Regional vs global physical therapy interventions to treat chronic pain in survivors of trauma: a randomized controlled trial

ABSTRACT

Background

A history of traumatic life events is associated with chronic pain in later life. Physical therapists utilize a variety of methods to treat pain, however, they have struggled to find effective interventions to improve patient outcomes.

Objective

To compare impairment-based, regional (REGION-PT) physical therapy (PT) to a global (GLOBAL-PT) model consisting of pain neuroscience education, graded motor imagery, and exercise for adults with chronic pain and history of trauma.

Design

Randomized Controlled Trial.

Methods

Adults ≥ 18 years of age with chronic pain and a history of ≥1 trauma identified through the Life Events Checklist received the allocated intervention once a week for six weeks. Treatment effects were assessed using linear mixed models.

Results

Ninety-eight participants completed the trial. There were no difference in outcomes between groups. There were significant interactions between race and intervention. Both interventions were associated with improvements in pain interference for white participants, but non-white participants experienced improvement only with GLOBAL-PT. Regardless of allocation, participants improved in physical function, six of the PROMIS-29 domains, and in pain interference measures.

Conclusion

Both interventions are reasonable strategies for individuals with chronic pain and a history of trauma.

Introduction

Twenty-five million Americans suffer from chronic pain, creating an urgent need to find interventions to help improve outcomes [1,2]. Pain treatment guidelines recommend a biopsychosocial multi-disciplinary approach, with prescription medications not being the focus of the intervention [3–5]. Multi-disciplinary pain centers are effective but costly, scarce, and often have long wait times [6]. Physical therapists have a variety of methods to treat acute pain, but have struggled to find effective interventions for chronic pain. An impairment-based model, focusing treatment to the region of the reported pain, may be appropriate for acute conditions. However, it appears limited for chronic pain, which may be driven by altered nervous system processing [5,7–9]. Guidelines for the treatment of chronic pain encourage active interventions that address biopsychosocial factors and focus on improvement in function [10]. Guidelines for chronic low back pain endorse the use of exercise and a range of other non-pharmacological therapies, alone and in combination, such as massage, acupuncture, spinal manipulation, Tai Chi, and yoga [10]. The guidelines are helpful if followed, but individuals living with chronic pain often demonstrate fear of moving due to concerns of causing or exacerbating the pain condition so a strong patient-clinician therapeutic alliance is also required.

Risk factors for chronic pain include a history of physical or emotional trauma [11]. Disparities in the treatment of women and nonwhite individuals with chronic pain may exacerbate symptoms [12]. Nonwhite individuals and women are more likely to report higher pain intensities yet have their pain intensities underestimated and undertreated compared to white male counterparts [12,13]. Women with a history of trauma report more somatic symptoms, and a co-diagnosis of depression compounds the somatic symptom severity [14]. The number of traumatic events experienced appears to increase the likelihood of chronic pain [11,14–17]. Post-traumatic stress disorder symptoms occur in 50% of individuals reporting chronic pain, and those who experienced sexual abuse exhibit lower pain pressure thresholds, greater temporal summation, a larger number of painful sites and increased disability when compared to controls [18]. Increased temporal summation and lowered pain pressure thresholds are indicators of possible changes in central nervous system processing [5,8,19–21].

Nociplastic pain characterizes discomfort arising from altered function of sensory pathways in the nervous system and can be accompanied by fatigue, sleep disturbances, hypersensitivity to external stimuli, and mood disturbances [22]. Predisposing factors of nociplastic pain include a history of psychological, sexual, emotional, or physical abuse. Multiple mechanisms contribute to the perception and maintenance of pain and can be driven by peripheral and central nervous system (CNS) causes, as well as by psychological issues. These mixed pain conditions have overlapping nociceptive, neuropathic, and nociplastic components [22]. Due to the coinciding contributors to chronic pain, biopsychosocial multimodal interventions targeting altered neural processing may be warranted, particularly for survivors of trauma [22].

