Contemporary Clinical Trials Communications 44 (2025) 101428

Contents lists available at ScienceDirect

Contemporary Clinical Trials Communications

journal homepage: www.elsevier.com/locate/conctc

Comparative main effects, mediators, and moderators of cognitive

behavioral therapy, acceptance and commitment therapy, and emotional
awareness and expression therapy for chronic spinal pain: Randomized
controlled trial rationale and protocol
John W. Burns a,* , Mark A. Lumley b , Kevin E. Vowles c , Mark P. Jensen d ,
Melissa A. Day e,d , Howard Schubiner f , Emma Jaszczak a , Britney Abro b,
Sarah H. Addicks g , Michael J. Bordieri h , Michael M. Dow i , Shoshana Krohner j ,
Zyanya Mendoza k , Eric C. Meyer l , Danielle Z. Miro m, Hallie Tankha n , David S. Tubman o ,
Jolin B. Yamin p , Dokyoung S. You q
a
Department of Psychiatry and Behavioral Sciences, Rush University Medical Center, Chicago, IL, USA
b
Department of Psychology, Wayne State University, Detroit, MI, USA
c
School of Psychology, Queen’s University Belfast, Belfast & Centre for Pain Rehabilitation, Belfast Health and Social Care Trust, Northern Ireland, UK
d
Department of Rehabilitation Medicine, University of Washington, Seattle, WA, USA
e
School of Psychology, The University of Queensland, Brisbane, Queensland, Australia
f
College of Human Medicine, Michigan State University, East Lansing, MI, USA
g
School of Health, Augsburg University, Minneapolis, MN, USA
h
Department of Psychology, Murray State University, Murray, KY, USA
i
Private Practice, Longmont, CO, USA
j
Department of Psychiatry, Montefiore/Albert Einstein College of Medicine, New York, NY, USA
k
Z Psychology Inc., Greater Los Angeles, California, USA
l
Department of Counseling and Behavioral Health, University of Pittsburgh, Pittsburgh, PA, USA
m
Private Practice, Reno, NV, USA
n
Department of Wellness and Preventive Medicine, Cleveland Clinic, Cleveland, OH, USA
o
Private Practice, Dayton, OH, USA
p
Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA
q
Department of Anesthesiology, Perioperative, and Pain Medicine, Stanford University, Palo Alto, CA, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Background: Chronic spinal (back/neck) pain is common and costly. Psychosocial treatments are available but
Chronic pain have modest effects. Knowledge of treatment mechanisms (mediators and moderators) can be used to enhance
Psychological therapy efficacy. Trials that directly compare different treatments are needed to determine which mechanisms are
Mechanism
treatment-specific, which are shared across treatments, and which contribute the most to outcomes.
Mediator
Moderator
Methods: We will conduct a 4-arm randomized, controlled clinical trial to compare the main effects, mediators,
and moderators of three pain therapies: Cognitive-Behavioral Therapy, Acceptance and Commitment Therapy,
and Emotional Awareness and Expression Therapy in adults with chronic spinal pain. Following baseline
assessment of outcomes variables (two primary outcomes: pain intensity and pain interference) and potential
mediators and moderators, we will randomize participants (up to 460) to one of the treatments or usual care
control. Treatments will be conducted individually each week for 8 weeks via telehealth. We will conduct weekly
assessments of both potential mediators and outcomes, as well as post-treatment and 6-month follow-up as­
sessments. We will test whether any of the therapies is superior to the others (Aim 1); identify mediators that are
specific to treatments and those that are shared across treatments (Aim 2); and identify baseline moderators that
are specific to treatments or shared across treatments, and moderated mediators of treatments (Aim 3).

* Corresponding author.
E-mail address: [email protected] (J.W. Burns).

https://doi.org/10.1016/j.conctc.2025.101428
Received 26 August 2024; Received in revised form 3 January 2025; Accepted 5 January 2025
Available online 6 January 2025
2451-8654/© 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

Discussion: The findings from this project can be used to improve the effects of psychosocial chronic pain
treatments by identifying the most powerful specific and shared mechanisms and revealing for whom the
mediator-outcome pathways are strongest.

