Going Remote: A Revised Study Protocol for a Pilot Randomized Controlled Trial for Biofeedback Treatment of Anxiety Associated With Chronic Spinal Cord Injury

Abstract

Background

The incidence of anxiety in adults with spinal cord injury/disorder (SCI/D) exceeds that of the general population. Heart rate variability (HRV) biofeedback training is a potential treatment associated with a reduction in stress and anxiety, however HRV training has not been explored in the SCI/D population.

Objectives

To describe a modified protocol piloting HRV training to reduce anxiety associated with SCI/D and detail the COVID-19–related modifications.

Methods

To test the feasibility of the biofeedback treatment, 30 adults with SCI/D will complete this pilot randomized controlled trial. Enrollment started in January 2020, halted in March 2020 due to the COVID-19 pandemic, and resumed in March 2021 with a modified protocol. Protocol modifications are documented using the Framework for Reporting Adaptations and Modifications (FRAME). Participants are allocated to the treatment or control arm and undergo eight sessions of physiological monitoring at home using a commercially available HRV sensor and mobile application, which also delivers biofeedback training for those in the treatment arm. Surveys are administered following each session to capture self-reported stress, anxiety, and mood. The study is approved by the HCA-HealthONE institutional review board and is registered with clinicaltrials.gov (NCT# 03975075).

Conclusion

COVID-19 has changed the research landscape, forcing scientists to rethink their study designs to address patient and staff safety in this new context. Our modified protocol accomplished this by moving the treatment setting and delivery out of the clinic and into the home. In doing so, we address patient and staff safety, increase external validity, and reduce participant burden.

Background

The incidence of anxiety in adults with spinal cord injury/disorder (SCI/D) exceeds that of the general population^1–3 and longitudinally has been found to remain elevated over the first 2 years post injury.¹ A possible treatment approach to anxiety for individuals with tetraplegia is self-regulation via biofeedback. One potentially feasible method of biofeedback for anxiety reduction is heart rate variability (HRV) training. HRV reflects the capacity of the heart to adapt to changing circumstances and is understood to be a reflection of overall cardiovascular health, having been viewed as a marker for disease and adaptability.⁴ HRV has also been shown as an indicator of psychological resilience and behavioral flexibility, reflecting the body’s capacity to effectively adapt to changes in social or environmental demands.⁵ Furthermore, it has been suggested that the phase relationship between heart rate, respiration, and blood pressure results in the most efficient gas exchange and oxygen saturation, yielding overall positive health benefits.⁶ Because HRV is mediated by the vagus nerve (cranial nerve X), which has its nuclei of origin located in the medulla, it is believed to be relatively unaffected by injury to the spinal cord and may represent a viable modality for treatment of anxiety in individuals affected by SCI/D. In HRV training, individuals learn how to effectively alter HRV levels using a technique called resonance frequency (RF) training to attain coherence between HRV and respiration rate/amplitude. HRV biofeedback training is associated with a reduction in self-reported stress, anxiety, depressive symptoms, and improved psychological well-being and sleep quality.^7,8 HRV training is a noninvasive, minimal risk treatment approach and represents a promising direction for developing a physiologically based treatment complementary to traditional cognitive behavioral therapy (CBT); as such, its feasibility and utility warrant exploration in the SCI/D population.

Accordingly, the Craig H. Neilsen Foundation funded a pilot randomized controlled trial (RCT) to investigate biofeedback treatment of anxiety associated with chronic SCI/D. Enrollment for the study began in January 2020, but the trial was temporarily halted in March 2020 due to COVID-19, a disease caused by the highly transmissible SARS-Co-V2 respiratory virus. What follows is a description of the study protocol and the modifications that were made in response to COVID-19–related restrictions so that research activities could safely resume.

Objectives

The objective of this article is to describe a protocol that was modified in response to the COVID-19 pandemic: a pilot RCT for biofeedback treatment of anxiety associated with chronic SCI/D. The premise of the pilot study is particularly relevant in the context of COVID-19 given that adults with SCI/D, who were already at a higher risk of experiencing anxiety,^1–3 are likely experiencing increased anxiety as a result of the pandemic.^9,10 We found it critically important to move forward with this research and modify the existing protocol in a manner that could safely be implemented in the homes of adults living with SCI/D. The specific aims, hypotheses, and data analysis plan of the modified protocol (version 9.0 February 2021) are summarized in Figure 1.

