our chemoradiation patients, their symptoms change very rapidly. Their side effects oftentimes escalate.”
Radiation Oncology Patients receiving radiation treatment need to be seen in-person to assess for adverse treatment reactions.	“During chemo-radiation therapy, I feel those patients have a high need, so I usually see them at least once per week.” “[Patients receiving radiation] have to be seen in person. They need to be seen at least twice a week in person for two weeks after radiation and then as needed, between 2 weeks and two months.”
Surgical Oncology In-person exams are crucial for surgical oncology as they dictate decisions made in the operating room (OR).	“The pliability, the thickness, the mucosal surface changes, those are integral to excellent head and neck surgery. Staring at it with loop magnifications and bright lights and agonizing over every extra millimeter that you take, or the mobility of something that’s stuck versus whether it moves, the texture of the skin, the doughiness, the induration or the suppleness.”
Factors Influencing Patient Suitability for Telemedicine
Disease Characteristics Patients with head and neck cancer are uniquely challenging for telemedicine due to the need for a thorough physical exam for many patients. Patients with thyroid cancer are ideal for telemedicine in the head and neck cancer population due to imaging and laboratory values leading most treatment planning and decisions.
	“[For patients with thyroid cancer] there isn’t really much of an exam. I don’t even scope those patients. I don’t think that’s useful. Thyroid would be fine with just telemedicine.” “…for me it’s thyroid cancer patients that are uncomplicated. We have lots of data on them. We have blood tests…ultrasound that can be done in our regionals, and the physician can review them…There’s very little need for in-depth office physical examination.”
Treatment Stage Patients on long-term follow-up are good candidates for telemedicine.	“…long-term follow-up for patients that are doing well and having minimal symptoms, I think there’s a role for telemedicine in that situation.”
Purpose of Visit Visits primarily focused on imaging and lab results are appropriate for telemedicine. Appropriateness of telemedicine for symptom management is dependent on the symptoms being reported.	“I think if they have good imaging, especially if they’re coming in from far away, I often feel comfortable reviewing that imaging with them [via telemedicine]. And I feel like the exam doesn’t always add that much or it’s rare that the exam and the imaging is discordant and that’s a way where you can save them a 2-hour drive if they’re coming in from far away.”
	“…some skin reactions or some blistering we can see remotely, if they’re in a room that has good lighting.” “There might be some situations like hearing issues or even neuropathy [that would be suitable for telemedicine].” “…nurses do some remote care immediately after completion of treatment because the first several weeks tend to be a really tough time for patients when they’re not being seen on a regular basis anymore.” “I think when patients are very symptomatic, I feel very compelled that the patient has to come in to be examined.”
Benefits of Telemedicine
Telemedicine facilitates increased access to care for patients.	“It has made it feasible to sometimes add on a patient on a day when I wouldn’t have necessarily had clinic time, but a patient needs to be seen.”
Telemedicine saves patients travel time.	“…if they are here for new consults and they saw the surgeon two days ago, then having them come in so I can examine them is probably not going to add a lot to what we already know. That’s a context where you can avoid the burden on the patient in terms of coming in for three separate appointments and taking three separate days off to do that.”
Telemedicine allows for continuity of care for patients who live far away.	“It certainly makes follow up for patients that live far away and would otherwise skip their visits possible. It’s been helpful in that circumstance.”
Barriers to Telemedicine
It is difficult to build rapport with patients over telemedicine due to lack of in-person cues such as eye contact and body language.	“When we’re seeing patients before surgery, that human rapport is so important beyond even feeling the tumor or feeling the number of nodes or looking at the pliability of the skin; just to have that really close, trusting relationship before someone empowers you and trusts you to open their neck up is something that’s indescribable.”
Mixing telemedicine and in-person clinic appointments creates bottlenecks in physician workflow and often results in patients waiting online for visits to begin.	“It’s hard to have telemedicine visits randomly inserted [amongst in-person appointments] because I’m getting paged to check films and there’s status checks and the telemedicine patients, they get really concerned if it’s time for their visit and I haven’t joined.”
Technology remains a barrier to effective telemedicine visits as some patients may not have the abilities or utilities required to have an adequate visit.	“Sometimes you have a tele-visit scheduled and then the patient’s video doesn’t work or something and then it just turns into a phone visit, which is even less useful.” “It’s mostly the elderly, but sometimes younger patients who just are not computer savvy, who cannot connect and/or they live in an area that has very bad reception.”

