Resolving the Burden of Low Back Pain in Military Service Members and Veterans (RESOLVE): Protocol for a Multisite Pragmatic Clinical Trial

Abstract

Background

Physical therapy (PT) is frequently used for the management of low back pain (LBP) within the US Departments of Defense (DOD) and Veterans Affairs (VA). However, variations in PT practice patterns and use of ineffective interventions lower the quality and increase the cost of care. Although adherence to the clinical practice guidelines (CPGs) can improve the outcomes and cost-effectiveness of LBP care, PT CPG adherence remains below 50%. The Resolving the Burden of Low Back Pain in Military Service Members and Veterans (RESOLVE) trial will evaluate the effectiveness of an active PT CPG implementation strategy using an education, audit, and feedback model for reducing pain, disability, medication use, and cost of LBP care within the DOD and VA health care systems.

Design

The RESOLVE trial will include 3,300 to 7,260 patients with LBP across three DOD and two VA medical facilities using a stepped-wedge study design. An education, audit, and feedback model will be used to encourage physical therapists to better adhere to the PT CPG recommendations. The Oswestry Disability Index and the Defense and Veterans Pain Rating Scale will be used as primary outcomes. Secondary outcomes will include the LBP-related medication use, medical resource utilization, and biopsychosocial predictors of outcomes. Statistical analyses will be based on the intention-to-treat principle and will use linear mixed models to compare treatment conditions and examine the interactions between treatment and subgrouping status (e.g., limb loss).

Summary

The RESOLVE trial will provide a pragmatic approach to evaluate whether better adherence to PT CPGs can reduce pain, disability, medication use, and LBP care cost within the DOD and VA health care systems.

Background and Rationale

Low back pain (LBP) is one of the most frequently reported pain complaints in both deployed and nondeployed service members (SMs) and is among the leading causes of medical visits and lost duty days [1]. LBP is a primary reason SMs seek medical attention while they are deployed [2], and it is the most prevalent cause of interrupted combat duty [3, 4]. Once LBP develops in SMs, it is the single greatest risk factor for disability within 5 years of onset [5]. Given the direct impact of LBP on military readiness, implementing effective strategies to address it is a top strategic priority for the US Department of Defense (DOD) [6]. Similarly, LBP is also a frequent pain condition and a leading cause of disability within the Department of Veterans Affairs (VA) [7, 8]. LBP reduces functional capacity [9] and quality of life [10]. Polytrauma and blast injuries, specifically, increase the prevalence of chronic LBP among veterans [11, 12], particularly those with military-related injuries resulting in lower limb loss, where LBP prevalence rates approach 77% [13, 14]. This cohort is especially vulnerable to high rates of self-reported functional limitations, psychosocial conditions, and diminished quality of life [14, 15].

Physical therapy (PT) is a common nonpharmacological approach for the management of LBP and has been recommended by most clinical practice guidelines (CPGs) as the first line of treatment [16–20]. As such, results from systematic reviews of randomized clinical trials generally support the efficacy of PT interventions for reducing pain, disability, medication dependency, and cost of care for patients with LBP [21, 22]. However, variations in PT practice patterns [23, 24] and high utilization of ineffective interventions that are not supported by evidence can often lower the overall quality and increase the cost of LBP care [25, 26]. As such, adherence to the published CPG recommendations has been advocated as a way of standardizing treatment approaches to improve patient outcomes and reduce overutilization of other health care services [27, 28]. To this end, the American Physical Therapy Association (APTA) has published a specific CPG for management of LBP [17] that emphasizes active treatments (e.g., exercise), manual therapy, and self-care over passive interventions such as ultrasound and traction. However, the large number of possible PT treatment options available to patients and clinicians complicates the ability to standardize treatments. For example, although adherence to more active vs passive treatments has been associated with enhanced clinical outcomes and reduced health care costs, only 40–43% of patients appear to receive CPG-adherent PT care for LBP [28, 29].

Evidence suggests that successful treatment of LBP should include psychological assessments along with biological treatments to comprehensively address the patient’s unique pain experience [30]. The current APTA CPG for LBP supports the use of cognitive behavioral approaches to address beliefs and behaviors that may contribute to LBP chronicity and disability [17]. However, the results of a recent systematic review suggest that although physical therapists support biopsychosocial approaches, inadequate training and lack of individualized mentoring contribute to their low confidence in delivering psychological interventions [31]. Thus, psychologically informed PT (PIPT) has been suggested as a clinical framework for integrating cognitive behavioral approaches with the more standard PT practices based on biomedical principles [32]. The central theme of PIPT is the recognition that the psychological processes affecting the perception and response to pain are an expected and normal part of the pain experience [33]. PIPT focuses primarily on enhancing pain management skills and emphasizes the resumption of activities despite the presence of LBP [34]. Despite historical awareness of psychological risk factors and their impact on management of LBP [33], clinical implementation of PIPT strategies by physical therapists continues to be challenging [35].

