Clinical Presentation of Adults with Traumatic Orthopedic Injuries Enrolled in a Multisite Psychosocial Trial

Abstract

Objectives:

Traumatic orthopedic injuries are a top cause of hospital visits in the U.S. The Toolkit for Optimal Recovery (TOR) is a brief mind-body intervention that targets catastrophic thinking and pain anxiety following orthopedic injury. This study examines the baseline presentation of adults with traumatic orthopedic injuries who were enrolled in our recent multisite feasibility RCT of TOR versus usual care at four geographically distinct Level 1 trauma centers. We also examine whether patient presentation varies by site.

Methods:

We recruited 181 adults (M_age=44.16, SD=16.5) from four Level I trauma centers located in the northeast (Site A; N=63), southwest (Site B; N=44), southeast (Site C; N=44), and southeast (Site D; N=30). At baseline, participants provided information about sociodemographic factors, pain and physical function, and physicians completed the Abbreviated Injury Scale (AIS). Descriptive statistics were used to characterize the sample, and one-way analysis of variance (ANOVA) and Chi-square tests were used to compare variables between sites.

Results:

The majority of the sample (88.4%) sustained a fracture, and the mean AIS score was 2.31 (SD=0.55). Age, race, sex, gender, occupation, or marital status did not differ across sites (ps>.05). Over half (63%) of the sample was treated surgically, and 28.7% endorsed taking narcotic pain medications. More participants at Sites B (75%) and D (96.7%) received surgery than participants at Sites A (41%) and C (61.4%). More participants at Sites D and B reported narcotic usage than participants at Sites C and A. Participants at Site D demonstrated greater functional impairment than participants at the other sites.

Conclusion:

Although sites were largely comparable, we did find key differences in surgical management, narcotic use, and functional disability which may have important implications for treatment response. This information will be used to iterate and refine TOR for a future multisite efficacy trial.

Introduction

Traumatic orthopedic injuries, including fractures, dislocations, and ruptures, are among the most common causes of hospital visits in the United States.¹ Following appropriate surgical or non-surgical medical management, up to half of all orthopedic trauma patients will develop subsequent chronic pain and dysfunction.^1–3 chronic pain is associated with enormous personal and financial burdens, including reduced well-being and loss of mobility, productivity, and independence. Unhelpful pain-related thoughts and feelings following orthopedic trauma include pain anxiety (e.g., pain-related worry, hypervigilance), pain catastrophizing (e.g., magnifying and ruminating on pain-related thoughts), and activity avoidance, and are associated with the development of chronic pain. Orthopedic trauma occurs in a highly diverse population,^4–6 and patient outcomes following orthopedic injury (such as the development of chronic pain), may differ based on sociodemographic and clinical characteristics.⁷ Pain anxiety and catastrophizing are consistently associated with worse orthopedic clinical outcomes, including increased pain^8–11 and decreased function,¹² over and above measures of injury severity.^13,14 Unhelpful pain-related thoughts and feelings are modifiable and present a viable opportunity to prevent chronic pain and dysfunction.

To date, there are no evidence-based interventions targeting unhelpful pain-related thoughts and feelings following traumatic orthopedic injuries. Intervening early after injury to prevent chronic pain, rather than once patients are in a cycle of chronic pain and dysfunction, may reduce excess hospitalizations and subsequent medical costs. Additionally, early interventions can teach patients generalizable skills to manage pain and maximize well-being during rehabilitation. With feedback from multidisciplinary stakeholders,^15–18 we developed the Toolkit for Optimal Recovery (TOR) to help patients return to valued activities and regain physical functioning after traumatic orthopedic injury. Grounded within the Fear Avoidance Model (FAM),¹⁹ TOR targets pain anxiety, pain catastrophizing, and maladaptive pain avoidance.^20,21 The 4-session TOR program includes chronic pain psychoeducation (e.g., myths versus facts, a simplified fear-avoidance model), relaxation, mindfulness (e.g., mindfulness of the breath), adaptive thinking skills (e.g., restructuring unhelpful thoughts about pain), pain acceptance (e.g., understanding the futility of trying to stop or control pain), activity pacing, and value-based activity engagement.