1. Introduction must conduct large trials of multiple, conceptually and technically

different treatments [19,20]. In this trial, individuals with chronic spinal
Chronic spinal or axial (back, neck) pain is a major public health pain will be randomized to CBT, ACT, EAET, or treatment as usual
concern [1,2]. Various psychosocial therapies show evidence of (TAU). The three treatments were chosen because of evidence of efficacy
reducing pain and/or improving behavioral and emotional functioning, and because they are based on divergent theories, propose different
but these treatments have been shown to have modest effects on average mechanisms of action, and use different techniques. The TAU control is
[3–5], which is a shortcoming of these otherwise promising approaches. needed to determine whether any specific/shared mechanisms are just
We can boost treatment efficacy of existing treatments, or develop ap­ artifacts of changes that naturally occur over time or with participation
proaches capable of producing enhanced outcomes by identifying the in a study.
most potent mechanisms of these treatments as well as discover how to Each treatment will consist of 8 weekly sessions; potential modera­
better match patients to specific treatments [6]. Our planned clinical tors will be assessed at baseline, and mediator and outcome variables
trial compares not only the outcomes but also the mechanisms and will be assessed at baseline, weekly throughout treatment (or TAU), and
predictors/moderators of three psychosocial pain therapies: Cognitive at 6-month follow-up. Pain intensity and interference are co-primary
Behavioral Therapy (CBT) [7]; Acceptance and Commitment Therapy outcomes, and post-treatment is the primary endpoint. We will
(ACT) [8]; and Emotional Awareness and Expression Therapy (EAET) compare the main effects of the three therapies with each other and
[9]. TAU, but more importantly, we will characterize mediator, moderator,
and mediated-moderator effects over the course of therapies. Doing so
1.1. A strategy to increase efficacy: the study of mechanisms will clarify which mediators change across different treatments, which
predict outcomes, and which are moderated by baseline characteristics.
One strategy to enhance the effects of interventions is to increase
understanding of treatment mechanisms, by identifying the nature and 1.3. Specific aims
effects of both mediators and moderators [10]. That is, to boost pain
treatment effects, we need to know by what processes, in general, Aim 1: Main effects. The literature usually finds few or no outcome
treatments work and for exactly whom, in particular, the processes are differences among therapies; thus, we hypothesize that the three treat­
most strongly activated. ments will result in better outcomes than TAU but not differ in outcomes
Mediators. Implicit in many psychosocial pain treatment theories is among themselves, although EAET is newer, and initial tests suggest
the assumption that different treatments affect outcomes via mediators possible better outcomes than CBT in some populations [21–23]
that are unique or specific to that treatment. Few studies, however, have (Table 1 presents the full list of primary and secondary outcomes as well
directly compared different treatments regarding their mediators, and as possible mediators and moderators.).
the limited evidence suggests that different pain treatments may operate Aim 2: Mediators. Hypothesized treatment-specific mediators (CBT:
mostly (but perhaps not entirely) via shared or general (rather than behavioral activation and pain control beliefs; ACT: pain acceptance and
treatment-specific) mediators [11]. Shared mediators include, for self-compassion; EAET: emotional approach coping and attributions of
example, changes in pain-specific cognitions, emotions, and behaviors pain to the brain rather than body) would change primarily in the tar­
[12–14], as well as general psychotherapy mediators such as positive geted treatment; and lagged analyses would show that previous-week
outcome expectancies and the therapeutic alliance [11,15]. mediator changes predict next-week outcome changes primarily in the
Moderators. Focusing solely on group averages misses information relevant treatment. In contrast, shared (rather than treatment-specific)
reflected in variability in treatment response—individuals respond with mediators would show change irrespective of treatment condition, and
different levels of improvement [16]. The search for treatment moder­ earlier mediator changes would predict later changes in outcomes across
ators, however, has had limited success, owing to at least five factors: (a) multiple treatments. Analyses will determine which shared mediators
most research has been conducted on samples too small to support are most critical for outcomes.
moderator analyses; (b) most analyses to date have been conducted Aim 3: Moderators. We expect that people high in pain catastroph­
post-hoc, exploring variables that happen to be available rather than izing will respond best to CBT; those low in catastrophizing and/or high
following a priori theory and hypotheses; (c) moderator research rarely in experiential avoidance will respond best to ACT; and those low in
compares different treatments; (d) the identification of moderators has alexithymia and/or with trauma history will respond best to EAET.
often proceeded independently of understanding treatment mediators; Aim 4: Moderated mediators. According to the LA&E model, individ­
and (e) research often finds non-specific “predictors” of treatment out­ ual differences among participants at baseline (i.e., potential modera­
comes; that is, baseline characteristics that predict outcomes similarly tors) exert their effects on treatment outcomes because they activate or
for multiple treatments [17]. This trial addresses these limitations, influence specific mediators within a treatment. For example, a medi­
including the use of a theoretical model to guide the identification of ator specific to EAET – increases in emotional approach coping – may be
potential moderators (and moderated mediation pathways) of psycho­ activated most robustly among people with a history of trauma (acting
logical interventions for pain: the Limit, Activate, and Enhance (LA&E) here as a moderator) thereby driving favorable outcomes for this sub­
Model [18]. This model supports theory-based predictions regarding group of participants. That is, we will test whether moderators exert
moderators as a function of whether treatments limit maladaptive re­ effects on outcomes among people in specific treatment conditions via
sponses, activate adaptive responses, and enhance treatment outcomes hypothesized mediators.
based on patient strengths and resources.
2. Methods
1.2. New directions and the present study
2.1. Participants
To test the presence of mechanisms (i.e., mediators and moder­
ators)—whether they are specific to or shared across treatments—one We aim to randomize 460 adults, recruited from the community and

2
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

Table 1 Table 1 (continued )