There have been very few attempts to investigate and identify effective PT CPG implementation strategies. The use of printed educational materials and other passive implementation techniques only minimally influences adherence to CPGs [36]. To the contrary, multifaceted active interventions that include dissemination of best evidence by opinion leaders [37], audit and feedback [38], and educational meetings and workshops [39] appear to better facilitate the movement of evidence-based knowledge into clinical PT practice. Currently, it remains unclear whether incorporating an active implementation strategy for PT CPG adherence will ultimately improve outcomes while reducing the cost of care for patients with LBP [40].

The main objective of the Resolving the Burden of Low Back Pain in Military Service Members and Veterans (RESOLVE) trial is to evaluate the effectiveness of an active CPG adherence strategy using an education, audit, and feedback model with specific training in the APTA CPG and PIPT (CPG + PIPT) group as compared with usual care (UC) for reducing pain and disability in SMs and veterans receiving PT care for LBP. The second objective of the study is to compare the analgesic medication use and medical resource utilization between CPG + PIPT and UC over 12 months. Finally, the RESOLVE trial will aim to identify the predictors of clinical benefits, including patient sociodemographic variables (e.g., age, gender, race), medical history characteristics (e.g., limb loss, depression) for SMs and veterans receiving PT care for LBP, and the demographics and training history of the treating physical therapists.

Methods

Study Design

The RESOLVE trial is a pragmatic multisite stepped-wedge clinical trial (Figure 1). Inclusion of eligible patients into the study will occur over 22 months. Eligible LBP patients receiving PT care will complete validated outcome measures as part of the standard of PT practice. This information will be used by the physical therapists for clinical decision-making and will be used retrospectively for research purposes. In addition to the clinical patient information, the physical therapists’ CPG adherence rates will also be monitored using data from the electronic health records. Data collected from the patients and physical therapists during the prerandomization period will constitute the UC condition. Using a predetermined randomization order, each clinical site will be entered into the active phase of the trial (CPG + PIPT) in a stepwise fashion. Once randomized, physical therapists will receive education in APTA LBP CPGs and PIPT as well as monthly feedback reports (“audit and feedback”) on their CPG adherence rates and patient outcomes until the conclusion of the study period.

Figure 1.

The stepped-wedge study design for the RESOLVE trial. RESOLVE = Resolving the Burden of Low Back Pain in Military Service Members and Veterans; U_C = usual care monitoring period (physical therapist training will occur at the end of this period); X_T = physical therapist training completed and audits for physical therapists’ feedback reports begin; X_F = feedback period.

Scientific Rationale for Study Design

The stepped-wedge study design for the RESOLVE trial was selected in consultation with the National Institutes of Health (NIH)–DOD–VA Pain Management Collaboratory’s Biostatistics and Study Design Work Group [41]. The stepped-wedge trial is a study design that is increasingly used in the evaluation of service delivery–type interventions [42]. The design involves a random sequential crossover of clusters from control to intervention until all clusters have received the intervention by the end of the trial [42]. The stepped-wedge design is a pragmatic study approach that can reconcile the need for robust evaluations with political or logistical constraints, and it is particularly suited for evaluations that do not rely on individual patient recruitment. Unlike cluster-randomized trials, stepped-wedge trials without individual recruitment and without concealment of allocation or blinding of the intervention are at a lower risk of selection biases and treatment contamination. Given the focus on a pragmatic design, the RESOLVE trial intentionally incorporates version 2 of the Pragmatic-Explanatory Continuum Indicator Summary (PRECIS-2) [43] to achieve the highest level of pragmatism (Figure 2).

Figure 2.

