Implementation of Psychologically Informed Physical Therapy for Low Back Pain: Where Do We Stand, Where Do We Go?

Introduction

There has been substantial growth in interest in biopsychosocial treatments for patients with musculoskeletal pain.¹ One reason for this increased interest is the American College of Physicians Clinical Practice Guidelines and the Centers for Disease Control and Prevention Guidelines for treatment of chronic pain.^2,3 Additionally, the revised International Association for the Study of Pain (IASP) definition for pain focuses on the influence of biological, psychological, and social factors on a person’s pain experience.⁴ This focus provides continued opportunities for non-pharmacological pain treatments that use psychological and behavioral components in the management of musculoskeletal pain.^2,3

Physical therapists have incorporated psychologically informed practice (PiP) into their practice patterns for low back pain management. PiP is a novel approach that integrates physical, behavioral, and psychological interventions with an overall goal of reducing pain-related disability.⁵ Since PiP was originally proposed as a treatment framework; there has been an evolution and refinement in practice approaches for low back pain. PiP was targeted towards secondary prevention and subsequently focused treatment goals on reduction of disability. Recently, for low back pain, there has been a shift to focus more on whole person or positive health targets as patient outcomes.⁶

PiP is well positioned to be responsive to this shift away from secondary outcomes to a broader consideration of patient health. For example, key features of a PiP framework are well aligned with enhancing whole person health. These key features include addressing a patient’s behavioral response to pain through any of the five major types of PiP interventions: 1) educational (threat reduction and activation; 2) behavioral (adapting behaviors in response to pain); 3) cognitive-behavioral (cognition and coping skills); 4) psychophysiological (stress reduction and mindfulness); and 5) contextual cognitive-behavioral (Acceptance and Commitment Therapy).⁷ The actual content of a PiP intervention varies and may best be thought of as a framework for therapies that integrate contextual, behavioral, and physical treatments.⁸ Psychologically informed physical therapy (PIPT) then is viewed as a subset of PiP when PiP is delivered by physical therapists. Therefore, in this paper when PiP is used it is a general reference to the treatment framework. When PIPT is used, it is referring to treatment delivery by physical therapists. Specifically, PIPT is an integration of traditional physical treatments (eg, spinal manipulation or exercise) with cognitive-behavioral treatments (eg, pain coping skills or progressive relaxation).

In this paper, we focus on PIPT. We fully acknowledge that other providers (eg, nurses, chiropractors, occupational therapists) have PiP models for delivery of treatments for chronic conditions. However, our implementation experiences primarily involve physical therapists, and we therefore provide insight based on our expertise. We acknowledge the important roles of other types of providers in PiP, but that is not the focus of this paper. PIPT can benefit patients by individualizing physical and psychological interventions to improve functional activities that had been limited due to pain. Although it is beyond the intended scope of this Perspective to exhaustively review PIPT, key treatment components include Realization, Relief, Regulation, Re-activation, and Reinforcement (5 R’s).⁹ These core components are drawn from conceptual models that have been used in mental health and then applied in an integrated fashion (ie, with treatment of physical impairments) for pain clinic and rehabilitation settings.^7,10,11 The conceptual models used in these clinical settings include the fear avoidance model, the ACT “hexaflex”, and the cognitive behavioral model.^12–14 This “5 R approach” to PIPT, including convergent examples for each of these conceptual models, are summarized in Table 1.

Table 1 Key Components Comprising Psychologically Informed Practice (PIPT)

There is growing evidence demonstrating the effectiveness of PIPT approaches for pain-related conditions. Early supporting evidence came from the United Kingdom, where a stratified care approach was investigated for managing patients with low back pain in a randomized controlled trial.¹⁵ Low-risk patients were given minimal treatment, medium risk patients were referred for traditional physical therapy, and high-risk patients received psychologically informed physical therapy. Results showed that the stratified care intervention group had favorable outcomes for disability scores, health quality of life, and cost savings. A trial by Beneciuk and George¹⁶ reported on the pragmatic application of risk stratification in outpatient physical therapy practices and found that physical therapists who were randomly selected to receive stratified care training had increased biopsychosocial treatment orientations and patients who received care from these clinicians had improved pain and disability scores. These trials demonstrated that physical therapists can effectively screen and deliver psychologically informed care which resulted in better patient outcomes.

