Feasibility and acceptability of a rural, pragmatic, telemedicine‐delivered healthy lifestyle programme

Summary Background The public health crisis of obesity leads to increasing morbidity that are even more profound in certain populations such as rural adults. Live, two‐way video‐conferencing is a modality that can potentially surmount geographic barriers and staffing shortages. Methods Patients from the Dartmouth‐Hitchcock Weight and Wellness Center were recruited into a pragmatic, single‐arm, nonrandomized study of a remotely delivered 16‐week evidence‐based healthy lifestyle programme. Patients were provided hardware and appropriate software allowing for remote participation in all sessions, outside of the clinic setting. Our primary outcomes were feasibility and acceptability of the telemedicine intervention, as well as potential effectiveness on anthropometric and functional measures. Results Of 62 participants approached, we enrolled 37, of which 27 completed at least 75% of the 16‐week programme sessions (27% attrition). Mean age was 46.9 ± 11.6 years (88.9% female), with a mean body mass index of 41.3 ± 7.1 kg/m2 and mean waist circumference of 120.7 ± 16.8 cm. Mean patient participant satisfaction regarding the telemedicine approach was favourable (4.48 ± 0.58 on 1‐5 Likert scale—low to high) and 67.6/75 on standardized questionnaire. Mean weight loss at 16 weeks was 2.22 ± 3.18 kg representing a 2.1% change (P < .001), with a loss in waist circumference of 3.4% (P = .001). Fat mass and visceral fat were significantly lower at 16 weeks (2.9% and 12.5%; both P < .05), with marginal improvement in appendicular skeletal muscle mass (1.7%). In the 30‐second sit‐to‐stand test, a mean improvement of 2.46 stands (P = .005) was observed. Conclusion A telemedicine‐delivered, intensive weight loss intervention is feasible, acceptable, and potentially effective in rural adults seeking weight loss.

Number: U48DP005018; Friends of the Norris Cotton Cancer Center at Dartmouth and National Cancer Institute Cancer Center Support, Grant/Award Number: 5P30 CA023108-37; National Institute on Aging of the National Institutes of Health, Grant/Award Number: K23AG051681

| INTRODUCTION
As a major public health crisis nationally and internationally, obesity rates continue to rise, exceeding 1 an estimated 38%. Obesity is known to adversely impact cardiometabolic factors, including hypertension, diabetes, and dyslipidemia, 2 ultimately increasing vascular risk and leading to disability 3 and death. 4 The escalating costs associated with obesity in the United States demonstrate a critical need to address this epidemic 5 given that direct and indirect costs account for 9.3% of the gross domestic product. 6 Conditions are even worse in rural areas of the United States where obesity prevalence rates are much higher 7 and patients often need to travel extensive distances to access health care services and specialist providers. [8][9][10][11] Disparities in accessing care are especially problematic in caring for patients with obesity, where regular interactions are needed to promote health behaviour change. 12 Mobile health interventions hold promise in engaging adults with obesity in behavioural change. For instance, self-monitoring using commercial applications has demonstrated an increased likelihood of short-term weight loss. 13,14 Goal setting through text messaging, 15 automated voice response systems, 16 or tailored self-monitoring platforms 17 can all enhance success and are cost-effective strategies to atrisk populations. However, engagement drops off after initial usage suggesting a need for more personalized approaches. 18 In fact, at 12 months, there are no differences in weight loss between digital and control arms. 19 Recent studies using just-in-time adaptive interventions also hold considerable promise in influencing patterns of behaviour for engaging and sustaining weight loss. 20 The emergence of telemedicine, two-way live video-conferencing, has been embraced by the Centers for Medicare and Medicaid services 21 as a different type of mobile health modality that can potentially surmount geographic barriers to health care delivery. With the advent of policy changes promoting rural broadband and cellular access, 22 telemedicine is increasingly available to both health care entities and patients alike. In rural obesity management, telemedicine is particularly promising because it reduces demands on patients' time by reducing the need to travel long distances and spend hours away from work 23 in order to attend high-intensity visits recommended by the 2013 guidelines. 12 While an initial investment is needed, the payoff is significant in that it may reduce costs. 24 The affordability can allow rural patients increased access to specialists, making it a plausible method to deliver care.
Few trials have evaluated the use of telemedicine in obesity management. The Veterans Affairs MOVE! trial has implemented telemedicine in effective and sustainable approaches. 25,26 Their programme, although, focused only on veterans with obesity across the United States and was not specific to rural areas. The delivery was based on using a telehealth monitor delivering electronic modules, rather than using a clinical care provider. Other studies focus on paediatric populations with hybrid models (in-person and remote), [27][28][29][30] or the potential efficacy of low-intensity models. 31,32 Studies have demonstrated mixed results in other populations, including pregnancy 33 or endometrial cancer survivors. 34 While diet-quality and obesity are strongly associated with rural health care resource use, 35 there is a lack of pragmatic research strategies for delivering high-intensity obesity therapy in rural areas. Our hypothesis in this pilot study was that an adaptation of an in-person, 16-week intensive lifestyle intervention could feasibly be delivered using telemedicine and would be acceptable and potentially effective for participants.