TOR has demonstrated single site feasibility, acceptability, and satisfaction at an urban Level 1 trauma center in the northeast.²¹ We recently completed a multisite randomized controlled trial (RCT) to further evaluate feasibility, acceptability, and satisfaction at four Level 1 trauma centers. Given that orthopedic trauma occurs in a highly diverse population and is particularly prevalent and disabling among racial minorities and socioeconomically disadvantaged individuals,^4–6 it is important to understand whether this multisite trial captured a similarly diverse population. Age, race, education, and socioeconomic status can influence openness toward psychological interventions,²² other clinical characteristics (e.g., psychiatric diagnoses, medication usage, pre-existing chronic pain) have been shown to influence treatment outcomes.^13,23–26 Thus, capturing granular information about the baseline clinical characteristics of individuals with orthopedic injuries may help us better tailor our psychosocial intervention to our target patient population, and to identify factors that can be leveraged to better address varied patient needs. These factors may have important implications for treatment response in this population. Understanding baseline participant presentation, and whether this varies by site, will also provide the groundwork for interpreting the results of the multisite feasibility RCT and allow us to iterate and refine our intervention, if indicated, for future multisite trials. The aims of the current study are to 1) examine the baseline presentation of adults with traumatic orthopedic injuries across the four Level 1 trauma centers and 2) examine differences in baseline factors by study site.

Materials and Methods

Study Design and Setting

Participants were recruited from four Level 1 trauma centers for a multisite feasibility RCT of TOR: Site A (northeast), Site B (southwest), Site C (southeast), and Site D (southeast). Potentially eligible individuals were identified through an electronic medical chart review. If eligible, participants provided verbal consent to screening procedures before enrollment. During the baseline assessment, participants provided sociodemographic information, information about current and prior pain management, physical functioning, and pain-related psychosocial factors. The present analyses include data collected during the baseline assessment. See Vranceanu et al.²⁰ for a detailed description of the full multisite feasibility RCT study protocol. This multisite study was conducted in accordance with ethical standards. The Institutional Review Board of the primary institution (Site A) approved all study procedures.

Participants

Participants were English-speaking adults who sustained a traumatic orthopedic injury within the last 1-2 months, were cleared by their orthopedic surgeon to participate, and scored ≥20 on the Pain Catastrophizing Scale²⁷ or ≥40 on the Pain Anxiety Symptoms Scale Short Form 20.²⁸ Exclusion criteria included: surgery complications or other serious comorbid injuries; diagnosis of a medical condition expected to worsen in the next three months (e.g., malignancy); lifetime history of serious mental illness; current substance use disorder; current suicidal ideation; engaged in mind-body practice (e.g., yoga/mediation) at least once per week for 45 minutes or more within the last three months; and pregnancy. Eligible, interested participants provided informed consent.

Measures

Demographic Questionnaire: Participants completed a demographic questionnaire assessing age, gender, race/ethnicity, marital status, income level, education level, employment status, study site location, prior orthopedic injuries, current psychotropic and pain treatments, and history of psychological and substance use diagnoses.

Pain: Pain at rest and with activity was assessed using the Numerical Rating Scale (NRS).²⁹ This 2-item measure employs an 11-point Likert scale ranging from (0) no pain to (10) the worst pain imaginable.

Physical Function: Patient perception of physical functioning was assessed using the Patient Reported Outcomes Measurement Information System - Physical Function, version 1.2.8b (PROMIS-PF, v. 1.2.8b).³⁰ This 8-item measure assesses perception of capability across several different activities using a 5-point Likert scale from (1) unable to do to (5) able to do.

We also utilized the Short Musculoskeletal Function Assessment – Dysfunction Index (SMFA-DI) to assess self-reported physical function.³¹ This 34-item measure

assesses physical function and musculoskeletal disability using a five-point Likert scale ranging from (1) not at all difficult to (5) unable to do, with high scores indicative of higher disability.

Pain Catastrophizing: Catastrophic thoughts about pain (i.e., rumination, magnification of the threat value of pain, perceived helplessness) were assessed using the Pain Catastrophizing Scale (PCS).³² This 13-item measure employs a five-point Likert scale ranging from (0) not at all to (4) all the time, with higher scores indicating greater pain catastrophizing.

Pain Anxiety: Pain-related anxiety was assessed using the Pain Anxiety Symptoms Scale Short Form 20 (PASS-20).²⁸ This 20-item measure assesses factors of pain-related anxiety (i.e., cognitive, fear, escape/avoidance, physiological) on a six-point Likert scale ranging from (0) never to (5) always, with higher scores indicating greater pain-related anxiety.