Evidence from individual randomized controlled trials has been reported for synthesis of PIPT approaches for musculoskeletal pain. For example, Hall et al conducted a systematic review that assessed the effectiveness of physical therapist-delivered cognitive behavioral interventions for low back pain.¹⁷ Five randomized controlled trials (n = 1390) were identified and meta-analysis revealed that cognitive-behavioral interventions had a greater effect on reduction of disability and pain than education or exercise interventions.¹⁷ Similarly, Archer et al conducted another review that investigated the use of psychologically informed physical therapy for different types of musculoskeletal pain and identified eight randomized controlled trials (n = 978).¹⁸ They found that PIPT is likely effective for low back pain, especially when it is targeted toward patients who are at high risk of poor outcomes due to fear of movement.¹⁸ Two years later, Coronado et al identified 18 randomized controlled trials (n = 2387) that used PIPT approaches for patients with musculoskeletal pain and found that PiP interventions had a greater effect on reduction of disability and pain than education or exercise-only interventions.^17,19

A notable observation for this evidence base is that explanatory clinical trials investigating PIPT for low back pain have a tendency to show favorable results; however, pragmatic trials have shown less promising outcomes. For example, the TARGET trial (77 primary care clinics, n = 2300 enrolled patients) investigated whether a risk-stratified approach would result in lower rates of chronic low back pain and improved self-reported disability. Patients in the high-risk group were referred to PIPT or usual care. At the conclusion of the trial, patients in the PIPT intervention group showed no differences in disability, care utilization, or transition to chronic low back pain.²⁰

An earlier study, the MATCH trial (6 primary care clinics, n = 603 enrolled patients), investigated whether use of the STarT Back risk-stratification strategy for patients with low back pain resulted in superior outcomes for physical function and pain severity.²¹ While clinicians used the STarT Back tool with half of participants, use of the tool did not change recommended treatments nor did the intervention have any significant effects on patient outcomes.²¹

Common themes for both the TARGET and MATCH trials are that they were conducted in the United States and reported implementation challenges for PIPT approaches. In TARGET, the authors report a significant “implementation shortfall” with the PIPT referral process resulting in only half of the patients in the stratified care group receiving a PIPT referral.²⁰ This implementation shortfall was attributed to electronic health record challenges with the referral process to PIPT clinicians, as well as a lack of study-funded resources for physical therapy visits.²⁰ Similarly, in MATCH, the authors discuss implementation difficulties due to the complexity of the PIPT intervention, as well as the nuances of the US healthcare system, when compared to the U.K.²¹

Accordingly, this perspective addresses issues of PIPT implementation within US health systems. The fact that the TARGET and MATCH trials had challenges implementing this type of intervention in a pragmatic trial is concerning for those interested in implementing PIPT in real-world clinical settings. One challenge limiting uptake of PIPT in care settings is that, in general, non-pharmacological interventions can be challenging for health systems to deliver, when compared to medications or surgery.^1,22 Clinician attitudes and beliefs about using behavioral and psychological interventions can also influence their willingness to use PiP approaches since they are still not commonly taught in pre-professional training for many provider types (eg, physical therapists).^23,24 Other challenges to PIPT implementation include physical therapists feeling unprepared to treat patients with complex pain conditions and a fear of not meeting patient’s expectations for carrying out traditional physical therapy care.²⁵

Comprehensive training in implementing PiP approaches for physical therapists would address some of the challenges of real-world adoption; however, most PIPT training has been completed in support of conducting clinical trials. Scant evidence exists on how to scale PIPT training so that it can be delivered for practicing physical therapists.¹⁹ For example, the ideal training mode, intensity, or duration for learning and implementing PIPT interventions within real-world clinical settings is unknown. Indeed, the recent clinical trial experiences from the US provide an indication that scaling PIPT to meet the demands of routine clinical delivery is primarily implementation related.^20,21

Therefore, the purpose of this Perspective is twofold. First, we discuss the current state of PIPT training and implementation. Second, we propose a roadmap for future implementation needs related to increasing delivery of PiP-informed approaches by physical therapists. Within this perspective, we have organized these into two concepts: Concept 1 – Current state of PIPT implementation and Concept 2 – Describe an implementation roadmap for PIPT.

