This paper is in the following e-collection/theme issue:

Telehealth and Telemonitoring (1664) Systemic Racism and Racial Bias in Health Care and Society (85) Telerehabilitation (158) Internet for Rehabilitation (74) Technology in Physiotherapy (134) Health Services Research (477) E-Health / Health Services Research and New Models of Care (513)

Published on in Vol 26 (2024)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/65354, first published .
Exploring the Importance of Race and Gender Concordance Between Patients and Physical Therapists in Digital Rehabilitation for Musculoskeletal Conditions: Observational, Longitudinal Study

Exploring the Importance of Race and Gender Concordance Between Patients and Physical Therapists in Digital Rehabilitation for Musculoskeletal Conditions: Observational, Longitudinal Study

Exploring the Importance of Race and Gender Concordance Between Patients and Physical Therapists in Digital Rehabilitation for Musculoskeletal Conditions: Observational, Longitudinal Study

Authors of this article:

Anabela C Areias1 Author Orcid Image ;   Dora Janela1 Author Orcid Image ;   Maria Molinos1 Author Orcid Image ;   Virgílio Bento1 Author Orcid Image ;   Carolina Moreira1, 2 Author Orcid Image ;   Vijay Yanamadala1, 3, 4 Author Orcid Image ;   Steven P Cohen5, 6 Author Orcid Image ;   Fernando Dias Correia1, 7 Author Orcid Image ;   Fabíola Costa1 Author Orcid Image
Article Authors Cited by Tweetations Metrics

    Original Paper

    1Sword Health, Inc, Draper, UT, United States

    2Instituto de Ciências Biomédicas Abel Salazar, Porto, Portugal

    3Department of Surgery, Quinnipiac University Frank H Netter School of Medicine, Hamden, CT, United States

    4Department of Neurosurgery, Hartford Healthcare Medical Group, Westport, CT, United States

    5Northwestern Feinberg School of Medicine, Chicago, IL, United States

    6Walter Reed National Military Medical Center, Uniformed Services University of the Health Sciences, Bethesda, MD, United States

    7Neurology Department, Centro Hospitalar e Universitário do Porto, Porto, Portugal

    Corresponding Author:

    Fabíola Costa, PhD

    Sword Health, Inc

    13937 Sprague Lane, Ste 100

    Draper, UT, 84020

    United States

    Phone: 1 385 308 8034

    Fax:1 801 206 3433

    Email: f.costa@swordhealth.com


    Background: Race/ethnicity and gender concordance between patients and providers is a potential strategy to improve health care interventions. In digital health, where human interactions occur both synchronously and asynchronously, the effect of concordance between patients and providers is unknown.

    Objective: This study aimed to evaluate the impact of race/ethnicity or gender concordance between patients and physical therapists (PTs) in engagement and the clinical outcomes following a digital care program (DCP) in patients with musculoskeletal (MSK) conditions.

    Methods: This secondary analysis of 2 prospective longitudinal studies (originally focused on assessing the acceptance, engagement, and clinical outcomes after a remote DCP) examined the impact of both race/ethnicity concordance and gender concordance between patients and PTs on outcomes for a digital intervention for MSK conditions. Outcomes included engagement (measured by the completion rate and communication, assessed by text interactions), satisfaction, and clinical outcomes (response rate, ie, percentage of patients achieving at least a minimal clinically important change in pain, measured by the Numerical Pain Rating Scale [NPRS]; anxiety, measured by the Generalized Anxiety Disorder 7-item scale [GAD-7]; depression, measured by the Patient Health Questionnaire 9-item [PHQ-9]; and daily activity impairment, measured by the Work Productivity and Activity Impairment [WPAI] questionnaire).

    Results: Of 71,201 patients, 63.9% (n=45,507) were matched with their PT in terms of race/ethnicity, while 61.2% (n=43,560) were matched for gender. Concordant dyads showed a higher completion rate among White (adjusted odds ratio [aOR] 1.11, 95% CI 1.05-1.19, P<.001) and Hispanic (aOR 1.27, 95% CI 1.08-1.54, P=.009) groups, as well as women (aOR 1.10, 95% CI 1.06-1.18, P<.001), when compared to discordant dyads. High and similar levels of interaction between patients and PTs were observed across race/ethnicity and gender dyads, except for Asian concordant dyads (adjusted β coefficient 5.32, 95% CI 3.28-7.36, P<.001). Concordance did not affect satisfaction, with high values (>8.52, 95% CI 8.27-8.77) reported across all dyads. Response rates for pain, anxiety, and daily activity impairment were unaffected by race/ethnicity concordance. An exception was observed for depression, with White patients reporting a higher response rate when matched with PTs from other races/ethnicities (aOR 1.20, 95% CI 1.02-1.39, P=.02). In terms of gender, men had a slightly higher pain response rate in discordant dyads (aOR 1.08, 95% CI 1.01-1.15, P=.03) and a higher depression response rate in concordant dyads (aOR 1.23, 95% CI 1.05-1.47, P=.01).

