Telemedicine-based cognitive behavioral therapy (TENACITY) for Veterans with chronic migraine: A pilot randomized controlled trial protocol

Abstract

Background

Behavioral interventions for people with headache (HA) diseases have demonstrated clinical efficacy. Despite strong empirical support and endorsement by professional practice organizations, behavioral interventions are not widely accessible to HA patients. Clinic-based behavioral treatments remain inaccessible to many. The objectives of this study are to conduct a pilot randomized controlled trial and to determine the feasibility of the implementation of a cognitive behavioral therapy protocol for HA (CBT-HA) delivered via telehealth (TENACITY) compared to enhanced HA usual treatment (EHUT).

Methods/design

TENACITY is a single-blinded, randomized, controlled, pilot trial with a single intervention arm and an EHUT control arm analyzed as intention-to-treat. This multi-site study had a 31-day run-in/screening period to confirm chronic migraine and an approximately 12-week CBT-HA treatment period. Follow-up assessments were conducted at 3- and 6-month post-treatment initiation.

Analysis plan

The primary outcome will be analyzed as the change in self-reported number of HA days from the 30-day baseline run-in period before treatment to the number of self-reported days at 3 months post-treatment using a generalized mixed effects model. Secondary outcomes at 3- and 6-month post-treatment will be analyzed in addition.

Trial Registration

IRP 20-002, TENACITY: Telemedicine-based cognitive therapy for migraine NCT04613362

Graphical abstract

This is a visual representation of the abstract.

Introduction

Individuals with chronic migraine (CM) routinely face barriers that limit their access to evidence-based (EB) behavioral headache (HA) treatment. ¹ With the Veterans Health Administration (VHA) infrastructure and efficient telehealth delivery models for patient care, the delivery of a behavioral intervention for HA via a telehealth platform is optimal to address barriers of in-person care delivery and holds considerable promise to reach and reduce Veteran patients’ HA days.

Behavioral treatments for HA have consistently been shown to be effective and have demonstrated clinically meaningful benefits,^1–3 leading numerous professional practice organizations to recommend behavioral HA treatments as front-line interventions for CM.

Despite strong empirical support and endorsement by professional practice organizations, behavioral interventions are not widely accessible to people with HA diseases. ⁴ Even when trained practitioners are available, clinic-based behavioral treatments remain inaccessible to many. ⁵ Novel delivery models that address patient barriers to HA care may also improve patient access to behavioral treatments. Telehealth virtual approaches for the delivery of CBT protocol for HA (CBT-HA) have not been widely studied for HA management. Therefore, we aimed to implement and test a telehealth-based delivery of CBT for CM employing our manualized migraine treatment protocol. ⁶

Research aims

We aimed to conduct a pilot randomized controlled trial (RCT) and determine the feasibility of the implementation of a CBT-HA delivered via telehealth (TENACITY) compared to enhanced HA usual treatment (EHUT) for Veterans diagnosed with CM across three VA medical centers (VAMCs).

The specific aims are threefold:

Aim 1: To develop a bundle of EB practice implementation strategies to engage three VAMCs [two Headache Centers of Excellence (HCoEs) and one general neurology service] and facilitate their local adaptation and implementation of CBT-HA (TENACITY) through a telehealth service platform.

Aim 2. To conduct a pilot RCT and determine the preliminary efficacy and feasibility of TENACITY compared to enhanced treatment as usual (EHUT) across three VA sites.

Hypothesis

1: Veterans receiving TENACITY will experience a statistically significant reduction in routine clinical HA metrics: HA frequency [HA days per month] (primary outcome), HA-related impairment, and psychological symptoms (secondary outcomes) compared to EHUT at 3 and 6 months.

Aim 3: To conduct an exploratory cost analysis of TENACITY from the Veteran's perspective, using inputs from the pilot RCT, and a two-year budget impact analysis (BIA) from the VHA's perspective, incorporating the costs of implementation as well as direct costs (and cost-savings) of providing the TENACITY intervention over all HCoEs to VHA.

Hypothesis 2:

TENACITY will be cost-effective and provide value to Veterans and VHA.

Methods

Design

TENACITY is a single-blinded, randomized, controlled, pilot trial with a single intervention arm and an EHUT control arm analyzed as intention-to-treat. This multi-site study has a 31-day run-in/screening period to confirm CM and an approximately 12-week treatment period. Follow-up assessments are conducted at 3- and 6-month post-treatment initiation (Figure 1). We received institutional review board approval at all three recruitment sites and the research team's headquartered site.

Figure 1.

Study design and participant flow.