Physical therapists embrace multimodal treatments in attempts to address the biopsychosocial complexities of patients with chronic pain conditions, and it has been suggested that assessment of combined treatments should be considered for clinical trials [23]. Conventional physical therapy (PT) used to treat chronic low back pain is said to include manual therapy, region specific strengthening exercises, and patient education regarding body mechanics, postural exercises, and lifting techniques [24,25]. This follows an impairment based, regional (REGION-PT) approach. Pain neuroscience education (PNE) and graded motor imagery (GMI) are PT interventions used to treat conditions such as chronic low back pain, fibromyalgia, phantom-limb pain, and complex regional pain syndrome, and are thought to target nervous system changes and nociplastic pain [7,26–34]. Educating patients about chronic pain mechanisms, the importance of physical activity, and how home and work stressors can impact the pain experience have been found to reduce catastrophization and fear of movement [28,35]. GMI is an intervention that attempts to address cortical alterations in the somatosensory and motor cortexes of the brain which are reported to be adversely altered in individuals with chronic pain conditions [21,32,33]. These interventions may better direct treatment to the cause of the chronic pain, and follow a global, biopsychosocial approach; however, many studies examining the effects of GMI and PNE have methodological issues [27,33,36,37]. Aerobic exercise and resistance training are beneficial in reducing fear avoidance behaviors, improving mood, and decreasing disability through possible central mechanisms, and follow a global approach; however, there are conflicting outcomes [38–40]. Although physical therapists can offer all of these modalities, there is a lack of high-quality evidence to make conclusive recommendations, particularly for individuals with chronic pain associated with trauma [27,40,41].

Due to the theory that chronic pain is driven by altered nervous system changes [7,22], we hypothesized that effective treatment must include interventions targeting the CNS pathology. We sought to compare a REGION-PT approach to a global (GLOBAL-PT) model using PNE, GMI, and exercise to address chronic pain dysfunction in survivors of trauma. We hypothesized that those who received GLOBAL-PT would have improved function and reduced pain interference as compared to those treated by REGION-PT.

Methods

Trial design

This single center single blinded parallel trial enrolled and randomized participants to one of the two treatment groups. The protocol was modified in 2020 due to the COVID-19 pandemic to include the option of delivering both interventions via telehealth. This trial was approved by the University of Vermont Committee on Human Research in the Medical Sciences (CHRMS #STUDY00000128). The trial was prospectively registered at ClinicalTrials.gov (NCT03933189) on 1 May 2019.

Participants

The participants were adults ≥18 years of age with chronic (≥6 months) musculoskeletal, neuromuscular, or rheumatologic pain conditions, and a history of physical or emotional trauma. They were recruited from the outpatient PT clinic where the study took place or through social media postings between June 2019 and May 2021. Trauma screening was conducted by the primary investigator using the Life Events Checklist (LEC).^42(p5) The LEC is a widely used self-report measure of trauma history designed to screen for potentially traumatic events in a respondent’s lifetime. It has adequate psychometric properties as a stand-alone assessment of traumatic exposure [42]. Participants had to have experienced, witnessed, or learned about at least one of the 16 events on the LEC, or had ‘Any other very stressful event or experience’ not included in the checklist, to be eligible for this study [43].

Participants who could not travel to the clinic were eligible to participate through telehealth. Telehealth PT has been found to be feasible, acceptable, and effective, as well as non-inferior to in-person rehabilitation programs [44,45]. Individuals were excluded if their pain was due to migraine, neurologic conditions such as Parkinson’s disease, or cerebrovascular accident, if they could not ambulate 20 feet with or without an assistive device, or if they did not have insurance or financial ability to cover the costs of PT services.

Intervention

Participants received six weeks of one-hour/week, direct PT sessions provided by a licensed physical therapist. Four licensed physical therapists each underwent four hours of training to deliver both interventions. Three of the therapists held a Doctor of Physical Therapy degree, three were board-certified orthopedic clinical specialists, and one had 20 years’ experience practicing in an outpatient setting. Therapists were given a research manual and met regularly with the primary investigator to ensure fidelity to trial protocols.