Clinical Pain Characteristics: Post-injury rehabilitation was assessed using the question “Have you engaged in physiotherapy (physical therapy or occupational therapy) in the past week? (Yes/no)”. Participants also reported if they had received surgical management of their injury (yes/no). History of pain was assessed by asking “In the past year prior to your current injury, did you have any serious pain or pain that required medical attention? (Yes/no)”. If they endorsed a history of pain, they were asked a follow-up question assessing chronicity, “Was this pain longer than three months? (Yes/no)”.

Injury Severity: Injury severity was reported by the physician using the Abbreviated Injury Scale (AIS).³³ This measure classifies each injury by body region on a six-point scale ranging from (1) minor to (6) maximal. AIS scores were assessed by orthopedic surgeons following enrollment.

Statistical Approach

We conducted all analyses in R Version 4.2.1³⁴ using RStudio.³⁵ To characterize the sample for Aim 1, we reported descriptive statistics for the demographic and clinical variables. For Aim 2, we conducted a series of one-way analysis of variance (ANOVA) for continuous variables and Chi-square test of independence for categorical variables by site. We report the effect sizes for ANOVA and Chi-square test of independence using partial eta squared and Cramer's V, respectively. Two-tailed p<0.05 were considered statistically significant. Effect sizes were reported as Cramer’s V for categorical variables (0 indicating no effect, >.25 indicating a large effect) and η_p² for continuous variables interpreted as small (.01), medium (.06), and large (>.14).^36,37

Results

Demographic Characteristics

Participants included 181 adults (M_age=44.2, SD=16.5) across the four study sites (Site A: N=63, Site B: N=44, Site C: N=44, Site D: N=30). 68.5% of participants identified as female (N=124), and 13.8% identified as Hispanic or Latino/Latina (N=25). Participants were primarily White individuals (76.2%) and just under half (49.2%) endorsed being employed full-time. In terms of psychological characteristics, 34.3% of participants endorsed depression diagnoses (N=62), 35.4% endorsed anxiety diagnoses (N=64), and 8.3% endorsed PTSD diagnoses (N=15). All demographic data and participant characteristics for the total sample are included in [Table 1].

Table 1.

