Swallowing Therapy With the Assistance of a Mobile Health Device in Head and Neck Cancer Patients: a Pan-Alberta Study

Overview

Dysphagia affects 22% of those over the age of 50, which equates to 250 million people worldwide and 360,000 in Alberta. At high risk are survivors of head and neck cancer (70%). Difficulties with swallowing are not only life threatening and resource-intense, but also socially limiting. To regain swallowing function and avoid or reduce the consequences of dysphagia, patients require regular, intensive therapy over many months to strengthen swallowing muscles and improve swallow coordination. This therapy is often coupled with visual biofeedback that uses surface electromyography (sEMG). Despite evidence that swallowing exercises are effective when provided with an intensive regimen and when coupled with sEMG biofeedback, patients rarely receive it. The primary aim of this work is to determine whether the use of a mobile system equipped with sEMG biofeedback affects adherence to home-based swallowing exercises. The secondary aim of this work is to determine if the exercise program results in improved patient reported outcomes related to dysphagia and nutrition. Our tertiary aim is to determine if previous findings of adherence can be replicated. Sixty adults with oropharyngeal dysphagia secondary to OPSCC treatment will be enrolled in the study. This study will follow a cross over randomized design such that all participants will be provided with both types of treatment: using pen and paper (Treatment Arm A) and using the mobile health system (Treatment Arm B).

Background Dysphagia affects 22% of those over the age of 50 \[1\], which equates to 250 million people worldwide and 360,000 in Alberta \[2\]. At high risk are survivors of head and neck cancer (70%), stroke patients (78%), people with Parkinson's disease (82%) and those who've had a traumatic brain injury (65%) \[3\]. Left untreated, dysphagia has serious health consequences, such as malnutrition, aspiration pneumonia, and even death. The associated longer hospital stays and complications result in an estimated annual burden of care of $547 billion in USA \[4\]. In Alberta, managing just one consequence of dysphagia (aspiration pneumonia) in patients with head and neck cancer and stroke costs the healthcare system $2.53 million CAD a year \[5\]. Difficulties with swallowing are not only life threatening and resource-intense, but also socially limiting. Patients with dysphagia avoid social situations that involve eating, are disproportionately affected by depression, and miss more than 3 times the number of workdays than individuals without swallowing difficulties. To regain swallowing function and avoid or reduce the aforementioned consequences of dysphagia, patients require regular, intensive therapy over many months to strengthen swallowing muscles and improve swallow coordination. This therapy is often coupled with visual biofeedback that uses surface electromyography (sEMG). sEMG biofeedback lets patients know how hard their muscles are contracting and for how long during a given exercise. sEMG biofeedback is used to teach patients how to perform exercises correctly, elicit maximum effort, and demonstrate incremental improvements in performance. Despite evidence that swallowing exercises are effective when provided with an intensive regimen and when coupled with sEMG biofeedback, patients rarely receive it. This limited access stems from poor patient and clinician resources. For patients, getting to the clinic to access a clinician and the equipment is inconvenient and costly. In order to cope with limited patient engagement, clinicians are more likely to manage dysphagia by providing diet modifications, by prescribing feeding tubes, and by giving patients maneuvers and exercises on a printed sheet of paper. These approaches to the management of dysphagia are inadequate to meet the needs of patients. Furthermore, a recent systematic review found that average adherence to dysphagia management recommendations was between 22% and 52% \[6\]. With the advent of strict distancing restrictions during the COVID-19 pandemic, remote care via telerehabilitation or mobile health (mHealth) technologies has received increased attention. AIMS: The primary aim of this work is to determine whether the use of a mobile system equipped with sEMG biofeedback affects adherence to home-based swallowing exercises. The secondary aim of this work is to determine if the exercise program results in improved patient reported outcomes related to dysphagia and nutrition. Our tertiary aim is to determine if previous findings of adherence can be replicated. Mobile swallowing system selected for study The Mobili-T is a mobile swallowing exercise system for patients with swallowing difficulties. This mHealth system can be used by patients for home-based swallowing exercises. It is comprised of a mobile device, a smartphone patient app, and a clinician portal. The patient wears the mobile device during exercise. The mobile device uses clinical grade surface electromyography (sEMG) and sticks under the chin with a medical-grade double-sided adhesive. The sEMG hardware comes with a charging dock that doubles as a carrying case. The sEMG signal is wirelessly transmitted to the smartphone app. Here, the sEMG is shown to the patient as visual biofeedback, where the deflection of the biofeedback signal reflects the strength and duration of muscle contraction. The muscles monitored are found under the chin (submental area) and are those targeted during typical