Chronic pain, a persistent problem for United States veterans, has a reported prevalence as high as 65.5% in the veteran population and is associated with limitations in mobility and daily activities, dependence on opioids, anxiety and depression, and poor perceived health [1,2]. Chronic pain is often inadequately treated with analgesics alone and results in substantial disability, reduced health-related quality of life, and increased health care use and costs. Musculoskeletal pain accounts for more than half of all patients with pain disorders presenting for clinical care and analgesic treatment is insufficient for many patients [3,4]. To improve pain-related outcomes, the Department of Veterans Affairs (VA) recommends an integrative approach to pain management, including nonpharmacological interventions [5,6].

Music therapy interventions target biopsychosocial outcomes and show promise as nonpharmacological options for pain. Music therapists use active (music making) and receptive (music listening) interventions to support pain management; however, much of the music and chronic illness literature comprises studies that use receptive interventions [7,8]. Studies of music listening interventions have demonstrated statistically significant reductions in self-reported pain, emotional distress, and opioid use, but the findings are inconsistent and although there are a few studies that address patients with chronic pain, the focus has been primarily on acute pain [9-13]. A majority of these studies evaluated recorded music listening programs, with inconsistent findings likely because of an absence of theoretical frameworks to guide music selection and tailored delivery, and educational components to encourage independent use of music. Music listening interventions often involve minimal interaction with a music therapist, and are primarily designed to be patient self-directed, empowering patients to use music whenever needed. However, some patients with complex pain and psychological comorbidities may need a more intensive and interactive music therapy intervention that includes ongoing support, symptom monitoring, and additional treatment components, such as imagery and verbal processing, to potentially enhance the therapeutic effects of music listening.

Compared with recorded listening programs, therapist-led music therapy interventions involve more interactive approaches to music listening that provide education about the therapeutic potential of music and often integrate additional treatment components (eg, lyric discussion, drawing or journaling) to address additional biopsychosocial factors that contribute to chronic pain [14,15]. Music imagery, a receptive music therapy intervention that combines music listening, imagery, and verbal processing, has been used to address a variety of patient needs (eg, symptoms related to posttraumatic stress disorder [PTSD], cancer, mood disorders, and chronic pain [16-19]). In pilot studies with nonpain populations (eg, PTSD, health care providers), participants found that music imagery enhanced their coping skills and ability to self-regulate [16,17]. Both music listening and music imagery have shown benefits for patients with fibromyalgia and other chronic pain illnesses, including improved well-being and decreased pain, anxiety, and depression [20-22]. Taken as a whole, the music therapy literature suggests that music listening and music imagery may improve pain outcomes, but gaps in the literature include inadequate evaluations among patients with chronic pain, theoretically grounded interventions, and lack of methodological rigor.

Although VA recommends using nonpharmacological interventions for chronic pain, there are significant barriers that limit access to creative arts interventions, including music therapy (eg, rural and homebound veterans [23]). Virtual delivery of health care can improve access, especially for the older adults, the mobility-impaired, and those living in rural areas or those with transportation barriers [24,25]. Case studies of virtually delivered music therapy have described veterans’ endorsement of telehealth sessions and patient-reported benefits [26]. Although these studies provide compelling preliminary data, more research is needed to develop and test specific music interventions for chronic pain that can be delivered through telehealth. In this context, we will evaluate 2 music interventions (music listening and music imagery) for veterans with chronic musculoskeletal pain.

The FAMILIA (Feasibility and Acceptability of Music Imagery and Listening Interventions for Analgesia) study addresses four limitations in music intervention research for chronic pain: few studies in veterans, evaluation of multicomponent music interventions, methodological rigor, and virtual delivery. Most previous studies described a passive, one-session, single-component music listening experience for patients with acute pain [9-13]. In contrast, the FAMILIA study will evaluate 2 music interventions in veterans. Compared with music listening, less is known about multicomponent music interventions, which we hypothesize are more robust in reducing pain owing to greater treatment intensity and more frequent interactions with a music therapist. The FAMILIA study will assess outcomes (feasibility, acceptability, pain, and associated outcomes) of a single-component, minimally interactive intervention (music listening) and a multicomponent, interactive intervention (music imagery).

Systematic reviews of music interventions for pain cite a lack of methodological rigor because of poorly described interventions, a lack of theoretical foundation, and no reporting of adverse events [27,28]. The FAMILIA study will address these methodological concerns in several ways. First, we will adhere to reporting guidelines for music-based interventions developed by Robb et al [29]. Second, we will use a theoretically based conceptual framework (Figure 1) to explore the hypothesized mechanism of action and outcomes. Third, we will systematically monitor for specific potential adverse events in both interventions.