Sociodemographic Characteristics

		Study Site
	Total (N=181)	Site A (N=63)	Site B (N=44)	Site C (N=44)	Site D (N=30)	p -value
Age	M (SD) 44.2 (16.5) N (%)	M (SD) 46.8 (17.8) N (%)	M (SD) 42.9 (15.2) N (%)	M (SD) 43.6 (15.7) N (%)	M (SD) 41.3 (16.6) N (%)	0.42
Sex Male Female	57 (31.5%) 124 (68.5%)	17 (27.0%) 46 (73.0%)	13 (29.5%) 31 (70.464)	14 (31.8%) 30 (68.2%)	13 (43.3%) 17 (56.7%)	0.45
Gender Man Woman Genderqueer/gender fluid/nonbinary Prefer not to say	56 (30.9%) 119 (65.7%) 5 (2.8%) 1 (0.6%)	16 (25.4%) 44 (69.8%) 2 (3.2%) 1 (1.6%)	13 (29.5%) 30 (68.2%) 1 (2.3%) 0 (0.0%)	15 (34.1%) 29 (65.9%) 0 (0.0%) 0 (0.0%)	12 (40.0%) 16 (53.3%) 2 (6.7%) 0 (0.0%)	0.60
Race American Indian Alaska Native Asian Black/African American More Than One Race White Choose Not to Answer	1 (0.6%) 9 (5.0%) 14 (7.7%) 11 (6.1%) 138 (76.2%) 8 (4.4%)	0 (0.0%) 7 (11.1%) 5 (7.9%) 2 (3.2%) 47 (74.6%) 2 (3.2%)	0 (0.0%) 0 (0.0%) 4 (9.1%) 3 (6.8%) 32 (72.7%) 5 (11.4%)	1 (2.3%) 1 (2.3%) 1 (2.3%) 3 (6.8%) 38 (84.4%) 0 (0.0%)	0 (0.0%) 1 (3.3%) 4 (13.3%) 3 (10.0%) 21 (70.0%) 1 (3.3%)	0.07
Ethnicity Hispanic Non-Hispanic Choose Not to Answer	25 (13.8%) 151 (83.4%) 5 (2.8%)	5 (7.9%) 56 (88.9%) 2 (3.2%)	15 (34.1%) 27 (61.4%) 2 (4.5%)	2 (4.5%) 41 (93.2%) 1 (2.3%)	3 (10.0%) 27 (90.0%) 0 (0.0%)	0.001*
Education Less than high school Completed high school, GED Some college/Associates Completed 4 years of college Graduate/professional degree Choose Not to Answer	1 (0.6%) 47 (26.0%) 42 (23.2%) 37 (20.4%) 7 (3.9%) 47 (26.0%)	0 (0.0%) 22 (34.9%) 7 (11.1%) 17 (27.0%) 2 (3.2%) 15 (23.8%)	1 (2.3%) 9 (20.5%) 6 (13.6%) 12 (27.3%) 2 (4.5%) 14 (31.8%)	0 (0.0%) 10 (22.7%) 13 (29.5%) 6 (13.6%) 1 (2.3%) 14 (31.8%)	0 (0.0%) 6 (20.0%) 16 (53.3%) 2 (6.7%) 2 (6.7%) 4 (13.3%)	0.003*
Employment Employed full-time Employed part-time School full or part-time Keeping House/Housemaker Other Retired Unemployed Choose Not to Answer	89 (49.2%) 24 (13.3%) 6 (3.3%) 3 (1.7%) 10 (5.5%) 17 (9.4%) 26 (14.4%) 6 (3.3%)	31 (49.2%) 6 (9.5%) 4 (6.3%) 1 (1.6%) 4 (6.3%) 9 (14.3%) 5 (7.9%) 3 (4.8%)	24 (54.5%) 7 (15.9%) 0 (0.0%) 1 (2.3%) 3 (6.8%) 2 (4.5%) 5 (11.4%) 2 (4.5%)	20 (45.5%) 6 (13.6%) 2 (4.5%) 1 (2.3%) 2 (4.5%) 4 (9.1%) 9 (20.5%) 0 (0.0%)	14 (46.7%) 5 (16.7%) 0 (0.0%) 0 (0.0%) 1 (3.3%) 2 (6.7%) 7 (23.3%) 1 (3.3%)	0.70
Income Less than $10,000 $10,000 to less than $15,000 $15,000 to less than $20,000 $20,000 to less than $25,000 $25,000 to less than $35,000 $35,000 to less than $50,000 $50,000 to less than $75,000 $75,000 or more Choose Not to Answer	26 (14.4%) 10 (5.5%) 4 (2.2%) 10 (5.5%) 14 (7.7%) 24 (13.3%) 18 (9.9%) 51 (28.1%) 24 (13.2%)	6 (9.5%) 4 (6.3%) 0 (0.0%) 2 (3.2%) 4 (6.3%) 10 (15.9%) 3 (4.8%) 23 (36.5%) 11 (17.5%)	6 (13.6%) 3 (6.8%) 0 (0.0%) 2 (4.5%) 6 (13.6%) 2 (4.5%) 7 (15.9%) 13 (29.5%) 5 (11.4%)	6 (13.6%) 2 (4.5%) 3 (6.8%) 6 (13.6%) 2 (4.5%) 9 (20.5%) 5 (11.4%) 9 (20.5%) 2 (4.5%)	8 (26.7%) 1 (3.3%) 1 (3.3%) 0 (0.0%) 2 (6.7%) 3 (10.0%) 3 (10.0%) 6 (20.0%) 6 (20.0%)	0.04*
Psychological Characteristics Depression Anxiety PTSD	62 (34.3%) 64 (35.4%) 15 (8.3%)	23 (36.5%) 26 (41.3%) 7 (11.1%)	14 (31.8%) 11 (25%) 3 (6.8%)	13 (29.5%) 19 (43.2%) 1 (2.3%)	12 (40%) 8 (26.7%) 4 (13.3%)	0.77 0.16 0.28

*Indicates statistically significant p<0.05

Baseline Characteristics

±±±

±±±±Table 2].

Table 2.