Measures in the trial and schedule of administration. Measure Screen Baseline Week/ Post-tx 6-
Measure Screen Baseline Week/ Post-tx 6- sessions 1 (week month
sessions 1 (week month to 7 8: TAU follow-
to 7 8: TAU follow- up
up
Inventory (6 items);
On-line screen items X ​ ​ ​ ​ EAET
Inclusion/exclusion X ​ ​ ​ ​ Psychological Insight ​ X X X X
criteria (Zoom Questionnaire (6
interview) items); EAET
Brain and Stress ​ X X X X
Attributions for Pain
PRIMARY OUTCOMES: ALL TIME POINTS (7 items); EAET
BPI-4 (4-item Pain ​ X X X X Pain Anxiety Symptom ​ X X X X
severity scale) Scale (4 items); ALL
PROMIS Pain ​ X X X X Pain Catastrophizing ​ X X X X
Interference SF 8a Scale 6-item
(McWilliams); ALL
SECONDARY OUTCOMES: ALL TIME POINTS
Physical function: ​ X X X X POTENTIAL MEDIATORS: DURING TREATMENT ONLY
PROMIS 4-item (from Adverse Events ​ ​ X X X
PROMIS-29) Working Alliance ​ ​ X X ​
Anxiety: PROMIS 4- ​ X X X X Inventory-Patient (6
item (from PROMIS- items)
29) Working Alliance ​ ​ X X ​
Depression: PROMIS 4- ​ X X X X Inventory- Therapist
item (from PROMIS- (6 items)
29) Therapist Checklist of ​ ​ X X ​
Depression: PHQ-8 ​ X X X X Patient Engagement
Fatigue: PROMIS 4-item ​ X X X X and Homework
(from PROMIS-29) Treatment Expectancy ​ ​ X X ​
Sleep disturbance: ​ X X X X and Credibility
PROMIS 4-item (from Questionnaire
PROMIS-29) (Borkovec 1972; from
Anger: PROMIS ​ X X X X MATCH)
Emotional distress-
Anger 5a
Positive affect (PANAS- ​ X X X X BACKGROUND/MODERATOR ONLY MEASURES
SF-5-item) Demographic/medical ​ X ​ ​ ​
PCL-5 Short Form (4 ​ X X X X history questions
items) Chronic Overlapping ​ X ​ ​ ​
Pain Stages of Change ​ X X X X Pain Conditions
Questionnaire (13 Medication ​ X ​ ​ ​
Prep/Action items) Trauma History ​ X ​ ​ ​
Opioid Use (past 7 days) ​ X X X X Questionnaire-
Patient Global ​ ​ X X X Modified (25 items)
Impression of Change Big-5 Inventory (25 ​ X ​ ​ ​
Post-treatment ​ ​ ​ X X items; Johns)
Satisfaction Multidimensional Scale ​ X ​ ​ ​
Questionnaire of Perceived Social
(Therapy arms only) Support (12 items)
Toronto Alexithymia ​ X ​ ​ ​
Scale-20
POTENTIAL MEDIATORS AND THEORETICALLY-SPECIFIED TREATMENT: ALL Pain Detect (7 items) ​ X ​ ​ ​
TIME POINTS General Pain Treatment ​ X ​ ​ ​
Pain Self-efficacy ​ X X X X Expectancies
Questionnaire 4-item
(McWilliams); CBT
SOPA-2 (Pain control, ​ X X X X MODERATOR/OUTCOMES (NON-WEEKLY CHANGES)
Disability, Harm, Levels of Emotional ​ X ​ X ​
Emotion scales); CBT Awareness Scale
Behavioral Activation ​ X X X X (Forms A then B)
for Depression Scale- Chronic Pain ​ X ​ X X
Short form; CBT Attributions Scale-
Self-compassion scale ​ X X X X Short Form
(12 items); ACT Other Pain Treatments ​ X ​ X X
Chronic Pain ​ X X X X and Health Care Use
Acceptance Inventory of ​ X ​ X X
Questionnaire (8 Interpersonal
items); ACT Problems-32
Chronic Pain Values ​ X X X X ACR Fibromyalgia ​ X ​ X X
Inventory (6 x 2 Diagnostic Criteria
items); ACT 2011
Emotional Approach ​ X X X X PSYFlex (Psychological ​ X ​ X X
Coping Scale (4 Flexibility Scale; 6
items); EAET items)
Emotional ​ X X X X TAPS Tool Part II (3- ​ X ​ X X
Breakthrough month window)
(continued on next page)

3
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

Table 1 (continued )
Measure Screen Baseline Week/ Post-tx 6-
sessions 1 (week month
to 7 8: TAU follow-
up

PCL-5 (PTSD Checklist, ​ X ​ X X

20 items)
Employment status ​ X ​ X X

clinics using a range of methods (e.g., flyers in clinics, postings to social

media, advertisement in patient advocacy groups, medical center call
waiting announcements, ResearchMatch). Participants will report spinal
pain—anywhere in the back or neck—as their most prominent pain
problem. For inclusion, pain must: (a) be at least 3 months’ duration
and; (b) occur at least half the days in the past 6 months; and (c) have
both self-reported intensity and interference ratings (past month) of at
least 3 (0–10 scale). Additional inclusion criteria are: (d) fluency in
English; (e) access to personal computer/tablet; (f) ability to attend
weekly telehealth sessions; (g) willingness to be randomized; and (h)
residence in a U.S. state participating in psychologist licensure reci­
procity (i.e., PSYPACT state).
Exclusion criteria are few so that we can recruit a diverse and
generalizable sample. They include: (a) pain thought to be secondary to
a disease process (e.g., cancer, autoimmune disease) or neuropathy (e.g.,
leg pain greater than back pain); (b) alcohol/substance dependency or
use disorder within past 2 years; (c) psychotic or bipolar disorder
anticipated to interfere with therapy; (d) suicidality; (e) cognitive
impairment (2+ errors on 6-item screener [24]); (f) current or past year
application for disability or pain-related litigation; or (g) major surgery
planned for the next 9 months.

2.2. Procedure Fig. 1. Design of clinical trial.

The trial was approved by the Rush University Medical Center IRB (copying the therapist, who is assigned based on availability) informing
(other sites relegated authority to Rush) and registered on ClinicalTrials. participants of their therapist (or to continue treatment as usual). Par­
gov (NCT06044649). All study activities will be conducted remotely: ticipants learn in Session 1 to which therapy they are assigned; a process
recruitment, screening, and other communication will use email, text, which should minimize attrition or other biases that can occur when
and Zoom; assessments will use REDCap, and telehealth interventions participants are informed about their treatment prior to meeting their
will use Zoom. therapists. We will also assess participants’ perceived treatment credi­
Fig. 1 presents the design of the trial as a flow chart. bility after session 1, at which time participants learn of their assigned
Screening. Potential participants will complete a brief survey on-line; therapy and its rationale, This measure can inform the success of
those passing this screen will have a Zoom interview to determine blinding and the bias of differential expectations. Subsequently, par­
eligibility. Staff will describe the study to those who are eligible; if they ticipants are sent standardized emails for all weekly, post-treatment, and
remain interested, they will receive the consent form and schedule a follow-up assessments, which are conducted via REDCap. There are no
subsequent consent meeting, at which staff will obtain consent. After personal/unstructured interactions between staff and participants that
consent, participants will receive a REDCap link to baseline (pre-treat­ could bias responses; thus, outcome assessment is essentially blinded/
ment) measures. unbiased. The study statistician will analyze the data blind to condition
Randomization. After completing baseline measures, patients will be assignment.
randomized to one of four conditions. Before the trial, the statistician
will generate (with a customized randomization algoritm) randomiza­ 2.3. Treatment conditions
tion assignments, stratified by pain severity (two strata, defined by the
average of screening pain intensity and interference ratings: “high” is > All three treatments (CBT, ACT, and EAET) will be equal in format
6, “low is ≤ 6); such stratification will increase the likelihood of and duration: 8 sessions (allowed to occur over a 10-week period), 60
equivalence among treatment conditions at baseline on these key out­ min long, held weekly (no closer than 4 days apart), conducted indi­
comes. Randomization will also be done in blocks, thus increasing the vidually and remotely by a study therapist. Each therapy is guided by its
likelihood that the four arms will have equivalent sample sizes. The own manual, which describes principles and techniques, and includes a
sealed randomization schedule will be accessed by a researcher who has patient workbook. This workbook is introduced to the participant in
no participant contact, who will inform the research assistant of the session 1 by the therapist, who will email it to the participant after
condition assignment. The assistant will then inform the participant that session 1.
they are assigned to receive a treatment (but not which one) and that Cognitive Behavioral Therapy. CBT’s overarching premise is that
they should schedule Session 1 with a specific therapist, or that the chronic pain leads to learned, maladaptive patterns of thinking,
participant should continue their usual care (if assigned to TAU). behaving, and feeling, which exacerbate pain intensity, dysfunction, and
Blinding. This trial uses several types of blinding. First, participants distress. The version of CBT used in this study is based on several
and study personnel are blinded to condition until after baseline empirically-supported protocols [25–27]. Chronic pain can be success­
assessment. Then, staff will send a standardized email to participants fully managed by teaching people self-regulation, including: (a)