FAMILIA addresses important knowledge gaps regarding virtual delivery of music therapy interventions. The COVID-19 pandemic has rapidly accelerated the widespread use of telehealth services for multiple chronic conditions, including chronic pain, but there are few published studies to guide virtual delivery of music therapy. We recently conducted the first enterprise-wide survey examining current practices in telehealth delivery by VA creative arts therapists. Initial results indicated that 76% of creative arts therapists have delivered virtual sessions, with 74% delivering >50 sessions in the past year [30]. Telehealth provides greater access, convenience, continuity of care, and support for veterans [31,32]. Early adopters of telehealth music therapy have provided general instructive examples, but there is a need for rigorous research of music interventions delivered virtually to determine their feasibility, acceptability, and efficacy [23,26,33]. We expect that our study will inform evidence-based practice and help expand access to music interventions for chronic pain.

The objective of this randomized controlled trial (RCT) is to examine feasibility and acceptability of the interventions, trial design, and theoretical model before conducting a fully powered efficacy or comparative effectiveness trial. Findings from FAMILIA will be used to refine the interventions and theoretical model, allowing for a comprehensive analysis of proposed mediators and outcomes in a subsequent trial. The 2 specific aims and the related key questions are shown in Textbox 1.

The FAMILIA study is a 3-arm, parallel group, RCT, as shown in Figure 2. After providing informed consent and completing their baseline assessment, participants will be randomized in a 2:1:1 allocation ratio to 1 of 3 groups. Twice as many participants will be randomized to the music imagery group than other groups because a key goal of this pilot study is to gain experience with and learn about patients’ perceptions of a multicomponent, interactive music therapy intervention. The goal sample size is 60 participants, with approximately 30 participants in the music imagery arm, 15 participants in the music listening arm, and 15 participants in the usual care arm. A sample size of 15 to 20 per group is sufficient to assess the feasibility and acceptability in pilot trials [34].

Randomization will be stratified according to biological sex (male vs female). Within strata, randomization with block sizes of 9 will be executed to ensure a balance of key baseline characteristics (eg, baseline pain intensity).

Listening to music can at times bring up strong emotions, which can increase stress. The potential of this risk is minimal because of the supportive structure of the interventions and the overall study. The therapists are trained to help participants manage any strong emotions that emerge and, if needed, will refer participants for additional psychosocial support.

At each follow-up assessment (1 month, 3 months, and 4 months), a member of the research team will systematically check for participant intercurrent illness, adverse events, and serious adverse events. At all assessments, including baseline, a member of the research team will screen for suicidal ideation guided by a 3-step algorithm consisting of the following: (1) item 9 suicide question of the Patient Health Questionnaire-9 (PHQ-9), (2) the Columbia Suicide Severity Rating Scale Screener if item 9 of PHQ-9 is positive, and (3) Comprehensive Suicide Risk Evaluation for moderate to high risk of suicide based on the Columbia Suicide Severity Rating Scale Screener [43-45].

To ensure intervention fidelity and minimize experimental drift, we will use the following strategies: (1) standardized intervention protocols, study manuals, and training; (2) self-monitoring of intervention sessions recorded in field notes with debriefings across music therapists; (3) quality assurance checklists to track protocol deviations; and (4) therapist field notes documenting session duration and frequency (dose).

As an RCT, it is expected that baseline characteristics of the study participants will be balanced across the 3 treatment groups; however, possible group imbalances will be assessed. Outcomes (feasibility, acceptability, pain, and other patient-centered outcomes) will be reported as means with SDs for continuous variables, or medians (IQRs) for nonlinear data and frequencies with percentages for categorical variables. Changes in pain, associated patient-centered outcomes, and proximal and distal mediators will be explored from baseline to follow-up assessments (1, 3, and 4 months). Bivariate analyses will be performed first to determine if there are significant differences in the 3 treatment groups for the antecedent, mediator, and outcome variables, which will also help to determine if there are group imbalances as discussed above. These analyses will be performed using ANOVA and chi-square tests for continuous and categorical variables, respectively. No post hoc P value adjustments will be made, as this is a feasibility study. Data transformations will be used if necessary and Fisher exact test will be used instead of chi-square when expected cell counts are small, with data analyses being performed using SAS software (version 9.4; SAS Institute [60]). Path analyses will then be performed to examine the path coefficients for the entire conceptual study model (Figure 1). The path coefficients, model fit indices, and covariance structures, along with the results from the bivariate analyses, will be used to determine model adjustments and parameter estimates necessary to determine the power for a larger full study. Path analyses will be performed using Mplus (version 8.7; Muthén & Muthén [61]).