Injury and Post-Injury Management Characteristics

	Total (N = 181)	Site A (N = 63)	Site B (N = 44)	Site C (N = 44)	Site D (N = 30)	p-value
	N (%)	N (%)	N (%)	N (%)	N (%)
Injury Type						0.72
Fracture	160 (88.4%)	55 (87.3%)	38 (86.4%)	38 (86.4%)	29 (96.7%)
Dislocation	3 (1.7%)	0 (0.0%)	0 (0.0%)	3 (6.8%)	0 (0.0%)
Rupture	3 (1.7%)	1 (1.6%)	2 (4.5%)	0 (0.0%)	0 (0.0%)
Multiple	11 (6.1%)	4 (6.3%)	3 (6.8%)	3 (6.8%)	1 (3.3%)
Not Identified	4 (2.2%)	3 (4.8%)	1 (2.3%)	0 (0.0%)	0 (0.0%)
Past Week Non-Narcotics						0.38
Yes	109 (60.2%)	36 (57.1%)	23 (52.3%)	30 (68.2%)	20 (66.7%)
No	72 (39.8%)	27 (42.9%)	21 (47.7%)	14 (31.8%)	10 (33.3%)
Past-Week Narcotics						< 0.001*
Yes	52 (28.7%)	11 (17.5%)	18 (40.9%)	7 (15.9%)	16 (53.3%)
No	129 (71.3%)	52 (82.5%)	26 (59.1%)	37 (84.1%)	14 (46.7%)
Physiotherapy						0.007*
Yes	57 (31.5%)	28 (44.4%)	9 (20.5%)	8 (18.2%)	12 (40.0%)
No	124 (68.5%)	35 (55.6%)	35 (79.5%)	36 (81.8%)	18 (60.0%)
Surgery						< 0.001*
Yes	114 (63.0%)	25 (39.7%)	33 (75.0%)	27 (61.4%)	29 (96.7%)
No	65 (35.9%)	36 (57.1%)	11 (25.0%)	17 (38.6%)	1 (3.3%)
No Answer	2 (1.1%)	2 (3.2%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
Past Year Serious Pain						0.01*
Yes	43 (23.8%)	24 (38.1%)	6 (13.6%)	7 (15.9%)	6 (20.0%)
No	138 (76.2%)	39 (61.9%)	38 (86.4%)	37 (84.1%)	24 (80.0%)
Past Year Chronic Pain						0.07
Yes	27 (14.9%)	15 (23.8%)	4 (9.1%)	4 (9.1%)	4 (13.3%)
No	16 (8.8%)	9 (14.3%)	2 (4.5%)	3 (6.8%)	2 (6.7%)
No Answer	138 (76.2%)	39 (61.9%)	38 (86.4%)	37 (84.1%)	24 (80.0%)
	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)
Injury Severity (AIS)	2.3 (0.6)	2.5 (0.7)	2.3 (0.5)	2.1 (0.3)	2.5 (0.5)	< 0.001*
Pain at Rest (NRS)	3.8 (2.4)	3.1 (2.1)	4.2 (2.5)	3.6 (2.1)	4.8 (2.7)	0.007*
Pain with Activity (NRS)	6.0 (2.5)	5.1 (2.5)	6.5 (2.3)	6.1 (2.3)	7.0 (2.7)	0.002*
Pain Catastrophizing (PCS)	23.2 (10.8)	20.8 (10.5)	23.1 (11.0)	23.1 (10.3)	28.3 (10.5)	0.02*
Pain-Related Anxiety (PASS)	55.4 (16.0)	54.4 (15.1)	55.3 (13.8)	54.2 (18.2)	59.2 (16.7)	0.53
Physical Function (PROMIS)	30.3 (6.9)	30.8 (7.3)	31.3 (7.1)	30.8 (6.4)	27.3 (5.5)	0.07
Physical Function (SMFA-DI)	46.7 (15.0)	44.6 (14.9)	44.3 (15.5)	45.4 (14.4)	56.4 (12.0)	0.001*

Note. Abbreviated Injury Scale (AIS); NRS = Numeric Rating Scale; PCS = Pain Catastrophizing Scale; PASS = Pain Anxiety Symptoms Scale

Site Differences in Baseline Characteristics

We observed significant differences in participant characteristics by site [Table 1]. More participants identified as Hispanic or Latino/Latina at Site B (southwest) than the other three sites (34.1% Site B vs 7.94% Site A, 4.55% Site C, 10% Site D; X²(6, N=181) =22.59, p=.001, V=.25. Education differed between sites, X²(15, N=181) =34.62, p=.003, with participants at Site D (southeast) reporting lower levels of education than the other sites (53.3% reporting “Some college/Associates degree; p=.001, V=.25). Additionally, there was a significant difference in income between sites X²≥≥