4
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

cognitive restructuring of maladaptive pain-related thoughts; (b) competence, and conviction—three quite disparate treatments, espe­
relaxation training and reassuring self-statements to reduce emotional cially those with contrasting views of the possibility and desirability of
distress and arousal; (c) activity-rest cycling to address excessive pain pain reversibility (EAET and ACT). The trainers/supervisors for the
behaviors (e.g., downtime); (d) pleasant activity scheduling to increase three therapies are Burns (CBT), Vowles (ACT), and Lumley and
the level and range of daily reinforcing activities; (e) problem-solving to Schubiner (EAET), each of whom has expertise and a history of training
identify barriers and develop solutions to cognitive and behavior professionals in their therapy. The use of separate supervision/training
changes; and (f) a written maintenance plan with short- and long-term experts and therapists for each intervention maximizes study “equi­
goals for applying these skills and dealing with possible setbacks. poise.” Also, the three lead investigators are committed to the success of
Weekly homework focuses on practicing these skills. each of their therapies; such “adversarial collaboration” reduces con­
Acceptance and Commitment Therapy. ACT seeks to decrease mal­ cerns about the bias that occurs with investigator allegiance effects.
adaptive avoidant behavioral responses to chronic pain, enhance All study therapists will receive training consisting of remote, 2-day
engagement in personally-valued activities, and decrease pain-related courses. Treatment sessions will be videorecorded (via Zoom) and
restrictions. This version of ACT is based on protocols by Vowles [28], downloaded/saved for supervision, fidelity assessment, and later pro­
which have been successfully used in previous trials [29–31]. ACT helps cess analyses. At the trial’s start, weekly supervision will be conducted
people identify areas of meaningful functioning that have been by the investigator leading each treatment, who will watch all sessions
adversely impacted by pain, learn methods to enhance pain acceptance of each therapist’s first participant to confirm the therapist’s compe­
in the service of these meaningful areas, and practice present-focused tence and adherence. Observation of additional cases and further su­
awareness/mindfulness skills to aid in taking effective action towards pervision/training will be conducted to maintain skills, as needed.
values and in a manner that is sustainable over the longer term. Sessions Ongoing verification of therapist adherence to the treatment pro­
include discussions of the impact of pain and distress avoidance efforts, tocols will be conducted during the trial by clinical psychology doctoral
identification of alternatives to this avoidance, establishment of plans students, who will review the 8-session course of therapy for 20 % of
for behavior change enabling progress towards valued living, behavior cases from each therapy and therapist, rating the presence of the core
change and experiential exercises (e.g., therapist demonstration; components of each therapy. The full course of each therapy will be the
role-playing), and homework to facilitate adaptive behavior change and target of fidelity ratings because ACT and EAET do not prescribe specific
aid in generalization. techniques to be conducted during specific sessions - unlike CBT - but
Emotional Awareness and Expression Therapy. EAET is based on rather follow principles that allow the delivery of techniques, as needed
contemporary pain neuroscience which views chronic pain (especially by individual patients, in various sessions. The percentage of compo­
primary pain) as a brain-based, predictive processing “alarm” system nents presented will be calculated for each therapy, and feedback from
that is unnecessarily activated when the person perceives threat or coders will be provided to supervisors and therapists to help prevent
danger—to either their physical body or their interpersonal and psy­ drift.
chological safety. Fearful (but incorrect) beliefs that pain reflects bodily
damage as well as fear and avoidance of emotions related to adverse life 2.5. Data collection and measures
experiences, trauma, and relationship conflicts can generate, maintain,
or exacerbate pain [32–36]. The goal of EAET, therefore, is to greatly Table 1 presents all measures and assessment points: baseline,
reduce or eliminate pain by having people unlearn these fears. This weekly during treatment/usual care, post-treatment (the week 8
version of EAET is an adapted version of the manual written for assessment), and 6-month follow-up (6 months after post-treatment).
group-based treatment of fibromyalgia [23,37]. EAET: (a) educates The measures are divided into: (a) outcomes (primary and secondary),
patients about pain neuroscience and the brain’s role in generating pain; which are assessed at all time points; (b) mediators, which are assessed
(b) uses pain provocation techniques and healthy affirmations to at all time points; (c) moderators, which are assessed only at baseline;
decrease the brain’s pain alarm related to perceived bodily injury or and (d) another set of potential moderators/secondary outcomes that we
damage; (c) facilitates disclosure of stressful or conflictual experiences assess only at the three major time points (baseline, post-treatment, and
and emotions; (d) processes emotions related to unresolved traumas and 6-month follow-up. Finally, there are several measures that apply only
or conflict (i.e., identify, experience, express, reflect on, and release to participants engaged in one of the three interventions (but not TAU),
emotions), particularly emotions related to power and autonomy (e.g., and these will be assessed after each session of treatment. Study thera­
anger), connection (tenderness, love), grief (sadness), and pists will also complete several measures after each session.
self-compassion; and (e) encourages healthy communication with This trial has two co-primary outcomes – pain intensity (assessed
others. Weekly homework exercises include engaging in reading or with the Brief Pain Inventory [38]) and pain interference (PROMIS Pain
watching videos, exploring stressors and emotions, and practicing new Interference scale [39]) – because the three therapies have different
ways of communicating. primary goals: ACT seeks to decrease pain interference, EAET seeks to
Treatment as usual. Participants assigned to TAU will not receive any reduce pain intensity, and CBT targets both outcomes. Our primary
of the three psychological treatments offered in this study. They will, study endpoint for main effects is post-treatment. All measures included
however, be asked to continue ongoing treatments for pain, including in this trial have evidence supporting their psychometric qualities. Most
medications and other therapies, as they and their care providers deem measures are participant self-reports, although therapists provide rat­
appropriate. Similarly, participants assigned to all three treatment ings after each session. Because there are numerous patient-report
conditions will be allowed to continue their usual care. measures, we will use brief measures or short forms to minimize
respondent burden, especially for the weekly assessments.
2.4. Therapist training, supervision, fidelity, and adherence Within 48 h of each treatment session (or weekly for TAU), partici­
pants will complete the online survey (REDCap), which contains the
Interventions will be delivered via Zoom by doctoral-level (PhD or battery of study outcomes and mediators. The data collected after ses­
PsyD), licensed clinical psychologists who will be “nested within” sion 8/week 8 will be considered the “post-treatment” assessment.
treatment rather than “crossed over” treatments; that is, a minimum of Participants will have the follow-up assessment 6 months later. Treat­
three psychologists with skill in and commitment to each therapy will ment completion is defined as completing at least 6 of the 8 planned
conduct it, rather than delivering all three therapies. Conducting only sessions. Participants who do not complete treatment will be encouraged
the therapy in which one has expertise is most ecologically valid and to provide post-treatment and follow-up assessments. Participants will
reduces concern about bias (diffusion, comparative motivation) that can be paid $50 for each of the baseline, post-treatment, and follow-up as­
occur when the same therapist tries to conduct—with confidence, sessments, and $25 for each of the 7 weekly assessments, for a possible