The PRECIS wheel consisting of individual scores for the nine domains of pragmatism for the RESOLVE trial. The PRECIS wheel can visually represent how explanatory or pragmatic a trial is on the pragmatic to explanatory continuum. Trials that take an explanatory approach produce wheels nearer the hub; those with a pragmatic approach are closer to the rim. PRECIS = Pragmatic-Explanatory Continuum Indicator Summary; RESOLVE = Resolving the Burden of Low Back Pain in Military Service Members and Veterans. Adapted by permission from BMJ Publishing Group Limited. [The PRECIS-2 tool: designing trials that are fit for purpose, Loudon K, Treweek S, Sullivan F, Donnan P, Thorpe KE, Zwarenstein M. BMJ 2015;350:h2147]

Participating Sites

PT clinics from the Naval Medical Center San Diego, the Marine Corps Air Station Miramar/Naval Base San Diego, and the Walter Reed National Military Medical Center will constitute the three participating DOD sites. Additionally, the participating VA sites will include the James A. Haley VA Medical Center and the VA New Jersey Health Care System. The selection of these facilities provides access to a large number of SMs and veterans receiving PT care for LBP, especially those with limb loss. Inclusion of these clinical sites also provides a high level of pragmatism for the PRECIS-2 organizational domain, as the intervention delivery is slotted into the UC and uses existing health care staff and resources [43].

Study Population

The RESOLVE trial will use two simple and broad inclusion criteria: 1) participants must be 18–64 years old and 2) participants must be receiving PT care for LBP. The lower age limit of 18 years is based on military entrance requirements, and the upper boundary was selected due to the confounding effects of Medicare services used outside the DOD and VA networks for individuals over the age of 64 years. In keeping with the pragmatic nature of the trial and the PRECIS-2 domain of eligibility [43], the inclusion criteria will allow inclusion of women who are pregnant and people with nonsevere compressive nerve root signs, both of which are typical exclusion criteria in studies of LBP. The exclusion criteria for the study will be limited to medical contraindications to PT care, such as signs and symptoms of potentially serious conditions that would lead to patient safety concerns.

Screening and Recruitment Procedures

Data from patients will be primarily used for clinical decision-making and will be used retrospectively for research purposes. Therefore, the RESOLVE trial will not require consent for enrollment nor an active recruitment and retention strategy for patients. The study team will monitor the number of patients receiving PT care for LBP within a given time period as well as those who completed the primary outcomes at baseline. If the enrollment rate at a site is low, the study team will work with the site to troubleshoot any challenges. The participating physical therapists at each site, however, will need to provide informed consent because their demographics and experience information will be evaluated as predictors of the clinical outcome. Only patients treated for LBP by a consenting physical therapist will be included in the final analysis, and the entirety of the sample will reflect the gender, race, and ethnicity proportions of the population being served at each clinical site, providing a high level of pragmatism for the PRECIS-2 recruitment domain [43].

Interventions

Physical Therapist Education

Twelve hours of curriculum (Supplementary Data) will be delivered during a 2-day continuing education course, which is consistent in length with courses regularly attended by physical therapists to maintain and develop clinical skills. As such, the RESOLVE training program is highly pragmatic per the PRECIS-2 domain of intervention delivery [43]. The educational sessions will be led by the DOD, the VA, and academic opinion leaders, and the content of the training sessions will focus on dissemination of specific components of APTA LBP CPGs, such as active treatment strategies and PIPT. The educational sessions will be supplemented with short video interviews with the DOD, the VA, and academic opinion leaders on the topics of LBP, PIPT, and guideline-adherent LBP care. Additionally, the training material will include online educational modules focused on optimizing the experiential nature of the interactive in-person workshops [44]. All therapists will also have the option to participate in 1-hour quarterly educational outreach webinars geared toward providing updates on the evidence and problem-solving regarding the use of CPGs and PIPT.

Provider Audit and Feedback Procedures

After randomization into the CPG + PIPT phase, each physical therapist will receive regular monthly feedback reports on their CPG-adherent intervention delivery and changes in their patients’ disability and pain scores over the past 30 days. A number of clinical care quality indicators, such as rates of CPG-supported interventions delivered during the episode of PT care and changes in LBP-related patient outcomes, will be monitored and reported back. Treatment billing codes for the entire episode of PT care will be examined, and the total number of CPG-supported codes (i.e., exercise, manual therapy, and patient education) and those not supported by the CPGs (e.g., ultrasound, traction, and so on) will be calculated. To determine the percentage of each code category, the total number of Current Procedural Terminology codes for a given category (i.e., exercise, manual therapy, patient education, passive modalities) during the episode of care will be divided by the sum total of all intervention codes and will then be presented as a percentage of total care delivered. The proposed performance evaluation and quality control are highly pragmatic per the PRECIS-2 adherence domain, as they include no special measures to specifically enforce engagement or compliance and are consistent with the standard of care at most DOD and VA facilities [43].