Physician Preferences for Telemedicine by Specialty

Providers from three subspecialties (medical, radiation, and surgical oncology) generally agreed that malignancy type and treatment stage were among the important factors for deciding the appropriateness of telemedicine. Specifically, most (but not all) medical oncologists stated that patients receiving active chemotherapy should not be seen on telemedicine due to rapidly changing symptoms and possibility of acute side effects. Radiation oncologists referred to the simulation visit as an opportunity to perform an in-person exam as supplement to telemedicine visit at consultation, reducing number of visits for patients. Surgical oncologists cited that in-person exams are particularly crucial to dictate decisions made in the operating room.

Factors Influencing Patient Suitability for Telemedicine

Disease Characteristics:

Physicians across all subspecialties described physical exam as a defining aspect of visit appropriateness. They agreed that cancer type plays a major role in determining telemedicine appropriateness. Thyroid cancer was reported to be most appropriate due to limited reliance on physical exam (versus laboratory tests) for treatment decisions and patients often being healthier at baseline. Other malignancies specifically referenced as suitable for telemedicine were salivary and skin HNC.

Treatment Stage:

Oncologists agreed that patients on active treatment are particularly unsuitable for telemedicine while those on long-term follow-up are often better candidates because patients are generally healthier and an exam being less necessary. Additionally, telemedicine aided long-term follow-up visits for those living far from MSK.

Purpose of Visit:

Physicians agreed that several visit types were appropriate for telemedicine. These include visits to review data (e.g., imaging and labs), to tapper medication, or to review treatment plans. These visits were thought to be suitable for telemedicine as they often do not require an in-person exam. Physicians also reported that telemedicine was suitable for certain symptom management . In general, management of skin reactions, hearing changes, or neuropathy were reported to be more suitable for telemedicine, whereas symptoms of the oral cavity would not given the need for in-person exam. In addition, allied health members such as nursing to perform remote care were thought to be appropriate use of telemedicine, especially once patients have completed treatment or when collecting history prior to a new consultation.

Benefits and Barriers of Telemedicine

Benefits to Patients:

Physicians noted that telemedicine shorten wait times, allowing for greater flexibility in connecting with new patients. Physicians also noted that telemedicine increases continuity of care with patients, both for those who live far away and for being able to see patients sooner. Avoidance of travel for patients was frequently cited, with physicians particularly willing to use telemedicine if benefiting a patient who lives far away, although not in substitution for an in-person exam if needed.

Barriers to Effective Use:

In addition, in-person exam being a primary deterrent for telemedicine use, oncologists referenced the difficulty of building rapport. They cited the lack of body language and human interaction as major barrier to building physician-patient relationship. Additionally, they found that when telemedicine was built into in-person clinic schedules, time management became a problem due to delays with in-person clinic appointments often resulting in delays for telemedicine appointments. Technology was another factor commonly cited as a barrier to effective telemedicine use. Physicians stated that their patient populations have varying levels of technological abilities and access to adequate devices.

DISCUSSION:

In this study, while most HNC patients were satisfied with telemedicine visits, only half (56%) would consider them the be the same or better than an in-person visits, whereas half (46%) would consider them to be not as good or unsure. Patients with thyroid cancer were more likely than patients with mucosal/salivary or skin HNC to rate their telemedicine experience as the same or better than an in-person visit. HNC oncologists also reported thyroid cancer patients as one of the most appropriate patient populations for telemedicine.

The consensus between physicians and patients on appropriateness of telemedicine by cancer type is important, particularly for facilitating visit triage. Specifically, HNC oncologists across medical, radiation, and surgical sub-specialties referenced patients with thyroid cancer as optimal for telemedicine due to the in-person physical examination playing less of a role in care and treatment decisions. Additionally, laboratory and imaging data, as well as patient-reported symptoms that tend to not require a physical exam, drive many treatment decisions for patients with thyroid cancer, all of which can be feasibly addressed via telemedicine. This appears consistent with specialty-specific trends in use of telemedicine showing endocrinology among the highest utilizers^{30, 31} in which thyroid disease is commonly managed. Meanwhile, skin and salivary cancer may also be appropriate, while mucosal HNC less optimal, unless an exam can be deferred to a subsequent pre-treatment visit as in radiation oncology.