Figure 1.

Modified aims, hypotheses, and data analysis.

Methods

COVID-19–related modifications

For transparent protocol reporting, we followed the CONSERVE-SPIRIT Reporting Guidelines for Trial Protocols Modified due to the COVID-19 Pandemic and Other Extenuating Circumstances.¹¹ We also used the Framework for Reporting Adaptations and Modifications-Expanded (FRAME)¹² to document the COVID-19–related protocol modifications (see Figure 2) that were implemented in March 2021 following a 12-month pause on study-related activities. The intervention is a 4-week (eight session) RF training protocol comprising self-regulated breathing techniques, which took place in person prior to the COVID-19 pandemic. Given that the SARS-Co-V2 respiratory virus is transmitted via air droplets that are produced during respiration, it was critical for the safety of the participants and for the research staff to adjust the protocol in a manner that would require the least number of in-person visits. The modified protocol requires one in-person initial assessment (as described later), with the remaining study activities completed in the participant’s home. What was originally an intervention delivered in a controlled setting using medical grade physiological monitoring equipment is now an intervention delivered at home using a commercially available product designed for consumer use. Given this transition, we also modified the metrics by which physiological improvement is defined. Details of these changes are reported in Figure 2. Documents related to the original study design can be found in eAppendix A.

Figure 2.

FRAME protocol modifications.

Trial design

This study is a pilot RCT currently being conducted at a neurological rehabilitation facility in the western United States. Enrollment started in January 2020 and is expected to end in April 2023. Figure 3 summarizes the design of the modified study and the inclusion and exclusion criteria.

Figure 3.

Modified study design and inclusion/exclusion criteria. DASS-21 = Depression Anxiety Stress Scale 21; HRV = heart rate variability; STAI = State Trait Anxiety Inventory; SUDS = Subjective Units of Distress Scale.

Study procedures

Up to 40 individuals with SCI/D and comorbid anxiety are currently being recruited for participation to obtain complete data for a total of 30 participants. To minimize in-person contact due to COVID-19 restrictions and health and safety concerns and/or to reduce participant burden, all study activities take place remotely with the exception of a single in-person initial assessment. To this end, the modified protocol utilizes Health Insurance Portability and Accountability Act (HIPAA)–compliant technologies such as the web-conferencing platform, Zoom for Healthcare, DocuSign for obtaining eSignatures, and a commercially available HRV sensor and smartphone application (app) to facilitate implementation of the intervention at home.

After a phone screening conducted by the study coordinator, those who meet the inclusion criteria (see Figure 3) are scheduled for a virtual meeting during which a trained consenter reviews all essential elements of the institutional review board (IRB)–approved informed consent form and responds to any questions. Those with the cognitive capacity to provide informed consent and who agree to participate in the study electronically sign the consent document using DocuSign. Following enrollment, the study coordinator completes baseline data collection with the participant via Zoom. A schedule of study measures can be found in Table 1.

Table 1.

Schedule of study measures

Measures	Baseline	Initial assessment	Pre sessions 1–7	During sessions 1–7	Post sessions 1–7	Pre session 8	During session 8	Post session 8
Demographics questionnaire	x
Medication list	x
Medication use			x			x
Symptom list			x		x	x
Physiological monitoring		x		x			x
State Trait Anxiety Inventory (STAI)					x			x
Depression Anxiety Stress Scale-21 (DASS-21)	x							x
Subjective Units of Distress Scale (SUDS)	x				x			x
Brief Pain Inventory (BPI): four pain severity items	x							x
BPI: current pain severity item pre/post each session			x		x	x
BPI: pain relief item (post each session)					x			x
Impact of study participation								x

Allocation

Following baseline data collection, participants are allocated into one of two study arms via the minimization method: the treatment arm or the control arm. Minimization is an allocation method used in clinical trials to balance the arms simultaneously over several prognostic factors (strata).^13,14 In the minimization method, the first participant is randomly allocated into an arm of the study, and subsequent participants are allocated to the arm that reduces prognostic factor imbalances among the arms. When there are several empty arms in which to allocate, the arm will be chosen by a random drawing method. For this pilot, minimization balances study arms based on baseline anxiety level defined as high anxiety (Depression Anxiety Stress Scale-21 [DASS-21] anxiety subscale ≥ 15) or low anxiety (DASS-21 anxiety subscale ≤ 14).¹⁵

Initial assessment (in person)

All participants complete a preliminary assessment in person at the rehabilitation facility in accordance with local health and safety guidelines and the guidelines issued by the facility’s COVID-19 Incident Command.