Primary Outcomes

The primary outcomes selected for the RESOLVE trial are meant to optimize relevance to patients, providers, and health care systems, thus meeting the pragmatism requirements outlined in the PRECIS-2 primary outcome domain [43]. To this end, the Oswestry Disability Index (ODI) and the Defense and Veterans Pain Rating Scale (DVPRS) were selected as the co-primary outcome measures for disability and pain intensity in patients with LBP. The ODI is a condition-specific, reliable, and valid legacy measure of functional status widely used in LBP research [45]. The DVPRS was developed to implement a standardized, reliable, and valid DOD and VA approach to pain management for SMs and veterans. The DVPRS uses a numerical rating scale enhanced by functional word descriptors, color coding, and pictorial facial expressions matched to pain levels [46]. The co-primary outcomes of the ODI and DVPRS will be collected at the time of initial PT evaluation and every 2 weeks thereafter, with the 4-week or discharge score (whichever comes first) as the primary outcome time point.

Secondary Outcomes

The RESOLVE trial will use DOD and VA electronic health records to quantify medication and medical resource utilization. Prescription patterns for key drug classes commonly used for LBP management will include nonsteroidal anti-inflammatory drugs, opioid analgesics, skeletal muscle relaxants, antidepressants, and drugs for neuropathic pain. The medical resources of interest will include 1) referrals to LBP specialists and other health care providers and 2) LBP diagnostic tests and surgical procedures. All of the aforementioned secondary outcomes (prescriptions, referrals, orders) will be categorized as present or absent and will be treated as dichotomous in all analyses. PT and other provider encounters and procedures will be identified through the DOD Medical Data Repository and the VA Corporate Data Warehouse. Surveillance of the LBP-related medication and medical resource utilization will be conducted for 12 months after completion of the initial episode of PT care. The RESOLVE trial is considered highly pragmatic per the PRECIS-2 primary follow-up domain, as it has no additional data collection or follow-up than is normal in UC [43].

Additional Measures

Demographics and patient characteristics will be recorded as potential predictors of treatment outcome and to describe the sample as required for the NIH Minimal Dataset [47]. The nine-item Subgroups for Targeted Treatment (STarT) Back Screening Tool (SBT) will be used to categorize patients’ risk for psychosocial prognostic factors [48]. The SBT has demonstrated adequate test-retest reliability for its total scores and for its psychosocial subscale [48] and can provide important prognostic information for subgrouping patients with LBP on the basis of their modifiable risk factors [49]. Physical therapists’ demographics and training characteristics such as age, gender, years of experience, and specialty clinical certifications will also be collected to assess their influence on CPG adherence and patient outcomes.

Statistical Methods

All analyses will use an intention-to-treat approach, and results will be reported using the Consolidated Standards of Reporting Trials (CONSORT) extension to cluster-randomized trials [50]. Additionally, all available data will be used for primary analysis from the start of the trial, during the training and transition periods, and until the end of the trial. This approach is highly pragmatic per the PRECIS-2 primary analysis domain, as it uses all available data and includes no special allowance for nonadherence, practice variability, and other factors [43]. To test the effect of the intervention on clinical outcomes as measured by the ODI and DVPRS, the analysis will use a linear mixed model with fixed effects for intervention and time, accounting for the clustering trial design with random clinic effects, correlation among patients within a provider with random provider effects, and correlation among two to three repeated observations during the treatment period from the same patient using random patient effects. Site differences in patient populations will be controlled for in the model with the random site effect. In addition, the intervention effect is estimated from within-site changes as well as the between-site difference, which lessens the bias that could occur with a purely parallel cluster-randomized trial imbalanced on patient populations. Primary analyses will also adjust for baseline ODI and DVPRS scores. Analyses of analgesic medication use and medical resource utilization will use generalized linear mixed models, with the distribution determined by the type of data while controlling for the stepped-wedge cluster-randomized trial design by having random effects for sites and providers. For the predictors of clinical benefit, an unadjusted analysis will be estimated using univariate general linear mixed models for each potential predictor, controlling for clustering. Baseline measures associated with outcomes at the P<0.15 level in unadjusted models will be added to multivariable models using a hierarchical modeling approach and backward elimination.

Sample Size Determination

Based on preliminary data, approximately 30 to 66 new patients per month at each clinical site (5–11 physical therapists) should be available. The effective sample size is between 13 and 30, assuming the correlation among patients from the same physical therapist is high (ρ=0.2) and assuming 10% attrition. With five clinical sites in a complete stepped-wedge cluster-randomized design, with 22 months (including 2 months for baseline, delayed effects for 2 months [0.5 design effects], and a 1-month lag for starting the next site) and five steps, the RESOLVE trial will have 80–90% power to detect a medium effect size of 0.33. The participating site Intraclass Correlation Coefficient is assumed to be 0.05. Per clinic, across the 22 months, the sample size will range between 660 and 1,452 for a total sample size ranging between 3,300 and 7,260.