Concept 1: Current State of PIPT Implementation

Most of the implementation of PiP approaches for physical therapists has been done in preparation for conducting clinical trials. In these instances, physical therapists are trained in a specific PiP intervention (eg, Pain Coping Skills Training,²⁶ Acceptance and Commitment Therapy²⁷) and use those skills to treat trial participants with a given pain diagnosis. These trainings consist of different combinations of didactic, experiential, and mentoring components. Examples of how PIPT trainings have been delivered for selected clinical trials are shown in Table 2. In the selected trials, training episodes were between one and six days and included didactic and experiential learning. Three of the five PIPT trainings also had ongoing mentoring and feedback components. Two manualized trials for PIPT also used fidelity checklists to assess physical therapist competence and adherence.

Table 2 PIPT Training Examples for Licensed Physical Therapists

One advantage of training for explanatory trials is the ability to focus on treatment fidelity through many different means, including protocol standardization and a dedicated provider pool. If PIPT interventions do not demonstrate positive effects in such highly controlled environments, then they are not likely to be effective in an environment with fewer resources and less time dedicated towards delivering the intervention. However, there are also disadvantages with this method of training. Explanatory trials can take a long time to complete and are expensive. Typically, clinicians who are trained in PIPT interventions are paid research staff. After trial completion, the clinic or health system no longer has access to those specially trained providers unless they have the resources to hire additional personnel.

Embedded pragmatic trials are one potential solution to the sustainability concerns of an explanatory trial. In an embedded pragmatic trial, clinicians who are integrated within an existing clinic or health system are trained to deliver the PIPT interventions used in the study protocol. This use of existing clinical resources can be valuable for increasing the likelihood of long-term sustainability of the PIPT interventions. Additionally, like PIPT training for explanatory trials, training for pragmatic trials is highly standardized. However, variation from the treatment protocol is more acceptable (and even expected) due to the practical focus.

There is convergence around core components of a PIPT training program (ie, didactic content, experiential learning, and mentoring with feedback) as shown by the different training protocols from Table 2. However, there is no consensus on how these training programs should be optimally delivered. For example, we do not know the most effective intensity of training delivery or how to increase uptake of PIPT practice into an existing health system. This lack of guidance for implementation may help to explain the small treatment effects of PIPT interventions in recent trials that were more pragmatic in nature and could be attributed to undertraining or not appropriately stratifying care.^19,28 Further research into the implementation of PIPT training could elucidate effective combinations of training modes and implementation strategies within specific contexts (academic medical center, private clinic, home health, etc.), as well as for specific patient populations.

Table 2 indicates where PIPT implementation has been focused to-date. However, as clinical trial results have been published, it is clear that integrating PIPT approaches into the US healthcare system is not a simple task.^20,21 The strategies that have been used to implement PIPT approaches for clinical trials in other countries have not generalized to effective real-world delivery for trials embedded in the US healthcare system. This issue may not be specific to PIPT, as other management approaches for low back pain have struggled with real-world adoption.^29,30

Concept 2: Road Map for PIPT Implementation

One way to make the leap toward real-world delivery of behavioral interventions has been to move from clinical trials to implementation-focused studies.^31–33 Brown et al have created a dissemination and implementation diagram to increase real-world relevance and adoption of evidence-based interventions (Figure 1).³⁴ The pre-intervention phase often consists of theory development and pilot studies to investigate if an intervention is feasible to carry out. For PIPT training approaches, there are examples available to guide recommendations for the future.

Figure 1 Dissemination and implementation framework. Notes: Reprinted from Brown CH, Curran G, Palinkas LA et al. An Overview of Research and Evaluation Designs for Dissemination and Implementation. Annu Rev Public Health. 2017;38:1–22. Copyright © 2017 Annual Reviews. This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 (CC-BY-SA).34

For pre-licensure physical therapists, Ballengee et al have proposed a training approach that is similar to the one used for practicing clinicians with an additional focus on foundational pain education and patient-centered communication.³⁵ This training method addresses the call for pre-licensure training to go beyond didactic content and provide more experiential training to address barriers with incorporating psychologically informed interventions.^36,37 The training focuses on addressing the complex nuances of behavioral-based non-pharmacologic pain treatments like building therapeutic alliance, reducing the perceived threat of pain, conceptualizing pain beliefs, and promoting self-efficacy.³⁸ Pre-licensure trainings like this one are being introduced throughout Doctor of Physical Therapy curriculums throughout the U.S.³⁹ One noteworthy limitation of addressing PIPT implementation challenges through this type of training is the potential lack of reinforcement upon graduation if the PT is not working in an environment with PIPT-specific mentoring.