Several injury and post-injury management variables were significantly different between the sites [Table 2±±±F (3, 175) =6.36, p<.001, η_p²=.10. Additionally, more participants at Sites B (75%) and D (96.7%) received surgery following their injury than at Site A (41.0%) and Site C (61.4%), X²(3, N=181) =30.24, p<.001, V=.41 [Figure 1]. More participants at Site D (53.3%) and Site B (40.9%) reported narcotic usage in the week prior to enrollment than at Site C (15.9%) and Site A (17.5%), X²(3, N=181)=19.50, p<.001, V=.33. More participants at Site A (44.4%) and Site D (40.0%) reported engaging in physiotherapy following their injury than at Site C (18.2%) and Site B (20.5%), X²(3, N=181)=12.00, p=.007, V=.26. In total, 23.8% of participants (N=43) across all sites reported experiencing serious pain in the year prior to enrollment, with participants at Site A (38.1%) reporting the highest rates (Site B: 13.6%, Site C: 15.9%, Site D: 20.0%; X²(3, N=181) =11.37, p=.010, V=.25). Of those with previous serious pain, 62.8% (N=27) reported this pain was chronic (lasting longer than three months). Sites did not differ in the number of participants endorsing chronic pain in the year prior to enrollment (p=.07). Additionally, participants across all sites did not differ in their usage of non-narcotic pain medications (p=.38).

Figure 1.

Injury variables of chronic pain (serious pain in the last year that lasted longer than three months), narcotic use, physiotherapy attendance, and surgical management by site

±±F (3, 177) =3.44, p=.018, η_p²±F (3, 177) =5.40, p=.001, η_p²± ±±p=.07) or PASS scores (p=.53) differed by site.

Discussion

Despite the known influence of psychosocial factors on pain and functional outcomes following orthopedic injuries,³⁸ no evidence-based interventions directly addressing catastrophic thinking about pain and pain-related anxiety are available. The present study sought to characterize the sociodemographic and clinical characteristics of individuals with orthopedic trauma enrolled in a multisite psychosocial trial. Identifying these baseline characteristics may help us better identify and address patient needs and elucidate factors that may have important implications for treatment response in this population. Given that unhelpful pain-related thoughts and feelings are a primary predictor of clinical outcomes (e.g., the development of chronic pain) following traumatic orthopedic injuries,³⁸ novel interventions tailored to address the specific needs of this population are needed. The 4-session TOR program was developed to fill this gap by providing patients with skills to encourage them to return to valued activities and regain physical function following traumatic orthopedic injury. TOR has demonstrated single-site feasibility, acceptability, and satisfaction,²¹ and we recently completed a multisite feasibility RCT of the TOR program. Our previous work highlighted the need to recruit a more racially and ethnically diverse sample from geographically diverse regions of the U.S.,²¹ and the present study sought to explore differences in patient characteristics, which may be associated with subsequent TOR treatment response, between each of the four Level 1 trauma centers.

Overall, participants were primarily White, middle-aged, non-Hispanic, and female. A sample with greater variability in race, ethnicity, and sex is desirable when testing the feasibility and efficacy of new interventions to infer potential generalizability. Traumatic orthopedic injuries are particularly disabling among racial minorities,^4,5 and race can influence openness toward psychological interventions²² which is important to consider when evaluating the future efficacy of TOR. To increase the feasibility of recruiting a more diverse sample, the future TOR trials will include Level 1 trauma centers with larger populations of racially and ethnically diverse patients, and a concerted effort will be made to enroll a more sex-balanced sample. Consistent with inclusion criteria, all participants reported moderate pain-related anxiety and/or pain catastrophizing, demonstrating the feasibility of recruiting a sample with the target mechanism for TOR at each of the four study sites. Additional participant characteristics including psychiatric diagnoses, medication usage, and pre-existing chronic pain can negatively influence injury recovery and response to TOR.^13,23–26 Our results are consistent with other orthopedic literature, with over one third of participants endorsing depression and/or anxiety,⁹ over a quarter taking narcotic pain medications,²⁵ and one in every six participants reporting chronic pain in the last year.³⁹ Attending physiotherapy also plays a critical role in physical recovery from traumatic orthopedic injury. In the present study, less than one third of participants were attending physiotherapy at the time of enrollment – one to two months following their traumatic orthopedic injury – despite the fact that 63% of participants had undergone surgery. This is lower than an expected 70% of cases with physiotherapy referrals.⁴⁰ Low physiotherapy attendance highlights the need for additional recovery support providing further justification for TOR.