    Conclusions: Race/ethnicity and gender concordance between patients and PTs does not translate into higher satisfaction or improvement for most clinical outcomes, aside from a positive effect on treatment completion. These results highlight the importance of other PT characteristics, in addition to race/ethnicity or gender concordance, suggesting the potential benefit of experience, languages spoken, and cultural safety training as ways to optimize care.

    Trial Registration: ClinicalTrials.gov NCT04092946, NCT05417685; https://clinicaltrials.gov/study/NCT05417685, https://clinicaltrials.gov/study/NCT04092946

    J Med Internet Res 2024;26:e65354

    doi:10.2196/65354

    Keywords

    musculoskeletal pain; physical therapy; telerehabilitation; eHealth; racial/ethnic concordance; patient–provider concordance; physical therapy; digital rehabilitation; musculoskeletal conditions 


    Individual characteristics, including race and ethnicity, are recognized as major determinants of health according to the World Health Organization. Health disparities among racial and ethnic minorities are well documented, revealing decreased health care access, poorer health status, and prognosis among these populations [1-3].

    Although future projections forecast an increase in population and diversity in the United States [4], diversity within the medical community is not growing at the same pace [5]. This imbalance has raised concerns about the ability to provide culturally competent care and address the specific needs of these patients, when the majority of the workforce is White [6]. As such, the concept of concordance, when a patient and a clinician have the same race/ethnicity or gender, has been explored to understand its potential impact on engagement and clinical outcomes.

    Previous studies, predominantly conducted in primary care and psychology settings, have focused on understanding patients’ perspectives regarding clinicians’ demographics and their impact on therapeutic alliance, communication quality, and overall satisfaction. Findings from these studies are varied: some suggest that patient-clinician racial/ethnic or gender concordance offers benefits [7-11], while others do not [12-14]. Moreover, a recent systematic review studying the association between racial, gender, and multifactorial concordance reported inconclusive results on patient outcomes, advocating for further research [15]. This heterogeneity might be explained by singularities in the population, conditions, and settings under study.

    In digital health, where care can be delivered both synchronously and asynchronously, the impact of concordance between patients and providers remains largely unexplored, particularly in the field of physical therapy. Digital rehabilitation has arisen as a solution to address geographical and time barriers, as well as clinician shortages. Previous research supports its effectiveness among several musculoskeletal (MSK) conditions [16-18] regardless of race/ethnicity [19] and socioeconomic backgrounds [20]. Given the rising adoption of digital rehabilitation programs, and the unique dynamics they introduce to the patient-therapist relationship (eg, duration and frequency of contacts, communication channels used), it is important to assess the association between concordance, engagement, and patient outcomes.

    This study aimed to evaluate the impact of race/ethnicity or gender concordance between patients and physical therapists (PT) in engagement and the clinical outcomes following an asynchronous digital care program (DCP) in patients with MSK conditions in a real-world setting. Additional insights into this topic can enable providers and constituents to develop strategies that consider these aspects.


    Study Design

    This was a secondary analysis of 2 prospective, longitudinal, single-arm investigations focusing on the acceptability, feasibility, engagement, and clinical outcomes of a remote DCP for a population with acute and chronic MSK conditions [20-24]. The overarching aim of the original protocol also included the exploration of associations between the outcomes and patients’ demographic and clinical profiles. Patients participated in this study between January 1, 2022, and December 15, 2023. This study was reported in accordance with STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines (Table S1 in Multimedia Appendix 1).

    Study Population

    Patients, who were beneficiaries of employer health plans, reporting MSK pain in any of the following areas were included in accordance with eligibility criteria in the original investigations: ankle, elbow, hip, knee, low back, neck, shoulder, or wrist/hand. Patients self-reported their race/ethnicity in accordance with 2020 U.S. Census Bureau categories by the U.S. Office of Management and Budget (OMB) [25] as non-Hispanic Asian (hereafter referred to as Asians), non-Hispanic Black or African Americans (hereafter referred to as Black), Hispanic or Latino (hereafter referred to as Hispanic), non-Hispanic White (hereafter referred to as White), American Indian or Alaska Native, and Native Hawaiian or Other Pacific Islander.