The REGION-PT protocol focused treatment to the region of reported pain and consisted of patient education, manual therapy, and region-specific exercise. Education included topics related to potential patho-anatomic causes of pain, ergonomic guidance, and advice to stay active. Therapists used pictures, diagrams, and skeletal models to aide in education delivery. Manual therapy (soft tissue mobilization, non-thrust joint mobilization) was performed to the painful region at each session. Strengthening and stretching exercises targeted the identified regional impairments. If the participant was seen via telehealth, additional stretching exercises replaced the manual therapy [46,47]. (Appendix A)

The GLOBAL-PT protocol followed a similar sequence as the REGION-PT group. Patient education consisted of PNE. Therapists used pictures, books, and ‘Why you hurt’ PNE tools^a. Graded Motor Imagery consisted of laterality exercises using the Recognize™ app^b or pictures in magazines, imagined movements, and mirror exercises [48,49]. If the participant was seen via telehealth, a home mirror was used for any mirror exercises. General strengthening and aerobic exercises were performed by each participant. Strengthening exercises targeting the body’s major muscle groups were progressed in terms of volume and mode over the course of the six visits according to guidelines from the American College of Sports Medicine [50]. (Appendix B) Participants in both groups were given written copies of their exercises to perform daily between PT visits.

Outcomes

The primary outcome was the difference of the group mean change in Patient-Reported Outcomes Measurement Information System (PROMIS)-Physical Function (PF) Computer Assisted Technology (CAT)v2.0 T-score. All PROMIS scores are presented as a standardized T-score, with a mean of 50 and a standard deviation of 10. The PROMIS-PF CAT is a reliable, valid, and responsive tool for assessing function in individuals with musculoskeletal and rheumatologic pain conditions [51–53].

Secondary outcomes were the difference in the group change in PROMIS Pain Interference (PI) CATv1.1 T-score, PROMIS-29v2.0 eight domain (physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate in social roles and activities, pain interference, pain intensity) scores, Brief Pain Inventory (BPI), Central Sensitization Inventory (CSI), and two quantitative sensory tests. The PROMIS-PI CAT and PROMIS-29 are valid, reliable, and responsive tools for the assessment of patients with chronic musculoskeletal pain [52,53]. The BPI is valid and reliable in assessing pain intensity and interference in patients with arthritis and back pain, and is sensitive to change over time [54,55]. The CSI is valid and reliable for quantifying the severity of symptoms for individuals with chronic musculoskeletal pain, osteoarthritis, and chronic pain with and without central sensitization syndrome [56–59]. While initially designed as a symptom screening tool, there is evidence of CSI responsiveness to treatment [56,60]. Quantitative sensory testing included two-point discrimination (TPD) and pain pressure threshold (PPT). Two-Point Discrimination has moderate to good intra-rater reliability; PPT has excellent intra-rater reliability (Appendix C) [61–64]. Sensory tests were not conducted if the outcome assessment was completed via telehealth.

An independent examiner collected all data before initiation of the randomly assigned protocol and after completion of 6 weeks of PT. The examiner completed four hours of practice and reliability training prior to the start of the trial, was blinded to the participant’s randomized grouping, and did not participate in the PT. The study data were stored and managed using Research Electronic Data Capture (REDCap), a secure, web-based software platform designed to support data capture for research studies [65].

Randomization

To ensure that groups were similar in the primary outcome of function at the start of the trial, participants who met the inclusion criteria, provided written informed consent, and completed the baseline assessment, were categorized into one of two strata: low function (PROMIS-PF T-score<35), or high function (PROMIS-PF CAT >35). Members of each stratum were assigned to a treatment group following a computer-generated randomization sequence with a block size of four. A priori the PI prepared and placed sequentially numbered cards with the randomization assignment into sealed, opaque envelopes to ensure randomization was concealed.

To ensure that randomization resulted in balanced groups, baseline characteristics, and outcome variables between groups were compared using Student’s t-test for continuous variables and Fisher’s exact test for categorical variables.

Statistical methods

We used linear mixed models to assess group-by-time differences for all outcomes. A priori, we suspected that the length of time someone lived with chronic pain (duration) sex, and race might moderate the effect of treatment, so we built models that included interaction terms for these variables. For the primary outcome, between group difference in PF, we reported 95% confidence intervals (CI) and P-values and took P < 0.05 to mean statistical significance. For the secondary outcomes, we did not adjust the P-value for multiple comparisons. We performed analyses examining the impact on outcomes of participants who received telehealth. Data analyses were performed using Stata Statistical Software: Release 15. (StataCorp. 2017. College Station, TX).