Setting

Patients were recruited from three geographically distinct VAMCs. Site one hosts a large, multidisciplinary HCoE and is neurology-led with a health psychology program. Site two hosts a smaller VA Headache Center, which is neurology-led, with a health psychology program. Site three has a general neurology service with no established HCoE and is a large-volume site with no health psychology program for HA.

We identified potential study participants by applying the inclusion and exclusion criteria electronically from the electronic health records and manually by research staff during recruitment.

Study cohort and participants

Inclusion criteria. Veteran patients within the VHA medical system with an International Classification of Diseases, Tenth Revision (ICD-10) diagnosis of CM who completed at least 28 days of a HA diary at baseline with a confirmed frequency of ≥8 HA days per month at baseline are eligible for this study. Exclusion criteria. Veteran patients with any of the following diagnoses are ineligible for this study: post-traumatic HA, cluster HA, other primary HA, post-whiplash HA, and trigeminal autonomic cephalalgia diagnosis. Also excluded are participants who have received ≥90 days opioid therapy from the date of chart screening; have a current diagnosis of severe cognitive impairment indicated by short portable mental status questionnaire (SPMSQ) ⁷ screened by research staff; have a diagnosis of traumatic brain injury ≤1 year before diagnosis of CM or worsening of CM as indicated by post-traumatic HA screening tool; suffer from a disabling psychiatric illness (as noted by clinician in the medical records); have had a psychiatric hospitalization in the last 6 months; relate daily suicidal ideation within the last 2 weeks as indicated by the patient health questionnaire-9 (PHQ-9) with an answer of “3” to question nine; have severe depression, as indicated by PHQ-9 score ≥20; are terminally ill (life expectancy of <12 months as noted by clinician); decline to or cannot use the VA Nurse Annie mobile application (VA Annie App, which utilizes a text based service regarding health risk factor self-management and patient self-assessments of functioning); had received CBT during the past year; have been behaviorally treated by a HCoE clinical health psychologist for HA in last two years.

Recruitment

This study recruited participants from three geographically distinct VAMCs. All recruitment procedures were conducted via phone, with no in-person communication, which will allow for greater reach to include the entire VAMC catchment area and eliminate any potential barriers due to the COVID-19 pandemic.

Potential participants were identified by either direct referral from a clinician or from a HCoE-generated recruitment list, which utilized the VHA Headache Data cohort. ⁸ The cohort was constructed using VHA electronic clinical and administrative data for Veterans with a HA diagnosis. If inclusion criteria are met, the Veteran is sent a recruitment letter by mail, which describes patient incentives for participation ($25) per assessment period and receipt of a follow-up phone call. During the phone call, research staff will provide a study overview and conduct further screening if the Veteran indicates interest.

Sample size

We recruited Veterans with a goal of reaching 100 eligible participants across the three VAMCs. We randomized eligible Veterans to participate either in the TENACITY intervention (n = 50) or EHUT (n = 50). The project coordinator made the assignment and notified the research team. This is a pilot feasibility demonstration to provide preliminary efficacy. The sample size estimate was based on the staff resources available to screen, invite, and recruit patients to reach 100 subjects over a 15-month period.

Randomization

We used a computer-generated randomization listing stratified by three sites with no blocking. We used a password-protected computer file to conceal random assignment. The program manager will designate the random assignment upon participant completion of the baseline assessment and notify the respective clinical program and schedulers to contact the participants for their respective treatments.

Blinding

A TENACITY protocol evaluation was conducted with participants at approximately 3 months post-treatment initiation by blinded research staff to protect data integrity. We also evaluated maintenance at 6 months post-treatment initiation. To protect research personnel's blindness, the research staff member who is responsible for randomization collected self-reported medical care received during the prior 3 months.

Intervention components

The TENACITY intervention consists of protocolled, manualized, short-term CBT-HA sessions, which are administered via a telehealth delivery platform. Each session is approximately 50 min and follows a similar structure: (a) agenda setting; (b) home practice check-in; (c) learning a new skill; and (d) goal setting for home practice. The treatment will be delivered over approximately six biweekly sessions. ⁶

TENACITY's manualized CBT sessions are adaptable through the delivery medium (telehealth vs. in-person), by the number of sessions, and by the clinical role of the interventionists. The EHUT comparator was chosen to allow each site the flexibility of delivering an appropriate behavioral therapy alternative to control for the time and attention provided to the patients. The EHUT intervention consists of a usual care referral to an existing behavioral pain therapy available at the respective, specific site. The therapies available were CBT for chronic pain (CBT-CP), integrated pain mindfulness, and dialectical behavioral therapy. These therapies were delivered either individually or in a telehealth group setting. CBT-CP differs from CBT-HA in the differences in behavioral treatment targets for acute, preventive, and overuse HAs.