    Exclusion criteria defined in the original investigations were (1) health conditions incompatible with >20 minutes of light-to-moderate exercise; (2) ongoing cancer treatment; (3) the presence of red flags, defined as signs or symptoms suggestive of serious pathology (eg, rapid progressive motor weakness or sensory alterations, or bowel or bladder dysfunction), if uncleared by a physician; and (4) missing patient or PT gender or race/ethnicity. Additional exclusion criteria specific for this secondary analysis were patient race/ethnicity from American Indian or Alaska Native and Native Hawaiian or Other Pacific Islander, given the small sample size of these subgroups that precluded statistical analyses. Patients who were compliant with the intervention but missed a given reassessment survey were included.

    Intervention

    This completely remote DCP, developed according to current guidelines [26-28], consisted of home-based exercise, education, and cognitive behavioral therapy (CBT) for up to 12 weeks (at least 3 exercise sessions per week were recommended), as described elsewhere [19,20].

    Briefly, after enrollment, patients completed a baseline form on a dedicated website collecting demographic and clinical information. At the end of this form, patients could choose from a pool of available PTs presented with a photo and a mini biosketch (Figure 1). PTs were diverse in gender, race/ethnicity (Figure S2 in Multimedia Appendix 1), and spoken languages (English, Spanish, French, Portuguese), having a minimum experience of 3 years. All PTs completed cultural safety training, encompassing implicit bias, cultural awareness and humility, motivational interview training and strategies to build rapport with patients from diverse backgrounds. During onboarding, patients had a synchronous video call with their PT for an average duration of 30-60 minutes. During this call, the PT collected further medical history, conducted a clinical assessment, established goals through shared decision-making, and defined a tailored program following a patient-centered approach. Subsequent interactions between the patient and the PT were mostly asynchronous, where the therapeutic relationship was fostered through bidirectional communication via a secure chat within a smartphone app (Figure 1). Nonetheless, additional video calls were scheduled, whenever deemed appropriate by either patients or PTs.

    ‎
    Figure 1. Screenshots of the patient app showcasing the content offered upon onboarding PT selection, during video calls, and during text messaging along the DCP (no real patients are portrayed in this figure). DCP: digital care program; PT: physical therapist.

    Patients received a Food and Drug Administration (FDA)–listed class II medical device composed of a dedicated tablet with a mobile app, motion tracking, and a cloud-based portal. The DCP was conducted autonomously and asynchronously at the patient’s convenience, with exercise sessions being displayed on the tablet through videos with real-time audio and visual biofeedback on performance. The PT monitored and adjusted the program asynchronously through the cloud-based portal. Education covering pathophysiology, pain reconceptualization, active coping skills, the role of exercise, and fear-avoidance behaviors [26-28] was provided through short articles or videos via a smartphone app. The CBT component, delivered through interactive and audio modules via email, focused on mindfulness, acceptance and commitment therapy, and empathy-focused therapy [29].

    Outcome Measures

    Patient engagement was evaluated by the completion rate, defined as the percentage of patients who completed the DCP, and by communication, estimated through the number of text interactions between patients and PTs.

    Program satisfaction was assessed through the question “On a scale from 0 to 10, how likely is it that you would recommend this intervention to a friend or neighbor?”, recorded at the last reassessment. Clinical outcome data were collected at baseline and at the 9th, 18th, and 27th sessions as long as participants reached that milestone.

    Clinical outcomes were analyzed for participants who reached the program end (ie, completers), where the last-available assessment data were used to determine the minimum clinically important change (MCIC), as described in Table 1.

    Table 1. Clinical outcomes and corresponding patient-reported outcome measures with MCICa thresholds.
    Response ratePatient-reported outcome measureMCIC threshold
    PainThe NPRSb specific to the symptomatic body region: “Please rate your average pain over the last 7 days, from 0 (no pain at all) to 10 (worst pain imaginable).”Improvement in pain by at least 30% according to IMMPACT (Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials) guidelines [30]
    Daily activity impairmentWPAIc questionnaire: General Health v2.0 (range 0-100) [31]Reduction of ≥9.1 [23]
    AnxietyGAD-7d (range 0-21) [32,33]Reduction of ≥3.8 [34]
    DepressionPHQ-9e (range 0-27) [33,35]Reduction of ≥5 [36]

    aMCIC: minimum clinically important change.

    bNPRS: Numerical Pain Rating Scale.

    cWPAI: Work Productivity and Activity Impairment.

    dGAD-7: Generalized Anxiety Disorder 7-item scale.

    ePHQ-9: Patient Health Questionnaire 9-item.