Participants, regardless of group allocation, undergo a 40-minute, two-channel photoplethysmograph (PPG) and respiration strain gauge (RSG) RF assessment using Thought Technology ProComp 2 T7400M Encoder (Montreal, Canada-Quebec) and T7900 Biograph Infinity Software v6.2 (see eAppendix B for PPG/RSG equipment set up). The assessment is conducted in a quiet clinic examination room free of distractions. Although electrocardiography (ECG) is considered the gold standard for HRV analysis, PPG is often used in clinical settings due to its pragmatic utility (ease of set up) and less invasive application (requiring only single finger placement rather than multiple chest sensor placements). The reliability of PPG as a surrogate for HRV is variable, although many agree that it is a stable alternative when used in a controlled setting with prolonged measurement duration and inactivity.^16–19

Treatment arm participants direct their attention to a computer monitor and practice paced breathing in 3-minute segments starting at 10 breaths per minute and incrementally reduce pace to end at four breaths per minute. Breathing is guided by an on-screen pacer in the form of a ball rising and falling, with rising indicating rate of inhalation and falling indicating rate of exhalation. Control arm participants do not complete the paced breathing task; rather, they are instructed to, “Choose a quiet activity that you would normally do to reduce feelings of stress or anxiety (e.g., read, listen to music). Please try to limit movement, conversation, and distractions as much as possible during this timeframe.”

Throughout the assessment, the study coordinator actively monitors for intrusions (e.g., movement, talking) that may impact the resulting data. Marking the time and reason of each intrusion facilitates the post-assessment artifact rejection process completed by the Biograph Infinity Software. For treatment arm participants, a RF report is generated using the Biograph Infinity Software and is sent to the study psychologist to determine the optimal breathing rate (i.e., RF) at which the participant will train for the eight biofeedback treatment sessions. The RF for each individual can be detected as the frequency at which maximum HRV is produced, when the system is rhythmically stimulated at that frequency. The RF heart rate for most individuals is close to 0.1 Hz, or about six cycles per minute.

Following the assessment, the study coordinator provides the participant with a home HRV sensor and assists with downloading the corresponding free mobile app on their smartphone. The mobile app synchronizes with the HRV sensor and is used for physiological monitoring; for those in the treatment arm, it is also used to deliver the biofeedback training. The study coordinator sets up the biofeedback training protocol within the app to ensure consistency among participants (see eAppendix C for set up specifications). To minimize in-person contact time, participants view a video at home demonstrating how to properly utilize the sensor for physiological monitoring.

Eight-session intervention

The intervention uses a commercially available HRV sensor and smartphone app, Mindfield eSense Pulse, to measure HRV and to deliver the biofeedback training. eSense Pulse offers one-channel ECG with 500 Hz sampling and a chest strap to measure heart rate. No identifiable information is collected and/or stored in the mobile app, and data can easily be exported and emailed to the study coordinator after each session. Both study arms undergo 20 minutes of self-administered physiological monitoring at home with the eSense Pulse twice a week for 4 weeks (eight sessions total). Each arm has a different session protocol, as described below.

Treatment arm. Participants self-administer physiological monitoring while completing biofeedback HRV training with the eSense Pulse.

• Prior to each session, the study coordinator meets with the participant via Zoom to provide an explanation of the process of biofeedback, ensure proper sensor placement and smartphone app set up, and collect pre-session data (see Table 1).

• To train and improve HRV, participants use the smartphone app to practice breathing using the optimal breath rate identified during the initial assessment. Feedback depicts real-time parameters of HRV where the participant follows a breath pacer, inhaling as a sphere expands and exhaling as it contracts. Participants use this controlled breathing to assist with reaching a relaxed state.