Assuming a standard deviation of 6.4 for the change in PROMIS-PF T-score and alpha = 0.05, a total of 84 participants (42 per group) have 80% power to detect a difference of 4.0 [52,66]. To accommodate possible loss to follow-up, we sought to enroll 100 participants. If a participant dropped out of the study, we enrolled additional participants to achieve our target sample size.

Results

We assessed 197 individuals for eligibility; 117 participants met the inclusion criteria, consented, and enrolled in the trial. After randomization, nine participants (8%) never initiated treatment and ten (9%) dropped out after starting PT (Figure 1).

Figure 1.

Consort Flow Diagram.

Ninety-eight participants completed the trial. Eighty-nine participants (91%) received all six therapy visits per protocol, nine (9%) received 1–5 visits due to unexpected insurance coverage issues (n = 4), moving out of state (n = 1), being too busy (n = 2), or employment constraints (n = 2). Of the 98 completers, 10 (five in each group) received all treatment and assessments via telehealth. An additional 15 participants (5 in GLOBAL, 10 in -PT) received their final assessment via telehealth due to scheduling or sickness, although they received the intervention itself in person.

The participants had a mean age of 52 ± 18, 69 (70%) were female, and 81 (83%) were white. The mean duration of chronic pain was 11.5 years ±11.9, with a median of 8 years. Mean number of traumatic events experienced was 4.8 ± 2.4, with a median of four. Baseline characteristics of the two treatment groups were similar (Table 1). For the primary outcome of PROMIS-PF T-score, the GLOBAL-PT participants improved by +2.9 points and REGION-PT by +2.2, for a T-score difference of +0.7 (CI: −1.4, +2.8; P = 0.52) (Table 2, Figure 2). In comparing the two groups mean change in PROMIS-PI T-score, the GLOBAL-PT group reduced by −5.2 points compared to −2.9 points in the REGION-PT group for a T-score difference of −2.3 (CI: −4.9, +0.3; P = 0.09) (Figure 2).

Table 1.

~TC~.

Baseline Characteristics by Treatment Group
	REGION-PT	GLOBAL-PT	P Value
Allocated N	53	45
Age in years, mean (SD)	51 (18)	53 (18)	0.64
Sex (% female)	40 (75%)	29 (64%)	0.44
Race (%)			0.28
White	44 (83%)	37 (82%)
Black	0	2 (4%)
Asian	2 (4%)	4 (9%)
Iraqi	1 (2%)	0
Native American	0	1 (2%)
Hispanic	1 (2%)	0
White Hispanic-US born	2 (4%)	0
Other/Not reported	3 (6%)	1 (2%)
Duration of pain in years, mean SD)	11 (10)	13 (14)	0.36
Traumatic events, mean (SD)	4.8 (2.6)	4.8 (2.2)	0.77
Telehealth, N (%)	5 (9%)	5 (11%)	>0.99
Region of pain, N (%)
Lumbar spine	24 (45%)	19 (42%)	0.84
Thoracic spine Cervical spine	3 (6%) 15 (28%)	4 (9%) 10 (22%)	0.70 0.64
Upper extremity	3 (6%)	1 (2%)	0.62
Lower extremity	4 (8%)	6 (13%)	0.51
Other	4 (8%)	5 (11%)	0.73
Number of regions of pain, N (%)
One	28 (53%)	21 (47%)	0.82
Two	21 (40%)	21 (47%)
Three	4 (8%)	3 (7%)
Diagnoses, N (%)
Fibromyalgia	5 (9%)	5 (11%)	>0.99
Ehlers Danlos Syndrome	0 (0%)	1 (2%)	0.46
Lyme Disease	0 (0%)	3 (7%)	0.09
Myofascial Pain Syndrome	0 (0%)	2 (4%)	0.21
PROMIS-PF, mean T-score (SD)	40 (7.0)	41 (7.5)	0.79
PROMIS-PI, mean T-score (SD)	61 (5.7)	61 (6.8)	0.97
PROMIS-29, mean T-score (SD)
Physical Function	42.1 (6.4)	42.6 (7.8)	0.70
Anxiety	59.3 (8.4)	59.0 (9.6)	0.70
Depression	55.5 (7.5)	55.1 (9.0)	0.80
Fatigue	58.0 (7.0)	56.7 (7.3)	0.37
Sleep disturbance	55.4 (5.4)	57 (6.2)	0.16
Social roles	44.8 (6.0)	45.7 (10.6)	0.59
Pain interference	62.5 (5.6)	60.8 (7.9)	0.24
Pain intensity	5.4 (2.1)	5.3 (2.2)	0.78
Brief Pain Inventory, mean (SD)
Worst Pain	6.1 (2.1)	5.9 (2.3)	0.78
Best Pain	2.6 (2.3)	2.7 (2.1)	0.67
Current Pain	3.5 (2.5)	4.2 (2.7)	0.17
Average Pain	4.8 (2.2)	4.5 (2.2)	0.63
BPI Pain interference	4.4 (2.1)	4.2 (2.6)	0.53
CSI Score, mean (SD)	54 (12)	50 (13)	0.32
Two Point Discrimination, mean (SD)*
Tibialis Anterior	n=49	n=36	0.82
		54 (17)	55 (21)
Cervical region	n=21	n=15	0.07
		48 (13.9)	56 (10.6)
Lumbar region	n=28	n=21	0.59
		61 (13.9)	59 (12.3)
Pain Pressure Threshold, mean (SD) †
Tibialis Anterior	n=49	n=36	0.56
		10 (6.9)	9.6 (5.0)
Cervical region	n=21	n=15	0.51
		9.4 (4.5)	11 (4.9)
Lumbar region	n=28	n=21	0.72
		12 (6.5)	12 (7.4)