swallowing rehabilitation. The app uses a smart software that walks patients through a set of clinically proven swallowing exercises \[7,8\], uses a back-end algorithm to detect whether or not the sEMG signal collected was elicited by a swallow (as opposed to other extraneous movements such as head movements), and calculates effort targets based on the patient's abilities. Clinicians can remotely monitor patient progress, adherence, and adjust prescriptions. One of the unique aspects of Mobili-T is that the investigators have involved patients and clinicians in the design from the inception of the project. The investigators also have conducted usability and feasibility research testing \[9-13\]. Our feasibility study was a quasi-experimental pretest-posttest trial with 20 patients with dysphagia secondary to head and neck cancer (submitted for publication). The investigators evaluated whether or not patients could successfully complete swallowing therapy using the Mobili-T at home under the remote clinical supervision of our speech language pathologist. The investigators found that adherence rates to home-based swallowing exercise remained at or above 72% over a six week treatment period and that dysphagia-specific quality-of-life improved following this six week treatment program. However, the study had limitations that precluded us from understanding whether the observations noted were a result of the use of the technology or a result of participant attitude and motivation in being part of a study. Our aim with the present study is to determine if our previous findings can be replicated, and also to identify if these findings are unique to the use of an mHealth technology. We are also interested in whether patient-reported nutrition and health change throughout the treatment block. To address these objectives, we will have two conditions within the study: 1) home-based dysphagia exercise using the current gold standard of printed instructions and a diary to track adherence, and 2) home-based dysphagia exercises using an mHealth system to provide biofeedback and track exercise completion. To ensure that all patient participants have equitable access to both conditions, a cross over randomized design will be used such that all participants will be provided with both types of treatment-delivery methods. This study design was previously used by Mertens et al. (2016) to determine if a mobile application improves therapy adherence rates in elderly patients undergoing rehabilitation \[14\]. The specific objectives of our study are as follows: Objective 1 (i.e., primary objective). Determine if adherence to home-based swallowing exercises is different when a mobile system equipped with sEMG biofeedback is used versus pen and paper (diary format). Hypothesis 1. There is no difference in average weekly adherence between the experimental group (mHealth system) and the control group (pen and paper). Hypothesis 2. There is no difference in average daily exercise dose (total # of exercise trials completed in the treatment block/ number of days in the treatment block) between the experimental group (mHealth system) and the control group (pen and paper). Objective 2. Determine if the exercise program results in improved patient reported outcomes related to dysphagia, nutrition, and health. Hypothesis 3. There is no difference in patient-reported outcomes on the MD Anderson Dysphagia Inventory (MDADI), caloric intake, and health between the experimental group (mHealth system) and the control group (pen and paper). Objective 3: Determine if previous findings related to adherence and patient-reported swallowing outcomes can be replicated. Hypothesis 4. Average weekly adherence rates to home-based swallowing exercise will remain at or above 72% over a six week treatment period. Twenty-five percent of participants will experience a minimally important differences in their pre- and post-therapy dysphagia-specific quality-of-life scores. General study details for ITA A total of 60 patient participants will be recruited, 30 per study arm. The study will last 3 months and one week (or 13 weeks). However, the study will remain open until all participants have been recruited and have completed the trial (expected to take 1 year). Sample size calculation is detailed below, in section 1.3. Participants 1.1 Recruitment. Participants will be recruited through speech-language pathologists (SLPs) at the University of Alberta Hospital, Cross Cancer Institute, Institute for Reconstructive Sciences in Medicine, Home Nutrition Support Program, Foothills Medical Center, Tom Baker Cancer Center and Central Alberta Cancer Center. Participants will be identified by their SLPs. Participants also will be recruited from a pool of individuals who have contacted the research team during the feasibility trial or following media events, wishing to be notified about upcoming trials. Participants will be consecutively recruited until the sample size is achieved. Recruitment An email alert (found in Documents) will be sent to physicians and speech-language pathologist (SLPs) in the province known to work with head and neck cancer patients. They will be told which patients are eligible for this study and will be provided information about the study. SLPs will also be provided with the study poster and study flyer so that they can share these with interested patients. Hence, participants will be identified by their physicians or speech-language pathologist (SLP) and recommended for swallowing therapy to the research group. Patients who are referred to the study by