Impact of COVID-19 on Study Design

The potential impact of the COVID-19 Public Health Emergency on clinical pain management and pain research continues to evolve [51]. During the first wave of COVID-19 disease transmission, physical therapists were encouraged to adopt virtual health visits to safely provide essential rehabilitation services for their patients. It is anticipated that virtual care delivery will transform PT practice through increased use of virtual communication–based services [52]. Virtual delivery of PT for LBP, however, has the potential to drastically change adherence to the published LBP CPG recommendations as the list of available interventions is reduced to only those that can be delivered virtually. For example, the use of CPG-supported active treatments (e.g., exercise and self-care) could be increased, whereas the use of passive interventions (e.g., ultrasound and traction) that are not supported by the LBP CPGs and cannot be delivered virtually could be reduced. To account for the influence of this potential change in PT practice patterns, we will closely monitor the patterns of virtual vs the more traditional in-person PT care delivery during the implementation of our study and will account for such changes (if they are in fact present) in our statistical analysis.

Another potential influence of COVID-19 on our trial could be related restrictions on large gatherings, which could influence the delivery of the education curriculum earmarked for our providers. The educational curriculum was originally designed to be delivered during a 2-day live continuing education course. However, live continuing education courses are now being converted to a remote delivery format due to the COVID-19 crisis. As such, the educational training for the RESOLVE trial will be converted to a synchronous online format that will take advantage of the advanced features of remote education, including virtual breakout sessions that incorporate small-group workshops and role-playing exercises. The RESOLVE investigators have already successfully presented other continuing education courses with similar course content in an online synchronous format and have experience with managing “remote classrooms.” Physical therapists are also becoming more accustomed to the emerging trend of continuing education courses moving to an online synchronous format.

Discussion

The RESOLVE trial will address current gaps in the care of patients with LBP in the DOD and the VA and will provide the framework necessary for widespread integration of findings within these health care systems. This pragmatic trial has been specifically designed to incorporate existing training options used by physical therapists to deliver a strategy that can be broadly applied to real-world clinical settings with minimal financial and time investment on the part of the individual clinics and clinicians. This trial will also use CPG-recommended and rigorously validated patient-reported outcome measures (disability and pain) relevant to patients and providers, as well as prescription medication use and health care utilization relevant to the DOD and VA health care systems. In aggregate, the findings from the RESOLVE trial will provide evidence that can improve clinical decision-making for patients with LBP and will have a sustained impact on how PT care for LBP is delivered throughout the DOD and the VA.

Supplementary Data

Supplementary Data may be found online at http://painmedicine.oxfordjournals.org.

 

Funding sources: This study was approved by the Naval Medical Center San Diego Institutional Review Board (NMCSD.2018.0034). The authors certify that they have no affiliations with or financial involvement in any organization or entity with a direct financial interest in the subject matter or materials discussed in the article. The US Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick, MD 21702-5014 is the awarding and administering acquisition office. This work was supported by the Assistant Secretary of Defense for Health Affairs endorsed by the Department of Defense through the National Institutes of Health–Department of Defense–Department of Veterans Affairs Pain Management Collaboratory Pragmatic Clinical Trials Demonstration Projects under Award No. W81XWH-18–2-0007. For more information about the Collaboratory, visit https://painmanagementcollaboratory.org/. This research was also supported by the National Center for Complementary and Integrative Health of the National Institutes of Health under Award No. U24AT009769.

Conflicts of Interests: There are no conflicts of interest to report.

Supplement sponsorship: This article appears as part of the supplement entitled “NIH-DOD-VA Pain Management Collaboratory (PMC)”. This supplement was made possible by Grant Number U24 AT009769 from the National Center for Complementary and Integrative Health (NCCIH), and the Office of Behavioral and Social Sciences Research (OBSSR). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NCCIH, OBSSR, and the National Institutes of Health.

Disclaimer: The views expressed in this manuscript are those of the authors and do not necessarily reflect the official policy of the Uniformed Services University of the Health Sciences, Departments of the Air Force, Army, Navy, Defense, or Veterans Affairs, nor the US government. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Copyright statement: I am a military service member or employee of the US government. This work was prepared as part of my official duties. Title 17, U.S.C., §105 provides that copyright protection under this title is not available for any work of the US government. Title 17, U.S.C., §101 defines a US government work as a work prepared by a military service member or employee of the US government as part of that person’s official duties.