*Mean of 2 trials in mm, left and right sides combined, †mean of 3 trials in lbs. of pressure left and right sides combined.

Table 2.

Unadjusted Outcomes by Treatment Group.

	REGION-PT Group Mean (n=53)				GLOBAL-PT Group Mean (n=45)				GLOBAL-PT-REGION-PT
	Baseline	6 Week	Change	P	Baseline	6 Week	Change	P	Difference (CI)	P*
Physical Function	40.3	42.5	+2.2	0.14	40.7	43.5	+2.9	0.09	+0.7 (−1.4, +2.8)	0.52
Pain Interference	61.2	58.3	−2.9	0.02	61.2	56.0	−5.2	0.001	−2.3 (−4.9, +0.3)	0.09
PROMIS-29 PF	42.1	44.3	+2.2	0.14	42.6	44.9	+2.3	0.17	+0.1 (−2.2, +2.4)	0.92
PROMIS-29 Anxiety	59.4	57.1	−2.3	0.17	58.6	54.9	−3.7	0.07	−1.4 (−4.1, +1.2)	0.29
PROMIS Depression	55.5	53.0	−2.5	0.09	55.1	53.2	−1.9	0.34	−0.6 (−2.2, +3.4)	0.67
PROMIS Fatigue	58.0	56.3	−1.7	0.23	57.0	54.0	−3.0	0.08	−1.3 (−4.1, +1.6)	0.38
PROMIS Sleep Disturbance	55.4	54.9	−0.5	0.56	57.0	55.8	−1.2	0.26	−0.7 (−2.8, +1.3)	0.49
PROMIS Social Roles	44.8	48.6	+3.8	0.01	45.7	50.4	+4.7	0.02	+0.9 (−1.9, +3.6)	0.53
PROMIS Pain Interference	62.5	59.3	−3.2	0.01	60.8	57.0	−3.8	0.02	−0.6 (−3.4, +2.0)	0.62
PROMIS Pain Intensity	5.4	4.2	−1.2	0.003	5.3	3.7	−1.6	0.001	−0.4 (−1.2, +0.4)	0.29
BPI Average	4.8	3.9	−0.9	0.02	4.6	3.5	−1.1	0.03	−0.2 (−0.9, +0.5)	0.57
BPI Best	2.6	2.1	−0.5	0.22	2.8	2.0	−0.8	0.11	−0.3 (−1.0, +0.5)	0.44
BPI Worst	6.1	5.3	−0.8	0.07	5.9	4.4	−1.5	0.003	−0.7 (−1.7, +0.2)	0.12
BPI Current	3.5	3.2	−0.3	0.47	4.2	3.0	−1.2	0.04	−0.9 (−1.7, +0.02)	0.06
BPI PI	4.4	3.4	−1.0	0.03	4.2	2.9	−1.3	0.02	−0.3 (−1.2, +0.4)	0.35
CSI	53.8	52.2	−1.6	0.53	50.3	52.8	+2.5	0.44	+4.1 (−1.5, +9.7)	0.15
Two-point discrimination **(AT)	54.1	51.9	−2.2	0.59	55.0	49.6	−5.4	0.24	−2.8 (−10, +4.7)	0.46
Pain Pressure Threshold Ɨ (AT)	10.3	11.9	+1.6	0.29	9.6	9.3	−0.3	0.75	−1.8 (−3.6, +0.2)	0.053