| Study design and setting
This was a single-arm, non-randomized study by enrolling participants

| Intervention description
The Healthy Lifestyle Program consisted of a 16-week programme delivered by a health coach, registered dietitian, and nurse exercise specialist (see Table 1) focused on health-behaviour change and based on the Diabetes Prevention Program. 37  after the initial evaluation, in lieu of in-person care. Other participants were eligible for bariatric surgery or medication management and did not enter this programme. The structure and operational infrastructure paralleled that observed within on-site care. In addition, participants were provided with a wearable fitness device during the study to enable them to track their physical activity as part of a separate research study.

| Telemedicine delivery
The D-H Center for Telehealth has an extensive infrastructure to support clinical initiatives within D-H and provided logistical and technical support for this project. All study staff (health coaches, nurse, registered dietitians) were participated in multiple, on-site training sessions to ensure familiarity with the telehealth platform. Live, mock sessions and ongoing on-site support was provided by the research assistant (RA) and by a technology consultant from the Center for Telehealth.

All communications were conducted through an HIPAA-compliant
Vidyo software platform that includes end-to-end data encryption using HTTPS (browsing), TLS (signalling), and AES encryption.
Coaching sessions were conducted in a private clinical space, using a T450s Lenovo laptop and Logitech H390 USB Headset with a noisecancelling microphone. Participants were provided with a Samsung Galaxy Tab A 10.1 tablet that was encrypted per institutional policies to conduct the intervention off-site (ie, home) with the same software allowing them to interact with the study personnel.

| Recruitment and enrolment
Clinic schedules were initially reviewed by the RA. New patients were approached by the clinician and introduced the study opportunity to assess interest. The RA then further described the study, obtained consent, and scheduled a 1-hour individual orientation for all subjects. if one was not available, the RA assisted in its creation. Participants were excluded if they were unwilling to participate, as well as those with a medical record diagnosis of dementia, life-threatening illness, psychiatric illness (untreated serious mental illness, suicidal ideation) precluding their participation in the study, or a history of bariatric surgery. All participants required medical clearance from their primary care provider and needed to provide voluntary written consent. The lead author (JAB) was responsible for training the RA during this process. All participants received a $20 incentive at each in-person outcome assessment.

| Outcomes
Baseline measures were chosen a priori based on their validity, brevity, use in routine clinical care, and availability in the electronic medical record. The RA obtained baseline demographic information and comorbid health conditions from the EMR. On-site assessments occurred at baseline and at 16 weeks, with additional surveys conducted at 4-week intervals (data not shown). Monthly weights were acquired using an A&D Medical Bluetooth enabled scale and captured using the application. Surveys were sent electronically using REDCap, a secure, web-based application designed to support research data capture. 38 Height was measured using a wall-mounted stadiometer (SECA 216, Hamburg, Germany), with the participant standing barefoot, against a wall, with their weight evenly distributed on both feet and heels together. Three height measurements were taken, and the average was used as the final value. Waist circumference was measured by a registered nurse using a tape measure placed around the abdomen, just above the iliac crest, snug, and not compressing the skin. The participant was asked to relax, and measurements were taken at endexhalation. A bioelectrical impedance analyser (SECA mBCA 514, Hamburg, Germany) was used to assess weight, body fat, muscle mass, and visceral fat. Participants were instructed to remove any outer clothing, jewelry, shoes and socks, or tights and stood on the metal electrodes on the base of the machine, facing forward. A 6-minute walk test was conducted by a nurse according to standard protocols, 39 representing satisfaction with telemedicine delivered interventions.
Additional acceptability questionnaires were asked at the conclusion of the study. An exit-interview at the end of the study was conducted that ascertained the participant's impressions of the overall programme and what they liked/disliked about the programme. These interviews were digitally audio-recorded and transcribed by www.
rev.com, a commercial transcription programme.