Interventionists

Two licensed clinical psychologists employed by two of the three target VAMCs trained to deliver CBT-HA manualized behavioral treatment provided the intervention. In addition, one of these psychologists provided a direct virtual telehealth clinic to the third site (VAMC), which is without any clinical psychologist trained in CBT-HA. The two psychologists met to coordinate the treatment protocol prior to the start of the intervention.

Interventions were initiated after patient randomization and subsequent scheduling with the appropriate clinician using the VHA clinical scheduler. Treatment was administered using the VA telehealth platform, VA Video Connect, and/or telephone in one-to-one clinical visits.

Concomitant care

There was no prohibited concomitant care during the course of the study. Participants were asked to continue their usual medical HA care.

Measurement outcomes

Data collection plan

Participants completed a baseline interview and health questionnaires via phone or electronically using VA Qualtrics. They were assigned a 31-day HA diary using VA Annie App. Research staff will obtain Veteran permission to audio-record the interviews. Each interview will be recorded and transcribed. The health questionnaires and HA diary will also be completed at 3-month and 6-month post-treatment initiation, along with a treatment satisfaction survey.

At the end of the 31-day run-in period, the participants were contacted by phone to review the HA diary. If the participant self-reported ≥8 HA days over the course of the prior 31 days and completed the first question of the HA diary on at least 28 days, s/he remained eligible for the study. Participants enrolled in the study were randomized to either the TENACITY arm or the EHUT control arm of the study (Figure 1).

The primary outcome to be evaluated is the change in the number of self-reported HA days during the past 31 days at 3 months compared to baseline.

Secondary outcomes included the change in HA-related impairment and psychological symptoms at 3 and 6 months compared to baseline. These include a battery of well-validated health questionnaires of functioning and symptoms that are tracked as part of routine HA care services and include the following: migraine-specific quality of life questionnaire, ⁹ migraine disability assessment, ¹⁰ HA-specific pain catastrophizing scale, ¹¹ HA management self-efficacy scale, ¹² post-traumatic stress disorder checklist for DSM-5 (PCL-5), ¹³ PHQ-9, ¹⁴ generalized anxiety disorder-7, ¹⁵ Veterans RAND 12-item health survey (VR-12) health-related quality of life, ¹⁶ and insomnia severity index. ¹⁷ We chose to include the PCL-5 and the VR-12, given our sample was comprised of Veterans receiving healthcare in the VHA.

Screening assessments conducted at baseline include the following: (a) SPMSQ ⁷ utilized to screen for cognitive impairment; (b) American Migraine Study/American Migraine Prevalence and Prevention (AMS/AMPP) diagnostic module, ¹⁸ the AMS/AMPP will be utilized to assess Veterans for CM according to international classification of HA disorders-2 criteria; (c) the PHQ-9 ¹⁴ is a reliable and well-validated self-report measure of depressive symptom severity and suicide risk; and (d) post-traumatic stress disorder checklist. ¹³

Participant timeline

Upon enrollment, participants began with a run-in period of a 31-day electronic HA diary. Next, participants completed a baseline interview and standardized questionnaires. On completion of baseline HA diary and baseline data collection, if all inclusion criteria were met, participants were randomized to either CBT-HA (TENACITY) or EHUT. Consults were placed to the appropriate clinical providers, with clinics contacting participants to schedule treatment. Participants completed the 31-day HA diary and health questionnaires again at approximately 3 and 6 months after treatment initiation, along with an additional treatment satisfaction survey (Figure 1). Participants were eligible to receive a total of $150 in gift cards for completion of all data collection measures.

Primary outcomes collection

Reduction of HA days in the 31 days prior to the 3-month post-start of treatment mark is the primary outcome; reduction in HA days in the 30 days prior to the 6-month post-start of treatment mark will allow for measurement of cumulative benefit and persistence of efficacy.

The VA Annie mobile application was utilized as the platform for a daily HA diary. ¹⁹ A research TENACITY protocol had been designed (Table 1) to send daily automated text messages to each participant for a total of 31 consecutive days. Participant responses are stored in the Annie system and are accessible by study staff. If a participant responds that they had a HA, eight additional questions are asked as indicated in Table 1. After being prompted with an Annie question, if the participant does not respond, a reminder message is sent every 20 min, with a maximum of three reminders per question.

Table 1.

VA Annie mobile application for daily headache assessment.

