We conducted a 4-week, double-blind, sham-controlled, transcutaneous auricular vagus nerve stimulation (taVNS) trial exploring the effects of at-home, self-administered taVNS to manage long COVID symptoms (Fig. 1). This study was designed as a two-part study, part one being a randomized, double-blind, controlled trial lasting two weeks, followed by part two in which all participants received two weeks of active taVNS in an open-label design. All taVNS was self-administered by participants in their home or office space, twice daily, six days a week, for the duration of the four-week trial. This study was approved by the MUSC Institutional Review Board (IRB) and is registered on ClinicalTrials.org (NCT04638673). All participants signed written informed consent.
Participants and Inclusion Criteria
We consented and 13 individuals (8 women) who previously tested positive for COVID-19 into this remote-monitored at-home trial. Because this study was conducted during the COVID-19 pandemic from November 2020 to August 2021, no procedures were conducted in person. One participant was unable to provide a laboratory confirmed COVID positive test, and was not randomized into the study, thus we randomized and treated 12 total participants. All study procedures conducted by participants (consent, outcome measures, stimulation, and vital monitoring) were completed via virtual methods (e-mail, telephone, HIPAA-compliant video conferencing), and the taVNS devices and monitoring equipment were shipped to their home.
The inclusion criteria were as follow: over the age of 18; laboratory-confirmed COVID-19 infection; afebrile; new onset of at least one of the following nine neuropsychiatric symptoms associated with long COVID (anxiety, depression, vertigo, anosmia, ageusia, headaches, fatigue, irritability, brain fog); no damage to either left or right ear anatomy; no unstable hemodynamic effects; no ischemic or haemorrhagic stroke after developing COVID-19; ability to provide informed consent, follow instructions, read, write and speak English; attest to having reliable access to Wi-Fi internet at home.
Mobile Neurostimulation Briefcase Containing All Study-Related Equipment
To safely and effectively deliver taVNS at-home, in a manner that is user-friendly and portable, the study team designed a hard-shell briefcase containing all study related equipment. This included stimulation equipment, real time physiology monitoring equipment, and telecommunications devices required for remote monitoring (Fig. 2a).
A detailed list of the contents of the briefcase are as follow: one taVNS stimulator (Soterix Medical, Inc), 100 + pre-packaged taVNS electrodes (Neotech Products, LLC), 100 + alcohol prep pads (70% ETOH), six rechargeable lithium ion AAA batteries and charger, one vital monitoring tablet (Caretaker, LLC), vital monitoring device (Caretaker, LLC)(30), three vital monitoring finger cuffs (Caretaker, LLC), pulse oximeter (Caretaker, LLC), one iPad for telehealth visits (Apple, Inc), embedded surge protector, one exercise arm band to facilitate portability of taVNS, one enlarged reference photo demonstrating correct taVNS electrode placement.
Initial Orientation and Virtual Training
After completing the virtual consent and baseline visit utilizing HIPAA-compliant telemedicine software (Doxy.me, Inc.), subjects were informed that they would be shipped a neurostimulation briefcase and simultaneously e-mailed three training videos that they were instructed to watch prior to a planned future orientation session. These videos covered overall guidance on how to use the mobile neurostimulation briefcase, including: 1) welcome and unboxing of briefcase, 2) how to use the vital monitoring device, and 3) how to use the taVNS device.
After receiving the briefcase, participants were instructed to charge all internal electronic components. Subsequently, participants then attended a virtual orientation session with a study team member during which they were trained on how to use the systems provided in the briefcase. During this orientation, the study team covered three main overarching topics: 1) the purpose of each piece of equipment; 2) how to operate the vital monitoring device; 3) how to place the taVNS electrodes, turn on the stimulator and input the code to activate the stimulation device for a treatment.
Upon successful completion of this orientation visit, the study team scheduled the first treatment visit. Participants were instructed that they will be virtually monitored for compliance and safety during at least their first three days (six total taVNS sessions). Furthermore, participants were given instructions on how to contact the study team during normal business hours for any ongoing technical support.