5
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

total of $325; treatment is provided for free. not specific to a treatment (i.e., shared mediators) will predict unique
variance in outcome changes.
2.6. Adverse events Baseline characteristics hypothesized to serve as moderators on
outcomes will be examined across all four conditions (Aim 3). Evidence
To ensure participant safety, we will monitor for adverse events via of a moderator will be revealed by the following: (a) a significant
patient self-report on the weekly surveys and by therapist report. Treatment Condition x Moderator × Time interaction for weekly
Adverse events are defined as any unfavorable or unintended symptom, changes in an outcome; (b) Moderator x Time simple effects such that
sign, or disease that may or may not be related to the treatment or study the moderator is related significantly to outcome change in one or two
participation. We will also monitor for serious adverse events (i.e., but not three treatment conditions. Significant interaction terms would
death, hospitalization, emergency room visits, suicide plans or attempts) indicate that differences in outcome across treatment conditions exist
at each treatment session. The funding agency (National Institute of based on these moderating factors.
Nursing Research) has approved our data safety and monitoring plan, Finally, we will examine whether associations between lagged
which calls for a person independent of the study and the study in­ mediator changes and outcome changes over the course of treatment
stitutions to serve as a Safety Officer. A full Data Safety Monitoring depend on hypothesized moderators; that is, moderated-mediator ef­
Board will not be convened. We will consult our Safety Officer within 24 fects (Aim 4). This approach addresses whether specific week-to-week
h upon discovery of any serious adverse event (whether or not it appears lagged relationships between mediator and outcome changes depend
to be study related) as well as any non-serious adverse events that are on levels of baseline characteristics.
judged by the participant or therapist as definitely or possibly study-
related. Such events may be reported to the IRB and/or NIH as deter­ 2.8. Trial status
mined by the Safety Officer.
The trial, funded with a 4-year NIH grant (originally planned as a 5-
2.7. Power and statistical analyses year grant) began screening participants in November 2023 and ran­
domized its first participant on December 7, 2023. It is expected that
Simulation-based power analyses were conducted to justify the randomization will continue through spring 2026, with follow-up and
sample size based on linear mixed model (LMM) analyses. We powered analyses occurring in the subsequent year.
for the fixed effects of treatment condition and interaction with time to
ensure power of at least 80 % at the 2-sided type I error rate of p = 0.05, 3. Discussion
and assuming attrition of approximately 15 % (based on our prior trials)
that is evenly spaced throughout program timepoints. This analysis This RCT is innovative and can expand knowledge of psychosocial
revealed that a sample size of 460 randomized participants should treatments in several ways. First, results will clarify what makes psy­
provide greater than 90 % power to detect differences among treatment chosocial pain treatments work. There is much value in examining
conditions (e.g., group × time interactions), provided endpoint group whether theorized mechanisms (mediators and moderators) are indeed
differences in outcomes that are of clinical significance (i.e., d = 0.4–0.5 actual mechanisms, whether certain treatments exert effects via mech­
or higher). Analysis of sample size for models using mediators as lagged anisms that are specific to a treatment, or whether certain mechanisms
time-varying covariates was conducted based on simulations [40,41], operate across different treatments. Such findings will reveal which one
which suggested that power nearly always exceeded 0.90 to detect or more mediators are driving change within each treatment, and for
outcome changes across time, dependent on change in mediators, with a whom, and whether other putative mediators may need to be targeted to
sample size of 460 and number of measurements exceeding eight. a greater degree. Second, RCTs comparing different pain interventions
Primary analyses include all randomized participants (intent-to- are very rare, but this trial compares three conceptually- and
treat). Analysis of treatment condition differences across treatment and technically-unique therapies, which have never been compared with
follow-up time points will use maximum likelihood-based LMM for each each other, and the trial does so with experimental equipoise. Such a
outcome [42]. Models will be adjusted for covariates as appropriate. To comparison will not only determine whether one therapy is more effi­
test for main effects (Aim 1), models will compare the four conditions to cacious than another, but importantly, test whether mediators and
detect differential slopes of change from baseline to post-treatment on moderators are specific or generalized and which should be targeted to
the two primary outcomes and secondary outcomes. Significant omnibus enhance treatment effects. Third, this trial will examine general mech­
(4-arm) tests will be followed by post-hoc comparisons of each treat­ anisms common to many efficacious interventions (e.g., the therapeutic
ment with TAU as well as each treatment with each other. Similar an­ alliance). Such findings could strengthen efforts to enhance these factors
alyses will compare treatment conditions from baseline to 6-month in pain treatments and direct training that not only focuses on tech­
follow-up. niques but reinforces the importance of, for instance, establishing and
We will test for mediators (Aim 2) that are, in theory, specific to each maintaining sound therapeutic relationships or expectations. Finally,
treatment, as well as for shared mediators, which operate across all this trial will be conducted entirely remotely. Almost all prior inter­
treatments. Evidence of specific mediators will be revealed by the vention research has been conducted in-person, but there are many
following (a) putative specific mediators (e.g., pain acceptance in ACT): limitations of that approach, including poor access for those living far
will show the largest pre-post changes in the relevant treatment condi­ from treatment centers, who have limited transportation or time, or
tion (e.g., ACT pain acceptance changes > CBT, EAET, TAU) (b) medi­ whose health problems or caregiver duties prevent participation. This
ators NOT specific to a treatment (e.g., pain acceptance in CBT); will trial will provide data to compare with prior in-person research while
NOT change substantially in that treatment; (c) in lagged analyses, simultaneously leading the field into the next generation of treatment
substantial changes in the mediators will precede and predict substantial delivery research and practice.
subsequent changes in outcomes. Evidence of shared mediators will be This trial has several potential limitations or challenges. First, the
revealed by the following: (a) mediators hypothesized as specific to a choice of control conditions for comparative RCTs is controversial. We
treatment as well as mediators thought NOT to be specific to a treatment include TAU as an ecologically-valid comparison because “usual care”­
will show substantial treatment changes in that treatment; (b) in lagged —without specialized additional psychological treatment—is what most
analyses, substantial changes in mediators specific to and not specific to patients actually experience. Also, active control conditions such as
a treatment, and general mediators (e.g., therapeutic alliance) will support or education, although controlling for non-specific factors, may
precede and predict substantial subsequent changes in outcomes; and (c) be efficacious themselves [43], thereby confounding comparisons that
changes in mediators specific to a treatment and changes in mediators seek to isolate the active ingredients of target treatments. Second,