• Directly after each session, the data collector, who is unaware of the assigned group allocation, calls the participant to collect post-session measures (see Table 1) and reminds the participant to export and email the de-identified session data to the study coordinator.

• After the final session, the data collector determines whether study participation impacted anything other than stress, anxiety, or pain by asking a single open-ended question.

Control arm. Participants self-administer physiological monitoring using the eSense Pulse sensor but do not undergo biofeedback HRV training.

• Prior to each session, the study coordinator meets with the participant via Zoom to ensure proper sensor placement and smartphone app set up and to collect pre-session data (see Table 1). Participants are provided with the following instructions: “Please monitor yourself with this equipment for 20 minutes twice a week for 4 weeks. Please try to limit movement, conversation, and distractions as much as possible while you are monitoring.”

• Directly after each session, the data collector, who is unaware of the assigned allocation, calls the participant to collect post-session measures (see Table 1) and reminds the participant to export and email the de-identified session data to the study coordinator.

• After the final session, the data collector determines whether study participation impacted anything other than stress, anxiety, or pain by asking a single open-ended question. The participant is also offered instruction on using the app for biofeedback training.

If a participant is unable to complete a session, the session is rescheduled, and subsequent sessions are shifted accordingly. If a participant misses two sessions and is unable to reschedule within 1 week, they are withdrawn from the study and a new participant is recruited to replace them. As remuneration for study participation, participants keep the eSense Pulse assigned to them at the beginning of the study.

Primary outcome measures

Percentage of time spent in the low frequency (LF) band is generated within the eSense Pulse software and emailed to the study coordinator after each HRV session. Optimal coherence is achieved via “RF training” within the LF band (0.04 and 0.15 Hz)²⁰; thus the greater percentage of time spent in LF is indicative of more HRV and greater health benefits.

The Depression Anxiety Stress Scale-21 (DASS-21) is a 21-item instrument designed to measure the core symptoms of depression, anxiety, and stress.^15,21 The DASS-21 is validated in clinical and nonclinical samples,^22–24 including SCI.²⁵ The anxiety subscale is used as the primary outcome measure in this study, although all subscales will be analyzed. The DASS-21 is administered twice: pre- and post-intervention.

Adverse events, protocol completion rates, and number of withdrawals due to undesirable symptoms are tracked to demonstrate feasibility. A symptom list is used to track any new symptoms that emerge during the study. Participants complete the symptom list prior to and after each session. The type and median number of new symptoms are calculated and monitored throughout the study to assess study feasibility.

Secondary outcome measures

The Subjective Units of Distress Scale (SUDS)²⁶ is a self-report Likert-type scale used to measure the intensity of feelings and other internal experiences, such as anxiety, anger, agitation, stress, or other painful feelings. It is validated as a global measure of both physical and emotional discomfort.^26,27 The SUDS is administered at baseline and after each of the eight sessions.

The State Trait Anxiety Inventory (STAI) is a 40-item self-report inventory comprising 20 items to assess trait anxiety and 20 items to assess state anxiety.²⁸ The STAI is used and validated with SCI samples.^2,29 The STAI is administered to screen for enrollment in the study and used to measure anxiety after each session.

Exploratory measures and covariates

The Brief Pain Inventory (BPI)³⁰ is one of the most widely used tools to assess pain. For this study, the four pain severity items are used to assess worst, least, average, and current pain. We ask one additional item regarding the amount of pain relief that treatment provided during the session.

Impact of study participation is a single, open-ended item used to determine if study participation impacted anything other than stress, anxiety, or pain. Following the final session, participants are asked, “Are there conditions other than stress, anxiety, or pain that were impacted by your participation in this study? If so, please list.”

A demographic questionnaire is administered to collect information regarding age, gender, ethnicity, marital status, education, occupational status, injury etiology, level of SCI, American Spinal Injury Association Impairment Scale (AIS) classification, and date of injury.

A medication list is obtained at baseline. Participants report current daily and as-needed medications (both prescribed and nonprescribed substances), dosages, the reasons prescribed, length of time at the current dose, route of administration, and time of day medication is taken. At all subsequent sessions, participants respond to questions regarding changes in medications, dosages, and date/time medications were last taken (medication use).