their SLP and are interested in participating will be asked by their referring clinician to complete the Consent to Recruit/ Be Contacted form (found in Documents). This form will be passed on to the Research Clinician by the referring clinician. Consent To Be Contacted by Researcher Form has an expiry date listed on the form. Participants also will be recruited from a pool of individuals who have contacted the research team during the feasibility trial, through word of mouth, or following media events, wishing to be notified about upcoming trials. They will be alerted of the study via an email (found in Documents) and the study recruitment poster. The study flyer will be shared only with anyone who is interested. If these patients express an interest in participating, they will still require a referral from their own attending SLPs (i.e., community clinicians). Participants will be enrolled in the study if they have a diagnosis of oropharyngeal dysphagia secondary to HNC treatment. The patient's present or past SLP will be used as a source of referral and validation that the patient is a good candidate for home based swallowing exercise using the effortful and/or the Mendelsohn swallows. For patients unknown to the investigators first, they will NOT be contacted directly until patients have provided consent to do so. Patients will be approached by their attending clinicians (physicians or SLPs). Clinicians will let patients know that a study is currently underway. They will share the title of the study, study poster, and study flyer with the patient and ask them if it is okay for a Research Clinician (Dr. Gabriela Constantinescu) to contact them with more information. The Research Clinician will contact the patient and ask them if they are still interested in participating. If yes, she will share the letter of information and consent form electronically (either emailed to the patient or to their SLP). A second call will be scheduled and the Research Clinician will then go over the letter of information with the patient in detail. If the patient is still interested in the study, 3 appointments in Edmonton will be booked. Consent will be obtained in person, at the first visit. If the patient affirms on the phone however, that they are no longer interested in the study, the patient's contact information will be shredded. Occasionally, patients may request to be contacted at a later date due to personal reasons. In this scenario, contact information will be kept in a secure location until the sample size for the study is reached. At that time, if the patient has not been contacted/ booked, their contact information will be shredded. None of the study coordinators /investigators/ research clinician are directly involved in the patients' care. Patients may simply tell their attending clinicians that they are not interested in hearing further details about the study and the Research Clinician will not be aware of this information. Patients will be reminded by the Research Clinician that participation is voluntary and that care will not be affected in any way by their decision. Any callback list is not publicly displayed. It will be kept in the study binder (in a locked room at the University of Alberta) until shredded as detailed above. The Research Clinician, Dr. Gabriela Constantinescu, will look after contact information of patients expressing interest, until shredded as detailed above. 1.2 Inclusion/ Exclusion Criteria. Participants will be enrolled in the study if they have a diagnosis of oropharyngeal dysphagia secondary to treatment for squamous cell carcinoma (OPSCC). SCCs make up 90% of all HNCs. Participants will be consecutively recruited, until the sample size is achieved. Participants will be recruited any time 3 months or later post-surgery or post-(chemo)radiation therapy. This time frame was selected to avoid duplication of services for patients who may receive swallowing therapy during their radiation therapy and to target patients with chronic and stable dysphagia. Participants will be included: 1) if they have received treatment for head and neck cancer (e.g., +/-surgery, +/-radiation therapy, +/- chemotherapy), 2) if their attending speech-language pathologist has confirmed from a Modified Barium Swallow (MBS) assessment or Fiberoptic Endoscopic Evaluation of Swallowing (FEES) that the patient is a candidate for the effortful swallow and/or the Mendelsohn maneuver therapy. These exercises are used routinely in clinical practice and target specific physiological impairments of the swallow that are common to OPSCC patients. Participants will be excluded: 1) if they have a history of cognitive delay, 2) a history of stroke or traumatic brain injury, and 3) if they cannot reliably navigate the Mobili-T system after the training session, and 4) if their SLP confirms planned swallowing exercises or procedures to address dysphagia like esophageal dilation. Participants also will be excluded if they have a beard that they are not willing to shave or partially shave (as the device adheres under the chin, on the surface of the skin). Finally, participants will be excluded if they are unable to travel to Edmonton 3 times during a 3 month period). In addition, in keeping with contraindications listed on the Mobili-T labeling, patient participants will be excluded if they: have an implanted electronic device of any kind, including cardiac pacemakers or similar assistive devices, electronic infusion pumps, and implanted stimulators; if they have irritated skin or skin with open wounds under the chin; if they have an allergy to silver. 