Examples of real-world adoption of PIPT training for licensed physical therapists have been shown in the United States. Keefe et al have developed a PIPT training program for practicing physical therapists that consists of a PIPT treatment manual, experiential workshops, and ongoing supervision with consultation and feedback.⁷ This multicomponent training approach gives clinicians the opportunity to build their knowledge and PIPT skills while also addressing treatment fidelity. This approach has been carried out within the Duke University Health Care System with 95 clinicians (80 physical therapists, 15 occupational therapists) being trained to-date. The program works with a group of 3–5 clinicians at a time and consists of 90 minute didactic and experiential sessions for 15 weeks with a pain psychologist. This has been implemented as a quality improvement initiative funded by the hospital system and outcomes are currently being evaluated on a clinician and system-wide level.

Another quality improvement initiative is taking place within the Department of Veteran’s Affairs Health System (VA) to train physical therapists in PiP approaches. This approach uses physical therapists who have already been trained in PIPT interventions to train other VA physical therapists (train-the-trainer approach) who are interested in learning components of the International Association for the Study of Pain (IASP) pain education curriculum.⁴⁰ The VA curriculum consists of didactic, experiential, and mentoring components. This quality improvement project will also evaluate the implementation strategies needed for wide-scale adoption after two pilot clinics go through the training. This initiative coincides with the American Physical Therapy Association’s (APTA) call to standardize pain education and advance therapeutic pain communication for physical therapists.⁴¹ Notably, both Duke University Health System and the Department of Veterans Affairs are well-resourced settings that may not be indicative of all practice settings across the United States.

Moving Towards Wide-Scale Implementation

While there are some examples of local attempts to implement PIPT approaches, there is no over-arching guidance on the most effective way to increase real-world clinical delivery. Figure 2 illustrates a conceptual PIPT Implementation Roadmap adapted from the Brown et al diagram that could bridge the gap between training for trials and real-world adoption by focusing on “Implementation Zones” between the different stages.³⁴ To-date, no implementation-focused studies like those suggested in Figure 2 have been done for PIPT interventions. One way to conceptualize implementation studies for PIPT would be to consider different types of research questions that would align with the Implementation Zones from Figure 2. Curran et al proposed three hybrid effectiveness-implementation study designs that could serve as a staged approach to move PIPT beyond pure efficacy trials.^42,43 Table 3 illustrates the three hybrid designs with corresponding PIPT-focused aims and research questions.

Table 3 Hybrid Effectiveness-Implementation Study Designs with Corresponding PIPT Research Questions, Clinical Outcomes, and Implementation Outcomes43

Figure 2 PIPT implementation roadmap.

Determining PIPT Effectiveness and Approaches for Implementation

An example of a Type 1 hybrid design for a PIPT study would be to investigate whether a PIPT approach resulted in improved functional outcomes for patients with low back pain in an outpatient physical therapy clinic. While the focus of a Type 1 study is on PIPT effectiveness, the study investigators would also gather information about barriers and facilitators to implementing the PIPT approaches to inform future studies. In a Type 2 hybrid design, the focus of the study would be on both the effectiveness of the PIPT approach on functional outcomes for low back pain patients, and on the feasibility of using a particular implementation intervention or strategy (eg, in-person, PIPT training). The unit of randomization in a Type 1 study would likely be the patient, while in a Type 2 study, patients would often be randomized for the clinical effectiveness aim and the provider or clinical unit would often be the unit of randomization for the implementation aim.