Q	Text	Response options	Action
1	Hello, it's Annie. Did you have a headache today?	Yes—HA 2 No—HA 1	If HA 2 → Q2 If HA 1 → Q9
2	On a scale from 0 to 10, with 0 = none and 10 = worst, how severe was the headache?	PN 0–PN 10	If response → Q3
3	Was the pain worse on one side of your head?	Yes—SD 2 No—SD 1	If response → Q4
4	Was the pain pounding/throbbing?	Yes—PD 2 No—PD 1	If response → Q5
5	Did light bother you?	Yes—LT 2 No—LT 1	If response → Q6
6	Did sound bother you?	Yes—SO 2 No—SO 1	If response → Q7
7	Did you feel nauseated or sick to your stomach or vomit?	Yes—VO 2 No—VO 1	If response → Q8
8	Was the pain made worse with routine activities such as walking or climbing stairs?	Yes—MV 2 No—MV 1	If response → Q9
9	On a scale from 0 to 10, with 0 = none and 10 = worst, how much did the headache interfere with your life today?	IN 0–IN 10

Analysis plans

Data management. Research staff entered participant assessment data directly into a secure database. In addition, we used a secure database (Qualtrics) for participants to enter their self-reported outcome data directly into a database.

Analyses comparing the TENACITY and EHUT control groups on primary and secondary outcomes were completed using linear mixed models for hierarchical and longitudinal data (patients are nested within site). The mediating and/or moderating effects of patient-level (e.g. age, sex, and depression) and facility-level variables will be examined.

We will conduct an intention-to-treat analysis. A generalized mixed effects model will evaluate the primary outcome of the number of HA days and the fixed effect of study arm, visit (time) (month 3 vs. baseline), and their interaction. Days will be clustered within Veterans, and a first-order autoregressive structure will be used for the relationship between adjacent days.

We will adjust for participant sex. We will evaluate missing data and will employ imputation for participants who complete at least 28 of the 31 days of the HA diary for our primary outcome. We will repeat these analyses with each of the secondary outcomes.

At 3 months, one participant from the CBT-HA arm and two participants from the EHUT arm did not complete treatment, and at 6 months, the CBT-HA arm did not lose any participants, and the EHUT arm lost one participant to follow-up.

Aim 3 cost data

We will employ microcosting techniques to assess costs ²⁰ in two phases: (a) startup costs to the organization for the bundled implementation strategy for TENACITY adoption, and (b) post-startup costs for delivery of the intervention and downstream costs of acute healthcare utilization for Veterans with migraine receiving TENACITY compared to EHUT. Costs for services within the VA will be assessed from the managerial cost accounting system for VA. ²¹

Cost analyses

We will evaluate the cost-effectiveness of TENACITY from the perspective of the individual Veteran and organizational costs at the VHA level. Cost analyses included both intervention and downstream medical costs at the Veteran level, along with implementation costs from the organizational perspective. Effectiveness outcomes included HA days per month and changes in health utility derived from general health-related quality of life (VR-12). We will evaluate organization costs to VHA with a BIA over a two-year time summing implementation costs to the HCoEs and the product of average cost (or cost-saving) per Veteran from the cost benefit analysis, with the number of Veterans who would benefit from receiving the intervention at all HCoEs for each year. ²³

Discussion

Telemedicine-based cognitive behavioral therapy (TENACITY) for HA (CBT-HA) is a single-blind, randomized, controlled, pilot trial with a single intervention arm and an EHUT control arm analyzed as intention-to-treat. While CBT-HA is an effective behavioral HA treatment, this study is one of the few studies to examine the effects of telehealth virtual delivery directly to patients with CM. Grazzi et al.²⁷ completed a small pilot study of virtual behavioral therapy for patients with HA during the COVID-19 pandemic. They found a significant reduction in migraine days, migraine medications, and migraine-related disability at 6 months; however, this effect was no longer present at 12 months. ²² There was no control group or randomization to compare to usual care.

We will also conduct an exploratory cost analysis of TENACITY from the Veteran's perspective, using inputs from the pilot RCT, and a two-year BIA from the VHA's perspective, incorporating the costs of implementation as well as direct costs of providing the TENACITY intervention.

Clinical implications

• CM is prevalent among Veteran patients and is related to poorer health-related quality of life.

• Patient access to CBT-HA is limited because a limited number of clinicians are trained in HA-specific behavioral treatments, and Veterans with HA report the negative barriers to treatment associated with the stress from traveling to onsite treatment.

• Delivery of Telehealth CBT-HA for CM may improve access and reach for a diverse patient population and improve health-related quality of life.