Project 4: Ambulatory Biofeedback and Voice Therapy for Patients With Vocal Hyperfunction

Preliminary Studies to Test the Effects of Ambulatory Biofeedback in Small Groups of Patients With Vocal Hyperfunction: Study 3

Vocal hyperfunction (VH) is the most commonly treated class of voice disorders by speech-language pathologists and voice therapy is the primary curative treatment. Patients and clinicians report that generalizing improved voicing into daily life is the most significant barrier to successful therapy. We will test if extending biofeedback into the patient's daily life using ambulatory voice monitoring will significantly improve generalization during therapy and if individual patient factors, like how easily they can modify their voice and engagement during therapy, moderate the effects of the biofeedback.

Vocal hyperfunction (VH) is ostensibly caused by and/or associated with pathological daily voice use and involves the most commonly treated voice disorders by speech-language pathologists, e.g., vocal fold nodules, muscle tension dysphonia. Voice therapy is the primary curative option for VH. For example, even when patients undergo laryngeal surgery to remove lesions, they are still thought to be at risk for recurrence unless they successfully complete post-surgical voice therapy. However, voice therapy suffers from high rates of patient dropout. Patients and clinicians report that generalizing desired vocal behaviors from the therapy session into daily life is one of the most significant barriers to successful voice therapy. Despite this critical barrier, voice therapy remains entirely dependent upon episodic delivery within an in-clinic or virtual session. Thus, this project will test if adding Ambulatory Voice Monitoring with Biofeedback (AVM-B) significantly addresses this generalization challenge, as it can directly extend therapeutic activities into the patient's daily life. A clinical trial will randomize patients with VH to receive an evidence-based therapy (Conversation Training Therapy; CTT) or CTT with AVM-B added. In Aim 1, it is hypothesized that, compared to patients who only received CTT, patients who receive CTT and AVM-B will demonstrate significantly better generalization during therapy which will be retained immediately after therapy and six months later. In Aim 2, we will explore patient factors that mediate the relationship between therapy and generalization, hypothesizing that stimulabilty-how easily a patient can modify their voice-and engagement-the patient's level of effort during therapy-will be positively correlated to the amount of generalization in daily life. If successful, this work would result in multiple paradigm-shifting impacts with potential to improve the efficiency of clinical practice. AVM-B would become one of the first evidence-based voice treatment activities taking place primarily outside the therapy session. Future work could investigate how AVM-B could transition voice therapy from once-a-week sessions into a continuous process integrated into the patient's daily life. Further inquiry could improve generalization by identifying evidence-based methods to tailor therapy based on individual patient factors such as stimulability and engagement. After discharge, AVM-B could provide a means for patients to "recalibrate" themselves and prevent relapse without having to see a clinician. Finally, implementation work could help clinicians adopt/adapt AVM-B and evaluate its effects on dropouts.

Conversation Training Therapy will be administered one time per week for 1 hour. Ambulatory Voice Monitoring with Biofeedback will be administered during the second therapy session and for 5 days between the second and third therapy sessions.

AVM-B is software on the Voice Health Monitor that uses a neck-placed accelerometer to sense neck skin vibrations during voicing in daily life. The AVM-B consists of cues every time the patient exceeds a subject-specific threshold, [100% frequency AVM-B] and/or summary statistics every 2 minutes of voicing [Summary AVM-B].

CTT is a 4-week (1 45-minute session per week) evidence-based voice therapy for patients with VH, developed primarily to maximize the amount and speed of generalization outside the therapy session. It consists of 3 required targets: Decreased overall auditory-perceptual voice severity, improved discrimination between baseline and improved voicing, and adherence to CTT recommendations in daily life. The three optional targets are decreased overall auditory-perceptual severity across an increased pitch range, loudness range, and decreased rate of speech. All voice-related targets are modified through practicing voicing with increased forward resonance and mean airflow in spontaneous speech/conversation.

Generalization will be represented as a "percent accuracy" where voiced frames of a feature that are above or below a percentile threshold (based on baseline/pre-therapy monitoring data) are divided by the total number of voiced frames.

Stimulability will be represented as a "percent accuracy" where voiced frames of a feature that are above or below a percentile threshold (based on baseline/pre-therapy monitoring data) are divided by the total number of voiced frames.

The RITS scale is one scale with 7 levels where the 1 represents the "absence of effort" and 7 represents "superior effort." The patient's treating clinician will judge his/her engagement during therapy after each session. These judgements will provide valid and reliable estimates of patient engagement.

The CAPE-V consists of 4 perceptually judged 100-mm visual analog scales (Overall auditory perceptual severity, Breathiness, Strain, Roughness). The patient's treating clinician will judge standard audio samples. These judgements provide gold-standard perceptual ratings of voice quality to evaluate if therapy and AVM-B are associated with improved voice quality.

The V-RQOL consists of 10 Likert-style questions ranging from 1-5 (1 = no problem at all; 2 = a small amount; 3 = a moderate problem; 4 = a lot; 5 = problem is as bas as it can be) that estimates how the subject's vocal function effects his/her daily life. This patient reported outcome measure will be used to evaluate if the patients perceive that therapy helped them function better in their activities of daily living.

Inclusion criteria for PVH: Diagnosis of vocal fold nodules and polyps. Secondary diagnosis of laryngopharyngeal reflux (LPR) or gastro-esophageal reflux disease (GERD) Secondary diagnoses commonly associated with phonotrauma like erythema, edema, varies, ectasia, laryngitis, secondary/reactive muscle tension dysphonia (MTD), hemorrhage, etc. Exclusion criteria for PVH: Any secondary diagnosis not directly related to phonotrauma, like cyst, sulk, cancer, bamboo nodule, known or suspected paralysis, etc. Inclusion criteria for NPVH: Diagnosis of primary MTD. Allowable secondary diagnoses are LPR and GERD. Exclusion criteria for NPVH: Any secondary diagnosis related to pathological structure (e.g., nodules, polyps, edema, cancer), neurology (e.g., paralysis, Parkinson's Disease), or respiration (e.g., chronic cough, paradoxical vocal fold motion).

De-identified data to be shared will include stimulability before therapy, average generalization each week of therapy, and patient-reported and clinician-rated outcomes before, after, and 6 months after voice therapy. These are the data that will be used in the analyses and the basis for manuscripts and results. Thus, they are the data necessary for the scientific community to replicate findings or integrate results into meta-analyses and data synthesis.

Data will become available concurrent with publications where applicable and after the study finishes in 2029.

To maximize the appropriate sharing of scientific data and protect research participants' privacy and confidentiality, reuse of this dataset should use the following Data Use Limitations (DULs) under Controlled Access that is made available by a data repository only after approval of the request by the Harvard Dataverse independent review panel process; including Health/Medical/Biomedical controls and IRB Approval Required. To protect research participants' privacy and confidentiality, data submitted to the repository will not include personally identifiable information such as names, addresses, etc. Additional protections, such as the approach for managing Health Insurance Portability and Accountability Act identified, will be used for de-identification or to provide a limited data set to minimize the risk of participant reidentification.