Determining Best Approach to Implement PIPT

In a Type 3 implementation study, researchers could investigate whether one implementation strategy is superior to another based on a pre-determined set of implementation outcomes. An example of a Type 3 PIPT implementation study could be to start with an existing PIPT curriculum that was been used for training clinicians in a pragmatic trial. To study the most effective way of implementing that PIPT curriculum within a health system, a Type 3 study could be conducted to compare an audit and feedback strategy against academic detailing and evaluate which strategy produced superior implementation outcomes.⁴⁴

Another Type 3 option might be to investigate implementing PIPT interventions in a stepped-wedge design where all sites receive a low-level implementation strategy such as a one-day course with a manual (academic detailing) and then evaluate where those sites are with their implementation outcomes at a given time point. Those sites that have not made sufficient progress would then be “stepped-up” to receive a higher level (ie, more resource-intensive) implementation strategy such as implementation facilitation by a trained facilitator.⁴⁴

Other Issues That Could Impact PIPT Implementation

Barriers to PIPT implementation should be considered at multiple levels (eg, individual, clinic-level, health-system, and beyond).^45,46 For example, in pragmatic trials in the US, a lack of primary care referrals have been observed as a barrier to patients receiving PIPT. Several of the challenges with PIPT implementation (ie, no psychosocial risk screening, lack of PT referral) start at the primary care visit where many patients do not have the opportunity to access psychologically informed care.^47–49 Instituting routine screening in primary care settings with automatic referral to PIPT could limit this barrier. However, such a process could create other problems (ie, not enough providers to deliver treatment, increase risk of overtreatment). Future research should investigate the impact of different referral models on the percentage of those actually receiving PIPT.

Another example of limiting PIPT implementation is existing policies that limit reimbursement for non-pharmacological pain treatments. Therefore, policy-related changes could be used to increase reimbursement for non-pharmacological management of low back pain.⁵⁰ However, the upfront resource investment (eg, time, money, personnel) required to implement non-pharmacological treatments, may be too much for a health system if there are no reimbursement incentives in place to offset those costs. One solution may be to conduct implementation-focused studies that carefully consider context, including current clinical workflow, to optimize delivery approaches, elucidate the most effective ways to implement PIPT in a given setting, and therefore optimize resource use – all with the goal of providing the best clinical care with the highest return on investment.

Another policy-related change could be to alter the way funding can be used for a trial. Often, funding policies are geared towards the start-up and conduct of the trial. These funds are used in study start-up for confirming the protocol, hiring study staff, completing training for intervention delivery, reimbursing participants, and setting up systems to collect outcome measures. During the conduct of the trial, funds are used to deliver the intervention, collect outcome measures, complete analyses to determine intervention effectiveness, and to disseminate trial findings. Typically, there are not sufficient funds after trial completion that would add to the sustainability of the intervention in the setting in which the trial was conducted. This type of funding would be especially appealing for interventions that demonstrate effectiveness during the trial, but it would require different funding mechanisms than are currently available.

Lastly, another important consideration for implementing PIPT approaches is context, or the set of unique circumstances and factors around implementation of PIPT (eg, staffing availability, culture around PT’s role, etc.). For example, each health-care setting has unique characteristics that could impact the success or failure of an implementation effort.^51,52 For non-pharmacologic approaches like PIPT, part of the contextual considerations that have impacted its uptake include provider beliefs.¹ On the positive side, physical therapists are known for their patient-centered approaches to care through the use of tools like shared-decision making and patient-centered communication.^53,54 However, physical therapy training is still heavily focused on biomedical approaches for treating musculoskeletal pain which could make adoption of PIPT more difficult.³⁹ Practicing physical therapists often have to self-teach or self-finance continuing education courses to learn about PiP approaches. Other factors related to physical therapist role clarity have been cited as potential barriers.⁵⁵ Some physical therapists report difficulty in getting buy-in from patients when using PiP approaches due to the perception that physical therapists only focus on physical activity and exercise.⁵⁶

Conclusion

Pain care in the United States has many improvements to make towards systematic implementation of effective, evidence-based, nonpharmacological treatments.^57,58 PiP delivered by physical therapists is one way to address this need, but while there is evidence to support its effectiveness for certain patient populations, evidence is lacking about the best way to implement these interventions into everyday clinical care. In this Perspective, we have presented a conceptual roadmap to move implementation of PIPT beyond the delivery of clinical trials and towards real-world adoption. This roadmap serves as guidance to increasing PIPT scalability and scope, which if fully realized could help improve routine delivery of guideline concordant pain care. Physical therapists can be the provider of choice for delivering PiP approaches, but only if the scale of training can be increased to meet existing needs.