6
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

studies vary in their targeted chronic pain problem; some are very het­ acquisition, Conceptualization. Howard Schubiner: Writing – review &
erogeneous (e.g., any musculoskeletal pain), impeding attempts to editing, Supervision, Methodology, Funding acquisition, Conceptuali­
match treatments to specific pain conditions, whereas others (e.g., low zation. Emma Jaszczak: Writing – review & editing, Project adminis­
back pain) seem homogeneous but ignore the fact that many patients tration, Methodology, Data curation. Britney Abro: Writing – review &
have other pain locations, commonly in the spinal/axial region. Thus, editing, Project administration, Methodology, Data curation. Sarah H.
we will target patients with spinal pain but also assess the extent of Addicks: Writing – review & editing, Data curation. Michael J. Bor­
musculoskeletal pain across the body and test this as a possible dieri: Writing – review & editing, Data curation. Michael M. Dow:
moderator. Third, a trial conducted remotely might raise concerns about Writing – review & editing, Data curation. Shoshana Krohner: Writing
equitable access due to technological differences; however, we think – review & editing, Data curation. Zyanya Mendoza: Writing – review
that access and retention will be at least as good, and likely better than & editing, Data curation. Eric C. Meyer: Writing – review & editing,
occurs in trials that require in-person attendance [44]. Although we do Data curation. Danielle Z. Miro: Writing – review & editing, Data
not know how these interventions delivered via telehealth will compare curation. Hallie Tankha: Writing – review & editing, Data curation.
to the larger literature generated via face-to-face interventions, we think David S. Tubman: Writing – review & editing, Data curation. Jolin B.
it is vital to obtain such remotely-delivered intervention data. Finally, Yamin: Writing – review & editing, Data curation. Dokyoung S. You:
because study participants may engage in other psychosocial or medical Writing – review & editing, Data curation.
pain treatments during the trial, we will assess for their presence and test
this factor either as a moderator or covariate. Funding source
This trial has implications for policy and practice. Results will show
which mechanisms should be activated to bring about treatment success, This trial is made possible by an R01 grant from the NIH/National
and conversely, which mechanisms are less important than previously Institute of Nursing Research: R01 020610; “Comparative Mechanisms
thought. Providers will have empirical bases for favoring certain tech­ (Mediators, Moderators) of Psychosocial Chronic Pain Treatments”;
niques and change processes and for de-emphasizing others. Results will Multiple PI: John Burns and Mark Lumley.
support new guidelines regarding patient characteristics contributing to
the best treatment matching, so that providers will gain empirically-
Declaration of competing interest
supported procedures for screening patients and recommending partic­
ipation in treatments best suited to them.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
4. Disclosures
the work reported in this paper.