*Mixed model repeated measures group x time interaction; **= millimeters, Ɨ = pounds of pressure; AT= Anterior Tibialis; BPI=Brief Pain Inventory; PF= Physical Function; PROMIS29 physical function & social roles higher score is better; anxiety, depression, fatigue, sleep & pain interference higher score is worse.

Figure 2.

Difference in Physical Function and Pain Interference by Group.

There were no between-group differences for the CSI, the PROMIS-29 domains, the BPI categories, or for PPT or TDP sensory testing. Sensory testing was not performed on the participants who completed the post-treatment assessment via telehealth (n = 25). Forty-six participants had PPT and TPD data analyzed in the lumbar spine region, 27 participants in the cervical spine region, and 73 with data from the tibialis anterior site (Table 2). Re-analysis excluding the 10 participants who received treatment via telehealth was essentially identical to the overall analysis.

Regardless of treatment allocation, all 98 participants, on average, demonstrated improvement in Physical Function (T-score mean change +2.5 points; CI: +0.35, +4.6; P = 0.02). For the secondary outcome of the change in PROMIS-Pain Interference, on average, participants had reduced T-score PI (−4.0 points; CI: −5.9, −2.0; P < 0.0001). Independent of the treatment group, participants improved significantly with either approach in all PROMIS-29 domains excepting depression and sleep disturbance (Table 3). The BPI improved overall with participants reducing by −1.1 points for worst pain (CI: −1.8, −0.5; P = 0.0007), −1.0 average pain (CI: −1.6, −0.4; P = 0.002), −0.6 best pain (CI: −1.3, −0.2; P = 0.04), −0.7 current pain (CI: −1.5, +0.001; P = 0.051) and −1.1 for PI (CI: −1.8, −0.4; P = 0.001). The CSI scores remained unchanged for participants (+0.3; CI: −3.7, +4.2; P = 0.88). Two-point discrimination improved slightly for participants overall (−4.0 mm; CI: −9.5, +2.3; P = 0.22). Pain pressure threshold did not demonstrate change (increased by +0.7 lbs.; CI −1.3, +2.6; P = 0.52).

Table 3.

PROMIS 29 Outcomes Overall.

PROMIS -29	Pre to post T-score change (CI)	P
Physical Function	+2.2 (+0.6, +4.4)	0.04
Pain Interference	−3.5 (−5.4, −1.5)	0.001
Anxiety	−3.0 (−5.5, −0.4)	0.02
Depression	−2.2 (−4.6, +0.1)	0.06
Fatigue	−2.3 (−4.5, −0.1)	0.04
Sleep Disturbance	−0.9 (−2.3, +0.5)	0.22
Social Roles	+4.2 (+1.8, +6.5)	0.001
Pain Intensity	−1.4 (−2.0, −0.8)	<0.001

When examining interactions between-group intervention and race, we found that race impacted the effect of the intervention on pain interference. Both interventions were associated with reductions in PI T-score for white participants (REGION-PT: −4.1; CI: −5.9, −2.3; GLOBAL-PT: −4.3; CI: −6.3, −2.3) but nonwhite participants experienced PI reductions only with GLOBAL-PT (−9.4; CI: −13.7; −5.0). Participants in the REGION-PT group had increases in PI (+2.8; CI: −1.3, +6.9). The interaction was significant with P=<0.001 (Figure 3).