| Analysis
All data were aggregated into REDCap. Descriptive statistics (means, standard deviations, medians, proportions, and range) were computed to assess feasibility and acceptability. The analysis focused on completers of the programme. Change in weight, percent weight loss, and waist circumference were our primary preliminary effectiveness outcomes. Paired t tests assessed change between baseline and followup. All qualitative interview data were inputted into Dedoose and analysed by two researchers using thematic data analysis [44][45][46] consisting of "open coding" of transcripts, a process of labelling text to identify concepts related to acceptability. 47 This process enhances rigor by allowing for different views. 48 Codes were determined both a priori and inductively derived. Text excerpts were aggregated by code to distill patterns and themes related to the intervention's acceptability.
The analysis was conducted using STATA v.15, Microsoft Excel 2017, and REDCap's data output for simple quantitative data analyses. While a P value < .05 was considered statistically significant, this pilot study was intended to investigate feasibility and was not powered to detect a statistically significant difference in our outcomes.

| Feasibility of recruitment and retention
Clinicians approached 62 participants seeking treatment at the centre ( Figure 1)

| Intervention adherence
All participants completed all study measures at baseline and followup points while enrolled, exceeding the a priori threshold of 80% considered as successful. The proportion of study participants completing greater than 75% of sessions was favourable among those enrolled (73%) and among those completing the study (100%). Approximately 93%, 96%, and 67% of participants attended greater than 75% of health coach, nurse, and dietitian sessions, respectively.

FIGURE 1
Consort diagram of all participants using telemedicine in a rural, academic, and obesity clinic Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease; CAD, coronary artery disease; NAFLD, nonalcoholic fatty liver disease; OSA, obstructive sleep apnoea.

| Baseline characteristics
Cohort characteristics, both enrolled and completers, are presented in Table 2. There were no significant differences between completers and noncompleters except for insurance status. Mean age among those completing the study was 46.9 ± 11.6 (range 27-64 years), and the proportion of females was high (88.9%). All participants represented themselves as white and not Hispanic. Mean body mass index was 41.3 ± 7.1 kg/m 2 and mean waist circumference was 120.7 ± 16.8 cm. Figure 2 presents data on the acceptability of telemedicine as a delivery modality. All responses were favourable (Table S2)  were delayed and only three (0.7%) were cancelled due to technical issues that included bandwidth issues or that a tablet was not charged.

| Qualitative inquiry on the programme's acceptability
Many themes emerged through our participant end-of-study interviews ( Table 3). The importance of time-savings was observed throughout many of the conversations. Participants were highly FIGURE 2 Select questions asked to participants on the acceptability of the intervention. Each question was rated from strongly disagree/ dissatisfied (1) to strongly agree/satisfied (5). Mean scores are indicated with error bars representing standard deviations positive about video-conferencing rather than commuting for an inperson evaluation. This enhanced control of their time, reduced anxiety and hassles, notably in enabling, and allowed for health consultations to occur within the context of a demanding job. Another theme included the simplicity of the video-conferencing technology.
The information delivery was helpful to all participants, and the programme provided considerable resources to enhance nutritional and behavioural strategies. In contrast, a significant criticism was the lack of peer-support by participants and that a programme wholly based on video-conferencing felt depersonalizing.

| DISCUSSION
An evidence-based weight loss intervention delivered using telemedicine was feasible and acceptable to rural adults with obesity. Importantly, the intervention led not only to weight loss but also to significant changes in visceral fat as measured by bioelectrical impedance with maintenance in appendicular muscle mass and improvements in strength measures. These results suggest that a future intervention using this delivery modality within a clinical setting can potentially overcome many hurdles/barriers to delivering high-quality, intensive obesity care in this rural population.
Many previously published obesity interventions occur within primary care environments 49 or in research centres. 50 Interventions such as the diabetes prevention programme are effective, 51

| CONCLUSION
A multicomponent, telemedicine-based obesity lifestyle programme appears to be feasible and acceptable to patients and is thus a promising approach for weight and visceral fat loss in rural populations. A randomized controlled trial is needed to evaluate this modality for future implementation and effectiveness as part of their routine practice.

ACKNOWLEDGEMENTS
This study was approved by the Committee for the Protection of Human Subjects #30240. All participants consented to participate.
The authors approve publication if accepted. The data that support the findings of this study are available from Dartmouth-Hitchcock, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available.
Data are however available from the authors upon reasonable request and with permission of Dartmouth-Hitchcock.
We thank the Center for Telehealth (Mary Lowry, Vanessa Brown, Fredric Glazer) for their assistance in developing the telemedicine component, and Tara Efstathiou, Laurie Gelb, Eugene Soboleski, Jane Brewer, Martha Catalona, Philip Oman, and Kaitlyn Christian, for their administrative assistance at the Weight and Wellness Center.

CONFLICT OF INTEREST
There are no conflicts of interest pertaining to this manuscript.