Data management, monitoring, and confidentiality

Presession data (e.g., medications) are collected by the study coordinator, and post-session outcome data are collected by the data collector, who is unaware of assigned allocation. After each session, participants export the HRV data from the eSense app and email it to the study coordinator. All data are then entered into a HIPAA-compliant web-based application, Research Electronic Data Capture (REDCap), where access to data is restricted based on study role to ensure allocation concealment. The study psychologists review the HRV data weekly to assess for inconsistencies (e.g., abnormal heart rate reading), so the study coordinator can troubleshoot with the participant, if needed. If a participant self-reports a new symptom during the current session or reports seeking medical attention for a new symptom following the prior session, the study physician and principal investigator assess reporting criteria for an adverse event and report to the IRB and study sponsor accordingly. Participants are identified by study ID, and personal identifiers are stored separate from study data and kept under lock and key. No identifying information will be used in analysis or publication.

Data analysis

This is an unpowered pilot study with a small sample (n = 30) that will use a conservative descriptive analysis approach to analyze feasibility objectives and provide estimations of a treatment effect. The data will be aggregated into ratios, frequencies, pre-post change scores, and medians (ranges) per arm to evaluate feasibility objectives and study aims. Participant flow will be presented in a CONSORT diagram with all exclusions clearly described. The characteristics of the enrolled sample will be presented in a demographics table by arm and by total sample.

Ethics and dissemination

The study was approved by the HCA-HealthONE IRB (study no. 1384215-2) and the internal research committee of the institution. The trial is registered with clinicaltrials.gov (NCT# 03975075). Upon completion of the study, an article with the results will be published in a peer-reviewed journal.

Discussion

The COVID-19 pandemic has changed the research landscape worldwide, including how research trials are conducted.³¹ The highly transmissible nature of the SARS-Co-V2 respiratory virus posed a significant barrier to the conduct of clinical research, resulting in halted trials and delayed startups. After months of trial interruptions, the research community voiced concerns around access to treatment, safety monitoring, the safety of patients and staff, and stalled trial progression. In response, the US Food and Drug Administration (FDA) provided emergency guidance on conducting clinical trials during the COVID-19 pandemic.³² However, modifying existing protocols to address these concerns whilst maintaining strong methodological rigor has proved challenging. Using FRAME,¹² we provide comprehensive documentation of the protocol modifications that were made in response to the COVID-19 pandemic, the stakeholders involved in decision making, and context around the reasons and goals for the modifications. Even with these modifications, we are able to maintain methodological rigor. Although moving the intervention out of the controlled clinical environment into the home of participants threatens internal validity of the study, external validity will be strengthened. Affordability, convenience, and ease of use are all benefits of utilizing a commercially available consumer product to deliver the HRV intervention. With this more pragmatic approach, we hope to hasten the 17-year research-to-implementation gap³³ by providing a low-burden, cost-effective, and easily accessible intervention. However, moving to this more practical delivery model means sacrificing the use of ECG to detect improvement in HRV, the current gold standard HRV measurement. Even so, our modified outcome measure (percentage of average time spent in LF) is an important indicator of HRV,²⁰ and we expect to see a signal of improvement in this outcome and in the subjective outcome measures of this study.

Conclusion

The modified protocol described in this article is not without limitations. Implementing an in-home treatment introduces the potential for distractions not typically experienced in a stimulus-controlled clinical setting. Further, conducting the initial assessment in the formal clinic setting may induce heightened feelings of anxiety for some individuals, which in turn might impact the derived RF. Both of these limitations may ultimately confound the results. Nonetheless, the potential that this home HRV biofeedback intervention is feasible and demonstrates a trend in reduction of anxiety among individuals with SCI/D has important clinical and treatment implications that will warrant further testing.

Funding Statement

Financial Support This study is funded by the Craig H. Neilsen Foundation (594559), and all study-related activities have been approved by the HCA-HealthONE Institutional Review Board (study no. 1384215-2) and listed on clinicaltrials.gov (NCT# 03975075). For reporting, we used the CONSERVE-SPIRIT Reporting Guidelines for Trial Protocols Modified due to the COVID-19 Pandemic and Other Extenuating Circumstances and the Framework for Reporting Adaptations and Modifications-Expanded (FRAME).