Figure 3.

Interaction of Race with Pain Interference by Group .

To assess for interactions between group intervention and the duration of chronic pain, we dichotomized duration of pain at the study population median of 8 years. Participants with <8 years of pain had improved function when treated with GLOBAL-PT (T-score +3.7; CI: +1.4, +5.9; REGION-PT: +0.95; CI: −1.1, +3.0) while those with ≥8 years of pain had a greater improvement in function with the REGION-PT protocol (+3.4; CI: +1.4, +5.3; GLOBAL-PT: +2.1; CI: −0.09, +4.3). The interaction was not significant with P = 0.07 (Figure 4).

Figure 4.

Interaction of Duration of Pain with Physical Function by Group .

Telehealth PT did not appear to interact with the effects of treatment on the primary outcome of physical function (P = 0.47). REGION-PT participants who received their intervention via telehealth improved their physical function T-score by +2.3, compared to +2.2 for REGION-PT participants who received the intervention in person. GLOBAL-PT participants who received their intervention via telehealth improved their physical function T-score by +0.7, compared to +3.1 for GLOBAL-PT participants who received the intervention in person. In comparing group differences, if participants received the allocated treatment via telehealth, there was a −1.6 difference in physical function (95% CI −8.7, +5.7; P = 0.63) favoring the REGION-PT group. If participants received the allocated treatment in person, there was a + 0.9 difference (95% CI −1.3, +3.3; P = 0.41) favoring the GLOBAL-PT group.

Discussion

We compared GLOBAL-PT and REGION-PT for individuals with chronic pain and a history of trauma. Although both groups improved over the course of treatment, we did not find significant differences between groups for our primary outcome of physical function.

We expected to see a greater improvement in function in participants receiving GLOBAL-PT, as it was hypothesized to address adverse nervous system changes, catastrophizing, and fear avoidance behaviors [21,27,67,68]. Watson, et al [35], found that use of PNE reduced catastrophizing and kinesiophobia. While we did not measure catastrophizing or kinesiophobia, we hypothesized that if these pathologic characteristics were reduced, participants would demonstrate an increased willingness to move and an improvement in their function. However, differences in function were not observed.

We considered whether the effect of treatment on physical function depends on how long a participant has lived with pain. For a shorter duration, the GLOBAL-PT group trended toward a greater functional improvement, while for a longer duration, the REGION-PT treatment had better outcomes. However, this observation was neither clinically nor statistically significant.

Our study did not evaluate participants for the presence of widespread pain, and this unmeasured variable may have impacted functional outcomes. For instance, in a recent study, individuals with chronic LBP who also had persistent widespread pain were more likely to have greater disability and a poorer quality of life than those who did not have widespread pain, and having widespread pain was highly correlated with moderate-to-high levels of pain catastrophizing [13].

Although pain interference improved in both groups, there was no significant difference between the two treatments at follow-up. Prior research has suggested that PNE, GMI and exercise may reduce pain in individuals with chronic MSK conditions [37,69]. Wood, et al [29], reported that when PNE was added to usual PT, pain symptoms improved. However, we did not see significant differences between the GLOBAL-PT and REGION-PT groups in any of our secondary outcomes. While there was a trend toward reducing PI in the GLOBAL-PT group, our results were not significant. We did find racial differences in PI outcomes with nonwhite participants benefiting from GLOBAL-PT, concordant with prior research suggesting that race may influence beliefs, cognitions, and behaviors around pain [70]. There is variability in the recommended duration of PNE, ranging from 30 min to 4 h [68]. The GLOBAL-PT approach combined PNE, GMI and exercise; our PNE duration over six visits was 90 min. We chose this duration to make the intervention feasible for usual clinical practice while including the other components of our GLOBAL-PT protocol. A longer duration or greater frequency of sessions might have been more effective, and this may have influenced our primary outcome of change in function.

We did not find any change in central sensitization symptoms as measured by the CSI or sensory testing. Perhaps there are subgroups of patients who do better with mental imagery techniques and might benefit from a GLOBAL-PT approach [71]. The success of mental imagery may require skills that vary across individuals [71]. Participants may not all have the ability to generate mental images which could influence results [71]. Likewise, the measurement tools used in our study may not be sensitive enough to changes in central sensitization, or our treatment time frame was not long enough to impart central nervous system changes, or the underlying model of sensitization is wrong.