Much of the existing literature on telemedicine in oncology focuses on the general oncology population. We found that many of these findings are consistent for the HNC population as well. For instance, the high satisfaction rate has been shown in other oncology populations^{3, 5, 25, 32, 33}, technology barriers impacting feasibility of telemedicine^{15, 17}, visit types dictate appropriateness (e.g., new consultations)¹⁸, as well as management of long-term treatment^{6, 34} and discussion of laboratory and imaging results^{18, 35} are feasible via telemedicine. We did observe improved patient experience on telemedicine compared to in-person visits over the study period, which may be due to improvement of telemedicine technology, or an increased in patient and physician familiarity with telemedicine platform; however, it is also possible that there is selection bias for patients later in study period due to an increasing option to have in-person visit.

Another finding consistent with prior literature was the perceived benefit to patients with prolonged travel time to in-person care. This was consistently cited among oncologists in the focus groups. They stated telemedicine provided a method to communicate with patients easily who would otherwise have an extended commute to the hospital. Interestingly, our survey found that patient distance to MSK campuses was non-significant, though this could be related to limitations in measurement of travel burden to MSK. It is possible that the way in which distance was measured was not reflective of patient travel time in a city with multiple modes of transportation. A cutoff of 50 miles from MSK may also be too large to detect a meaningful difference for this geographic location, in contrast to prior studies.³⁶

We found that oncologists commonly referenced telemedicine as a barrier to developing patient-rapport; however, consistent with other studies^{35, 37}, this concern did not impact patient satisfaction on telemedicine. Interestingly, literature has also suggested that some patients may build better rapport with their physicians with telemedicine.^{32, 38} These variations underscored that telemedicine experience are shaped by multifaceted factors, including telemedicine use context and individual preferences.

Importantly, our findings highlight the challenges in caring for patients HNC via telemedicine. The need for an exam is particularly crucial in the HNC population given subtly of critical complications such as airway compromise and pathologies (i.e. intraluminal lesions) may be difficult to visualize on imaging. Our focus group also suggest that HNC patients on active treatment are less suitable for telemedicine, in that patients often are treated with multimodal therapy (i.e. combination of surgery, chemotherapy and radiation), where concerns for rapidly evolving symptoms make assessment of acute side-effects difficult. Granted, literature in other cancer populations have suggested patient- and system-level benefits of remote systemic therapy monitoring in reducing emergency department visits, unplanned admissions, and cost of care.^{39, 40} This highlights the importance of cancer-type specific guidelines for assessing telemedicine care suitability over a “one-size-fits-all” approach.

Limitations

We acknowledge several limitations. First, our survey response rate was 24%, although consistent with other survey-based research^{5, 13, 25}, limited generalizability. Our survey was in English-only and our study was conducted at a single, multi-site institution; thus, results may not be representative of other healthcare settings, or patient populations with different sociodemographic, language, or technological literacy background. Survey responses were dichotomized to ease computation and limited by statistical power of our sample size. Given resource constraints, we did not collect specific data in our survey on when patients prefer to have telemedicine or in-person visits. Also, we did not include patient interviews or additional perspectives from allied healthcare members (nursing, etc.). These limitations should be considered for exploration in future research.

CONCLUSION:

This study highlights the intricacy involved in determining when and who telemedicine is appropriate for the HNC population from the perspective of oncologists and patients. Our results suggest that patients with thyroid cancer may be better suited for telemedicine compared to other HNC. As telemedicine becomes more commonly adopted, future research to optimize patient experience and selection is needed to ensure successful integration of telemedicine in oncology practices.

Context Summary.

Key Objective

In head and neck oncology, what factors influence patient experience and physician perception when determining appropriate use of telemedicine compared to in-person visits?

Knowledge Generated

About half of all surveyed patients with head and neck cancers (56%) reported telemedicine as same or better compared to in-person visits, whereas the other half (44%) reported not as good or unsure. Among them, patients with thyroid cancer had overall better experience with telemedicine, consistent with physicians identifying them as better suited for telemedicine because their care emphasize less on in-person exams. Notable factors that influence physician perception on telemedicine use are diverse, including assessment for clinical suitability (treatment status, visit purpose, exam necessity), patient benefits (travel time, access) and barriers (technology, rapport-building).

Relevance

Understanding patient experience and physician perception on telemedicine appropriateness help improve personalization of care as telemedicine continue to become more integrated into oncology practices.