Mark A. Lumley is a paid research consultant to CognifiSense, Inc,

Data availability
which develops virtual reality interventions for chronic pain. He also
facilitates workshops to train clinicians in pain treatments and some­
No data was used for the research described in the article.
times receives fees for doing so.
Kevin E. Vowles provides paid chronic pain treatment training and
References
supervision for organizations that provide treatment services for chronic
pain disorders, including Kaiser Permanente Southern California (USA) [1] M.P. Jensen, D.C. Turk, Contributions of psychology to the understanding and
and Connect Health Ltd (UK). treatment of people with chronic pain: why it matters to ALL psychologists, Am.
Mark P. Jensen has written or edited several books on the use of an Psychol. 69 (2) (2014) 105–118.
[2] U.S. Department of Health and Human Services, Pain Management Best Practices
intervention that can be used for chronic pain management (therapeutic Inter-Agency Task Force Report: Updates, Gaps, Inconsistencies, and
hypnosis) and received royalties from the sales of these books. He also Recommendations, 2019, May.
facilitates workshops to train clinicians in the use of therapeutic hyp­ [3] ACdC. Williams, E. Fisher, L. Hearn, C. Eccleston, Psychological therapies for the
management of chronic pain (excluding headache) in adults, Cochrane Database
nosis for pain management, and sometimes receives fees for doing so. He
Syst. Rev. (8) (2020).
also owns equity in a company that is developing a digital therapeutic [4] L. Hilton, S. Hempel, B.A. Ewing, et al., Mindfulness meditation for chronic pain:
that can provide therapeutic hypnosis. systematic review and meta-analysis, Ann. Behav. Med. 51 (2) (2017) 199–213.
Howard Schubiner has written or edited several books on chronic [5] L.S. Hughes, J. Clark, J.A. Colclough, E. Dale, D. McMillan, Acceptance and
commitment therapy (ACT) for chronic pain: a systematic review and meta-
pain treatment and received royalties from the sales of these books. He analyses, Clin. J. Pain 33 (6) (2017) 552–568.
facilitates workshops to train clinicians in chronic pain treatments and [6] H.C. Kraemer, G.T. Wilson, C.G. Fairburn, W.S. Agras, Mediators and moderators of
sometimes receives fees for doing so. He owns equity in two companies treatment effects in randomized clinical trials, Arch. Gen. Psychiatr. 59 (10) (2002)
877–883.
that provide education for patients and clinician training: Freedom From [7] D.M. Ehde, T.M. Dillworth, J.A. Turner, Cognitive-behavioral therapy for
Chronic Pain and Oviddx.com. individuals with chronic pain: efficacy, innovations, and directions for research,
Shoshana Krohner facilitates workshops to train clinicians in chronic Am. Psychol. 69 (2) (2014) 153–166.
[8] L.M. McCracken, K.E. Vowles, Acceptance and commitment therapy and
pain treatments and sometimes receives fees for doing so. She co-owns a mindfulness for chronic pain: model, process, and progress, Am. Psychol. 69 (2)
company that provides consultation and training to clinicians who treat (2014) 178–187.
chronic pain. [9] M.A. Lumley, H. Schubiner, Emotional awareness and expression therapy for
chronic pain: rationale, principles and techniques, evidence, and critical review,
No other authors have disclosures. Curr. Rheumatol. Rep. 21 (2019). Article 30.
[10] A.E. Kazdin, Mediators and mechanisms of change in psychotherapy research,
CRediT authorship contribution statement Annu. Rev. Clin. Psychol. 3 (1) (2007) 1–27.
[11] M.J. Lambert, Early response in psychotherapy: further evidence for the
importance of common factors rather than “placebo effects”, J. Clin. Psychol. 61
John W. Burns: Writing – original draft, Supervision, Project (7) (2005) 855–869.
administration, Methodology, Funding acquisition, Conceptualization. [12] J.W. Burns, M.A. Day, B.E. Thorn, Is reduction in pain catastrophizing a
therapeutic mechanism specific to cognitive-behavioral therapy for chronic pain?
Mark A. Lumley: Writing – original draft, Supervision, Project admin­
Translational Behavioral Medicine 2 (1) (2011) 22–29.
istration, Methodology, Funding acquisition. Kevin E. Vowles: Writing [13] B.E. Thorn, J.W. Burns, Common and specific treatment mechanisms in
– review & editing, Supervision, Resources, Project administration, psychosocial pain interventions: the need for a new research agenda, Pain 152 (4)
Funding acquisition, Conceptualization. Mark P. Jensen: Writing – re­ (2011) 705–706.
[14] M.A.M.A. Day, B.E.P.A. Thorn, J.W.P. Burns, The continuing evolution of
view & editing, Methodology, Funding acquisition, Conceptualization. biopsychosocial interventions for chronic pain, Journal of Cognitive
Melissa A. Day: Writing – review & editing, Methodology, Funding Psychotherapy, suppl Special Issue on Chronic Pain 26 (2) (2012) 114–129.