1.3 Sample Size. No previous literature exists on a clinically significant difference in adherence or dose to guide sample size calculation. The researchers used a previous study \[15\] to guide our sample size selection. In the cited study, a total of 79 OPSCC patients were randomized to 3 treatment arms (26 or 27 patients participants per arm), where one of the treatment arms included swallowing therapy using an app. As the present study has 2 treatment arms, we will recruit 52 to 60 participants. In addition, we have consulted an epidemiologist regarding sample size calculation. The following calculator was used: http://hedwig.mgh.harvard.edu/sample\_size/js/js\_crossover\_quant.html along with the following parameters: 0.05 significance level, two-sided, within patient standard deviation of 35.5 \[6\], power of 0.9, and a minimal detectable difference in means of 22. This sample size calculation is based on the primary objective. 2.0 General Procedure: Sixty adults with oropharyngeal dysphagia secondary to OPSCC treatment will be enrolled in the study. This study will follow a crossover randomized design (AB/BA) such that all participants will be provided with both types of treatment: using the mobile health system and using pen and paper. Randomized allocation will be determined using Sealed Envelope, a Randomization and online database for clinical trials (https://www.sealedenvelope.com/) Once a patient expresses interest in the study/ is referred to the study by their clinician (e.g., SLP), the Research Clinician, Dr. Gabi Constantinescu, will contact the participants with more information and screening questions (e.g., do you have a beard?). An electronic or hard copy version of the Letter of Information and Consent form will be shared with the patient. The information and questions are meant to ensure that patients interested in the study fit most exclusion/ inclusion criteria before traveling to Edmonton. If patient participants self-refer to the study via word of mouth (and were not referred by their attending SLPs), consent to contact the patient's attending SLP will be obtained before participants are invited to Edmonton for their first appointment. The patient's attending SLP will be contacted to ensure that services are not duplicated elsewhere, that the patient is not booked for other procedures that may impact outcomes in this trial (e.g., esophageal dilation), that a swallow assessment was recently completed and that the SLP believes the patient will be a good candidate for the swallowing therapy offered through this study. All pre-, washout, and post-treatment sessions will take place in Edmonton. 2.1 Pre-treatment session. The Letter of Information and Consent will be reviewed again. The MD Anderson Dysphagia Inventory (MDADI) and the Health Questionnaire (EuroQoL EQ-5D) (both described in section 3.0 Outcome Measures) will be administered. Next, Dr. Constantinescu will: (1) determine the order of treatment blocks for the participant (i.e., Group1 - AB, or Group 2 - BA); (2) introduce the Mobili-T system (if applicable); (3) introduce the swallowing exercises and allow time for practice; (4) explain the treatment block format and associated documents. The swallowing therapy exercises will be explained and practiced until patients can demonstrate competency on their own. This ability is defined as the successful completion of five consecutive trials of each exercise, where the success will be judged by clinical visual inspection. Participants will be sent home with the Three-Day Dietary Intake Record to complete and return at their next appointment. If patient participants are randomized to the Treatment Arm B with the mHealth system first, then training with the system will occur in the presence of the Research Clinician and will conclude once the patient appears comfortable with the device, has shown that he/ she can set up the device and start the software independently, and has demonstrated competency with the exercises. The Research Clinician in this study (i.e., registered SLP) will remotely monitor the session data from participants who are using the mHealth system and will be available to answer questions and troubleshoot technology issues. Patients will not be identified on the device, app, or clinician portal. The app simply records the serial number on the device hardware (e.g., 001). This is noted in the researcher's records (e.g., J Smith - 001). Participants will further be identified by a participant code. 2.2 Treatment blocks A and B. Patients will be asked to practice at home for 3 months targeting a daily total of 8 sets of 3 exercise types with 3 repetitions of each. The daily total target will be 72 trials (8 sets \* 3 exercise types \* 3 repetitions of each). Exercise type and dose will be the same for all participants, for a daily total of 24 regular saliva swallows, 24 effortful saliva swallows, and 24 Mendelsohn maneuver saliva swallows. Treatment Arm A will consist of using pen and paper to complete the exercises. This is considered standard care. Treatment Arm B will consist of using Mobili-T to complete the exercises. 2.3 Post-treatment session. Following the first Treatment Arm (A or B), there will be a one-week washout period. During this week, as well as following the second Treatment Arm, a post-treatment session will be booked and the Three-Day Dietary Intake Record provided for home completion. This session will consist of administering the MDADI, the Health Questionnaire (EuroQoL EQ-5D), and either returning or providing the Mobili-T system.