Self-Administration Procedures for taVNS
All taVNS was administered using adhesive hydrogel electrodes (Neotech Products, LLC) attached to the left ear of participants (Fig. 2B). First, participants were instructed to clean their ear using provided alcohol wipes (70% ETOH). Next, participants peeled the plastic lining off the electrodes and using either a mirror or digital screen, applied the electrodes to the two intended left ear targets (cymba conchae and tragus). To support proper placement, participants were instructed to refer to the enlarged reference photo demonstrating correct placement. This placement was informed based on prior trials and current flow modelling conducted by our group which demonstrated a highest likelihood of stimulation of the underlying auricular branch of the vagus nerve (ABVN)(31). After electrodes were placed, participants connected the electrode wire to their taVNS system and were ready to self-administer stimulation (Anode- Cymba Conchae, Cathode – Tragus). The stimulation systems were double-blind, and required a code to initiate stimulation, which was sent to participants via email the morning of each taVNS treatment day. Once the participant received the code, they entered it into the system to receive their appropriate stimulation session (active or sham).
taVNS Parameter Settings and Blinding
All taVNS sessions utilized the following parameters: 25Hz, 500us pulse width, tonically on for 1 hour, twice per day, 6 days per week. Current intensity was set to 2X individual perceptual threshold (PT), self-determined via methods described in our prior work(32). Perceptual threshold was conducted once, and the current intensity was pre-set for the entire study. After the PT was established, the research staff recorded which blinded group the participants were assigned to (A or B) and issued the participant group-specific stimulation codes. Neither the staff nor the participants knew that group A codes were programmed to deliver sham stimulation (0mA of stimulation), whereas group B codes were programmed to deliver active stimulation (2X PT mA). The device screen indicated stimulation was being administered with a real-time 60-minute countdown regardless of group. This code system enabled the participant and research staff to remain blinded throughout the experiment. Furthermore, codes being required to activate the device enabled compliance monitoring and regulated the number of stimulation sessions a per day.
Safety monitoring Using a Telemedicine Approach
We monitored the safety of taVNS by measuring blood pressure and heart rate during the first six 1-hour taVNS sessions for each participant. Safety monitoring was conducted in real-time using a dedicated tablet connected to WIFI, which communicated with a wrist-worn pneumatic device that measured heart rate and blood pressure in real time using a finger cuff attached to the middle finger of the nondominant hand of the participant (Caretaker Medical). The research staff was able to remotely connect and monitor the vitals using a web browser. Monitoring was performed for the duration of each monitored session. The study staff-maintained visual and phone communication with the participant to understand qualitative metrics of comfort, tolerability, and adverse events. All quantitative heart rate data was collected and stored for analysis of bradycardia events.
Symptom Improvement Tracking and Outcomes
We created a battery of outcomes to track heterogenous symptom improvements. We primarily focused on mood and anxiety effects, however due to the broad range of long-COVID impacts on the body and brain, we also included several other measures of fatigue, smell, and cognition. Our battery included the following assessments: General Anxiety Disorder-7 (GAD7), Clinical Global Impression Improvement (CGI-I), Clinical Global Impression Severity (CGI-S), Global Assessment of Functioning (GAF), COVID questionnaire, Connor Davidson Resilience Scale (CDRS), abbreviated PTSD Checklist (PCL6), Perceived Stress Scale (PSS10), Patient Health Questionnaires (PHQ4, PHQ9), pain scale, mental fatigue scale, fatigue severity scale, modified fatigue impact scale, subjective smell assessment, and the Montreal Cognitive Assessment (MOCA). Participants completed all assessments, which were virtually administered by research staff remotely via provided tablet.
We enrolled and consented 13 individuals, however one participant was dropped out of the study before being randomized to receive any intervention due to an inability to provide a positive COVID test. Thus, 12 individuals were randomized, and all participants completed the study without dropout.
This study was designed for primary outcomes of feasibility and safety, with secondary outcomes exploring symptom improvement in nine pre-identified Long COVID symptoms. Feasibility measures were assessed by objectively rating self-administration procedures during remote-monitored pre-stimulation orientation visit, as well as the first six treatment sessions. Research staff indicated whether a participant required assistance during self-administration and scored the assistance needed for each participant for both the physiology monitoring and taVNS on a scale of 0 = no help needed, 1 = minor help needed 2 = major help needed.
Tolerability and safety outcomes were quantified by analysing participant real-time heart rate (hr) during stimulation in their first six sessions receiving at-home stimulation, regardless of randomization group assignment. For clinical measures, individual symptom improvement trends are described in a raw, individual comparison format to show trends for future trials to build upon. All statistical analysis were conducted in performed using GraphPad Prism version 8.0.0 for Windows, GraphPad Software, San Diego, California USA.
Of randomized participants, there were no dropouts, and the blind was maintained until the final participant completed all study procedures, after which the entire study team was unblinded to the conditions.