Several important differences emerged between study sites. Participants at Site D presented with a greater clinical burden, including higher pain catastrophizing and lower physical function scores, as well as lower education and income levels. These disparities at Site D may indicate a greater need for efficacious treatments to prevent chronic pain and disability. There were also noteworthy differences in medical management based on site. For example, while nearly all participants at Site D received surgery for their injury, less than half received surgery at Site A, demonstrating a divergence in management of equivalently severe injuries. For the same or worse injury severity than the other sites, Site A had the lowest rate of surgery, the highest rate of physiotherapy, and low rates of opioid use, while also having the highest rate of prior serious pain. This suggests that usual care at Site A may consider factors beyond injury severity (e.g., history of serious pain) to dictate treatment course, resulting in greater utilization of physiotherapy and lower rates of opioids and surgical management.

Future work will examine whether these geographic variations in baseline participant characteristics will impact the feasibility, acceptability, or satisfaction of the TOR program. The increased clinical need (i.e., higher pain catastrophizing and lower physical function) and contributing socioeconomic factors (i.e., lower education and income levels) identified at Site D are important to consider when analyzing the multisite feasibility data and planning for future multisite efficacy trials, as these patients may have additional barriers to care (e.g., financial concerns, difficulty getting to medical appointments) that could impact their treatment engagement and response. Subsequent longitudinal examinations of the TOR multisite feasibility study may assess if these varied patient groups responded differently, and if injury severity moderates the relationship between pain-related anxiety/catastrophizing and physical function.

Limitations

While we included four geographically diverse sites, we did not include U.S. sites located in the Midwest or West Coast. Future multisite trials of TOR will further increase geographic diversity to examine the feasibility and acceptability in diverse and lower-resourced clinics. Site D began recruitment later than the other 3 sites and enrolled fewer participants as a result. This discrepancy in participant volume should be considered when interpreting the results. Additionally, 68.5% of the participants identified as female. As help-seeking behaviors and pain processing differ between sexes,^41–44 the primarily female sample may have had an impact on the results. Future studies will aim to recruit a sex-balanced sample by utilizing more targeted recruitment methods (e.g., medical record review, diversifying recruitment locations, and amending study materials to cater to specific populations).

Conclusion

We reported on the baseline characteristics and geographic site variation for a multisite feasibility RCT of the TOR program, which is designed to help patients with elevated pain anxiety or catastrophizing return to valued activities and regain physical function following traumatic orthopedic injury. Participants were predominantly White and female and exhibited lower levels of physiotherapy attendance than expected. We found that sites were largely comparable, with key differences in narcotic use, physiotherapy attendance, and physical function, which may have important implications for treatment response. This baseline information will be used to interpret the results of the multisite feasibility RCT, iterate and refine the TOR intervention, and conduct a future multisite efficacy clinical trial.

Acknowledgement

The authors acknowledge Jafar Bakhshaie, David Ring, Thuan Ly, Caitlin Conley, Cara Vantrease, Robert Parker, Ellie Briskin, Lucy Bowers, Niels Brinkman, Amanda Priest, Terence Penn, and Brooke Duarte for their important contributions to the Toolkit for Optimal Recovery Study.

Authors Contribution:

J.E.H. contributed to conception, design, data acquisition, data analysis and interpretation, drafting the article, and final approval. K. N. J. contributed to conception, design, data analysis and interpretation, drafting the article, and final approval. R.A.M. contributed to acquisition of data, analysis and interpretation of data, drafting the article, and final approval. J.D.D. contributed to acquisition of data, analysis and interpretation of data, drafting the article, and final approval. J.R.B. contributed to analysis and interpretation of data and drafting the article. A.M.V contributed to conception, design, and final approval of the article.

Declaration of Conflict of Interest:

The authors do NOT have any potential conflicts of interest for this manuscript.

Declaration of Funding:

Dr. Jochimsen is funded by a National Center for Complementary and Integrative Health K23 Mentored Patient-Oriented Research Career Development Award (K23AT011922). Dr. Mace is funded by a National Institute on Aging K23 Mentored Patient-Oriented Research Career Development Award (1 K23 AG075257-01). NIH/NCCIH U01AT010462; R01AT010462.

Declaration of Ethical Approval for Study:

This study was approved by the Massachusetts General Hospital Institutional Review Board (IRB) under Protocol # 2020P000095.

Declaration of Informed Consent:

There is no information in the submitted manuscript that can be used to identify patients.