Using guidelines from the thematic analysis by Braun and Clarke [62], the qualitative analysis of veteran interviews and music imagery participant listening logs will occur in 2 broad phases: open coding and focused coding. Open coding facilitates development of a code list for further analysis. In this phase, selected transcripts are read to gain a general understanding of the data and variation across participants. Then, each line will be independently labeled with initial codes to describe the data. The research team will meet to discuss and compare initial codes. The team will then examine the codes, looking for emerging categories to sort and organize the data. As new transcripts are reviewed and coded, categories may be altered or codes are restructured within the categories. This process will occur iteratively until agreement on, and shared understanding of, emergent thematic categories is established. In phase 2, focused coding, the categories derived in the first phase will be applied to transcripts. A subset of transcripts will be coded by all analysts using the agreed-upon categories. Team members will meet in pairs to ensure coding consistency, with discrepancies resolved by consensus. Representative quotations will be presented for each category.

FAMILIA was reviewed and approved by the Indiana University institutional review board and the VA research and development committee (IRB #12794).

The FAMILIA research team consulted with the VA Health Services Research & Development Service Pain/Opioid Consortium of Research Veteran Engagement Panel facilitation team and outlined a plan for how the panel can inform this project. The panel group includes 12 veterans from across the country who serve as patient advisors to researchers to enhance patient-centered research. All panel members have personal experience with chronic pain or opioids. Veteran members are a diverse group of men and women across a wide age range and include students, retirees, and current workers with varied military, volunteer, and professional experiences. Panel members provided patient-centered feedback that helped inform the study recruitment materials, study design, and anticipated feasibility issues.

Members of the research team conducted a single-arm study of 8 veterans with chronic pain to develop and pilot the telehealth MI intervention proposed for FAMILIA. Qualitative data were collected regarding the accessibility and feasibility of virtual delivery and the outcome measures of pain, depression, and anxiety. Findings from this small pilot study helped refine the music imagery intervention and its delivery for FAMILIA. For example, owing to some veterans’ lack of technology skills, a technology questionnaire was added to the eligibility screening.

There are several potential results and interpretations from the trial. It is expected that both music interventions will prove acceptable to veterans and that feasibility benchmarks will be reached. However, it is possible that one or both music interventions will not be embraced by all participants or may not meet all specified benchmarks. The qualitative interviews will identify the range of participant perceptions and experiences of both interventions and provide insight into the reasons for unacceptability or feasibility concerns. Although it is hypothesized that both music interventions will be more effective than usual care on pain and associated outcomes, differences in outcome scores may be observed but are unlikely to reach stochastic significance (because FAMILIA is not powered to evaluate efficacy). This potential scenario is acceptable in the context of pilot clinical trials where the primary objective is to field-test logistical aspects of the future RCT and to incorporate these aspects into the study design.

FAMILIA received institutional review board approval in October 2021 and approval from VA Research and Development in November 2021. Recruitment began in March 2022 and as of July 2022, a total of 16 participants were enrolled.

FAMILIA not only addresses the limitations of prior research but also responds to important clinical and research initiatives to develop and test safe and effective nonpharmacological interventions for chronic pain. In clinical practice, analgesics are widely used for pain, but suboptimal pain relief is common as are the side effects. Adverse events related to analgesics, especially nonsteroidal anti-inflammatory drugs and opioids, are well known. Given the modest benefits of analgesics and their safety concerns, novel nonpharmacological options are needed to reduce clinician and patient reliance on analgesics and to reduce the enormous burden of chronic pain on patients.

We selected an active comparator (music listening) to inform whether the design of the subsequent RCT should be a comparative effectiveness trial. We also designed FAMILIA to include a usual care control rather than other comparator options (eg, wait-list control or attention control). A usual care control group is appropriate given the paucity of studies and limited efficacy data of music interventions for chronic pain, especially for virtually delivered interventions. The study sample was drawn from a single VA medical center and may not be representative of all patients with chronic musculoskeletal pain.

The long-term objective of this research team is to develop and test music interventions that improve chronic pain management and can be implemented in various clinical settings. Findings from FAMILIA will inform refinement of interventions and trial design in advance of conducting a fully powered trial to identify putative mechanisms and test efficacy.