Having a trauma experience may exacerbate chronic pain symptoms [72]. Children who experience trauma, neglect, or sexual abuse, are at increased risk of chronic pain as adults [72,73]. A history of trauma is common in individuals with fibromyalgia, with 30–55% of patients reporting childhood trauma [74]. Participants in our study experienced, on average >4 traumatic life events; impact was not measured. Regardless of the degree of trauma, our participants improved on many outcomes, independent of treatment group. This could be attributed to the effect of PT, but an enhanced therapeutic alliance must also be considered; a trusting relationship and a customized treatment plan considering individual values has been found to have a beneficial effect on chronic musculoskeletal pain [75]. Conversely, regression to the mean, secular trends, or other unknown causes are also possible explanations.

While we did not find differences between groups, we did find that on average, all participants demonstrated significant improvement in physical function and in many of the secondary outcomes. Our findings are similar to others who compared interventions and did not find one superior to another in improving function suggesting that choice of intervention should be individualized by patient preference and presentation [76–78]. That physical function, PI, pain intensity, anxiety, fatigue, and social roles improved after just six PT visits, in individuals with a trauma history who had been living with chronic pain for many years, speaks to the potential of structured PT to effect change in those with chronic pain conditions. Individualized assessment, clinical reasoning and appropriate evidence-based intervention may improve the outcomes on longer term.

Limitations

We hypothesized that chronic pain is centrally mediated with a common expression independent of the specific anatomic locale involved. Therefore, we included many chronic pain conditions typically seen in an outpatient PT clinic. This created a heterogeneous population which may have affected our results; however, we are confident that our randomization scheme balanced out participant allocation. Likewise, there may be unmeasured variables impacting our findings. However, randomization mitigates this as well.

Eligibility criteria included having health insurance or financial means to cover the costs of PT; findings are only generalizable to the insured population, or those with financial means to pay for PT. The study population was largely white, reflecting local demographics, and female, reflecting chronic pain populations around the world. There may be unidentified sub-populations with a different response to either intervention. Trauma history was identified via the LEC; we did not quantify the impact of the experienced trauma.

The lack of a concurrent untreated control group makes it difficult to discern if either treatment approach was better than watchful waiting in this setting [79].

Conclusion

We observed few group differences between the outcomes of those treated by GLOBAL-PT compared to a REGION-PT approach. However, participants improved on average with either treatment, suggesting that both are reasonable strategies in treating individuals with chronic pain and a history of trauma.

Biographies

Justine McCuen Dee is a PhD Candidate in the Larner College of Medicine Clinical and Translational Science program as well as a clinical associate professor in the University of Vermont Doctor of Physical Therapy program. She specializes in treatment of patients with chronic pain conditions, particularly those who have experienced physical and emotional trauma.

Benjamin Littenberg is an Internal Medicine Specialist whose research interests center on technology assessment and quality improvement. Recent projects include new ways to measure quality of care in chronic disease, novel strategies for reporting test results to patients, the effect of the built environment on health outcomes, strategies to address health literacy, assessment of diagnostic test strategies, and registry-based approaches to quality and safety improvement.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

Justine Dee is the co-owner of the outpatient PT practice where the research study treatment took place. The authors have no other potential conflicts or financial affiliations to disclose.

Abbreviations

BPI

brief pain inventory

CAT

computer assisted technology

CI

confidence interval

CNS

central nervous system

CSI

central sensitization inventory

GMI

graded motor imagery

LEC

Life Events Checklist

PF

physical function

PI

pain interference

PNE

pain neuroscience education

PPT

pain pressure threshold

PROMIS

patient reported outcome measurement information system

PT

physical therapy

QST

quantitative sensory tests

TPD

two-point discrimination

US

United States

Suppliers

^a Why you hurt” PNE tools. Louw A. Orthopedic Physical Therapy Products; 2013

^b Recognize app. NOIgroup, www.noigroup.com

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/10669817.2022.2159615