7
J.W. Burns et al. Contemporary Clinical Trials Communications 44 (2025) 101428

[15] L.G. Castonguay, M.J. Constantino, M.G. Holtforth, The working alliance: where [30] K.E. Vowles, K. Witkiewitz, K.J. Cusack, et al., Integrated behavioral treatment for
are we and where should we go? Psychother. Theor. Res. Pract. Train. 43 (3) veterans with Co-morbid chronic pain and hazardous opioid use: a randomized
(2006) 271–279. controlled pilot trial, J. Pain 21 (7) (2020) 798–807.
[16] M.A. Day, L.C. Ward, B.E. Thorn, et al., Mechanisms of mindfulness meditation, [31] H. Mo’tamedi, P. Rezaiemaram, A. Tavallaie, The effectiveness of a group-based
cognitive therapy, and mindfulness-based cognitive therapy for chronic low back acceptance and commitment additive therapy on rehabilitation of female
pain, Clin. J. Pain 36 (10) (2020) 740–749. outpatients with chronic headache: preliminary findings reducing 3 dimensions of
[17] A.E. Kazdin, Understanding how and why psychotherapy leads to change, headache impact, Headache J. Head Face Pain 52 (7) (2012) 1106–1119.
Psychother. Res. 19 (4–5) (2009) 418–428. [32] N. Afari, S.M. Ahumada, L.J. Wright, et al., Psychological trauma and functional
[18] M.A. Day, D.M. Ehde, M.P. Jensen, Psychosocial pain management moderation: the somatic syndromes: a systematic review and meta-analysis, Psychosom. Med. 76
limit, activate, and enhance model, J. Pain 16 (10) (2015) 947–960. (1) (2014) 2–11.
[19] ACdC. Williams, E. Fisher, L. Hearn, C. Eccleston, Evidence-based psychological [33] G.J.G. Asmundson, H.D. Hadjistavropolous, Addressing shared vulnerability for
interventions for adults with chronic pain: precision, control, quality, and comorbid PTSD and chronic pain: a cognitive-behavioral perspective, Cognit.
equipoise, Pain 160 (8) (2021) 2149–2153. Behav. Pract. 13 (1) (2006) 8–16.
[20] J.-P. Laurenceau, A.M. Hayes, G.C. Feldman, Some methodological and statistical [34] D.A. Davis, L.J. Luecken, A.J. Zautra, Are reports of childhood abuse related to the
issues in the study of change processes in psychotherapy, Clin. Psychol. Rev. 27 (6) experience of chronic pain in adulthood?: a meta-analytic review of the literature,
(2007) 682–695. Clin. J. Pain 21 (5) (2005) 398–405.
[21] B.C. Yarns, N.J. Jackson, A. Alas, R.J. Melrose, M.A. Lumley, D.L. Sultzer, [35] G.T. Jones, C. Power, G.J. Macfarlane, Adverse events in childhood and chronic
Emotional awareness and expression therapy vs cognitive behavioral therapy for widespread pain in adult life: results from the 1958 British Birth Cohort Study, Pain
chronic pain in older veterans: a randomized clinical trial, JAMA Netw. Open 7 (6) 143 (1–2) (2009) 92–96.
(2024) e2415842. [36] M.B. Yunus, Fibromyalgia and overlapping disorders: the unifying concept of
[22] B.C. Yarns, M.A. Lumley, J.T. Cassidy, et al., Emotional awareness and expression central sensitivity syndromes, Semin. Arthritis Rheum. 36 (6) (2007) 339–356.
therapy achieves greater pain reduction than cognitive behavioral therapy in older [37] M.A. Lumley, H. Schubiner, Emotional Awareness and Expression Therapy (EAET):
adults with chronic musculoskeletal pain: a preliminary randomized comparison A Group-Based Treatment Manual for Patients with Fibromyalgia and Related
trial, Pain Med. 21 (11) (2020) 2811–2822. Centralized Pain Conditions, Wayne Sate University, Detroit, MI, 2012, Version 2.
[23] M.A. Lumley, H. Schubiner, N.A. Lockhart, et al., Emotional awareness and [38] C. Cleeland, K. Ryan, Pain assessment: global use of the brief pain inventory,
expression therapy, cognitive-behavioral therapy, and education for fibromyalgia: Annals of the Academcy of Medicine 23 (1994) 129–138.
a cluster-randomized controlled trial, Pain 158 (2017) 2354–2363. [39] D. Amtmann, K.F. Cook, M.P. Jensen, et al., Development of a PROMIS item bank
[24] C.M. Callahan, F.W. Unverzagt, S.L. Hui, A.J. Perkins, H.C. Hendrie, Six-item to measure pain interference, Pain 150 (1) (2010) 173–182.
screener to identify cognitive impairment among potential subjects for clinical [40] X. Basagaña, X. Liao, D. Spiegelman, Power and sample size calculations for
research, Med. Care 40 (9) (2002) 771–781. longitudinal studies estimating a main effect of a time-varying exposure, Stat.
[25] B.E. Thorn, Cognitive Therapy for Chronic Pain: A Step-by-step Guide, Guilford Methods Med. Res. 20 (5) (2011) 471–487.
Press, New York, 2004. [41] X. Basagaña, D. Spiegelman, Power and sample size calculations for longitudinal
[26] D.C. Turk, D. Meichenbaum, M. Genest, Pain and Behavioral Medicine: A studies comparing rates of change with a time-varying exposure, Stat. Med. 29 (2)
Cognitive-Behavioral Perspective, Guilford Press, New York, 1983. (2010) 181–192.
[27] J.W. Carson, F.J. Keefe, G. Affleck, et al., A comparison of conventional pain [42] A.L. Duckworth, E. Tsukayama, H. May, Establishing causality using longitudinal
coping skills training and pain coping skills training with a maintenance training hierarchical linear modeling: an illustration predicting achievement from self-
component: a daily diary analysis of short- and long-term treatment effects, J. Pain control, Soc. Psychol. Personal. Sci. 1 (4) (2010) 311–317.
7 (9) (2006) 615–625. [43] B.E. Thorn, J.C. Eyer, B.P. Van Dyke, et al., Literacy-adapted cognitive behavioral
[28] K.E. Vowles, J.T. Sorrell, Life with chronic pain: an acceptance-based approach. therapy versus education for chronic pain at low-income clinics: a randomized
Therapist Guide and Patient Workbook, University of Bath, 2007. controlled trial, Ann. Intern. Med. 168 (7) (2018) 471–480.
[29] K.E. Vowles, J.L. Wetherell, J.T. Sorrell, Targeting acceptance, mindfulness, and [44] M.A. Day, M.A. Ciol, M.E. Mendoza, et al., The effects of telehealth-delivered
values-based action in chronic pain: findings of two preliminary trials of an mindfulness meditation, cognitive therapy, and behavioral activation for chronic
outpatient group-based intervention, Cognit. Behav. Pract. 16 (1) (2009) 49–58. low back pain: a randomized clinical trial, BMC Med. 22 (1) (2024) 156.