    Statistical Analysis

    Descriptive statistics were reported using means and proportions. Comparisons between concordant and discordant groups were performed using independent-sample t tests or Mann-Whitney U tests for continuous variables (with Bonferroni correction) or chi-squared tests for categorical variables. Completion and clinical outcomes were binary variables, while satisfaction was a continuous variable. Two predictive variables describing concordance on race/ethnicity or on gender were created. Concordance was defined as an exact match on race/ethnicity between the patient and the PT, or a match on gender, depending on the variable of interest. Response rates were calculated using the last patient reassessment. In the case of missing reassessment at the discharge time point, the last available reassessment was carried forward.

    Due to variability in the distribution of members allocated to each PT, a consequence of some PTs having been in the study longer, using PTs as random effects was not feasible. Generalized linear or logistic regression models were used to assess the effects of concordance on both race/ethnicity and gender on the described outcomes. Models of race/ethnicity concordance included an interaction term between concordance and patient race/ethnicity to explore the effect of concordance on outcomes stratified by race/ethnicity. Similarly, models of gender concordance included an interaction term between concordance and patient gender to explore the effect of concordance on outcomes stratified by gender. Given the baseline differences observed between groups in variables previously reported as predictors of recovery in patients with MSK conditions after physical therapy [37-39], all models were adjusted for the BMI (body mass index); age; education levels (dichotomized as low [no college degree] or high [at least a bachelor’s degree]); employment status (dichotomized as employed either full- or part-time or not currently employed); rurality; acuity (acute: pain<3 months; chronic: pain≥3 months) [40]); and baseline levels of pain, anxiety, and depression. Models assessing clinical outcomes were also adjusted for the discharged time point. A false discovery rate was used to adjust for multiple comparisons, and statistical significance was set at P<.05. Descriptive statistics were performed using commercially available software (IBM SPSS v22), and statistical analyses were performed using R Studio version 2023.09.1+494 (R Foundation for Statistical Computing).

    Ethical Considerations

    All research was performed in accordance with relevant guidelines and regulations set by the Declaration of Helsinki. The original studies were prospectively approved by the New England (120190313) and Advarra (Pro00063337) independent Institutional Review Boards (IRBs) and registered on ClinicalTrials (NCT04092946, NCT05417685) on September 17, 2019, and June 14, 2022, respectively. All patients provided informed consent. Both IRB approvals and respective original consent covered secondary analysis without need for additional consent.

    Prior to data analysis, all collected data underwent a rigorous anonymization process to safeguard the privacy of participants involved in the research. All individuals whose images are included in this publication (Figure 1) provided explicit written consent for the display of identifiable features in this manuscript, in accordance with journal guidelines and ethical standards. All data collection and analysis methods complied with relevant guidelines and regulations. Participants were not provided with any form of compensation.


    Participant Details

    Of 84,326 eligible participants, 71,201 (84.4%) started the program, of which 49,877 (70.1%) completed the intervention (Figure 2). Of those who started the program, 63.9% (n=45,507) matched their PT’s race/ethnicity, while 61.2% (n=43,560) shared the same gender. Patients’ demographic and clinical characteristics stratified by race/ethnicity and gender concordance are reported in Table S2 in Multimedia Appendix 1. Race/ethnicity and gender distributions of the PTs (Figure S1 in Multimedia Appendix 1) were similar to the membership reported by the American Physical Therapy Association (APTA). The cohort was predominantly composed of middle-aged (mean 48.7 years, SD 11.8) women (n=42,464, 59.6%), who identified as White (n=49,920, 70.1%) and had a BMI above 30 (n=28,337, 39.9%). These individuals tended to have a high level of education (n=45,483, 63.8%), lived in urban locations (n=62,405, 87.6%), and reported full-time employment (n=61,298, 86.1%).

    ‎
    Figure 2. Study flowchart.

    Regarding race/ethnicity, concordant and discordant groups were different in some demographic characteristics, such as age (mean 46.9, SD 11.5, vs mean 49.7, SD 11.8; P<.001), employment status (employed: n=22,898, 89.1%, vs n=38,400, 84.4%; P<.001), and rurality (n=1761, 6.9%, vs n=6708, 14.7%; P<.001). Differences were also observed among baseline clinical data, namely acuity (chronic: n=19,461, 75.9%, vs n=36,013, 79.3%; P<.001) and pain (mean 4.9, SD 2.0, vs mean 4.7, SD 1.9; P<.001).

    Regarding gender, the concordant group had a larger proportion of women (n=33,854, 77.7%, vs n=8610, 31.1%; P<.001), patients working part-time (n=2215, 5.1%, vs n=770, 2.8%; P<.001), and patients with depression (mean 9.6, SD 4.3, vs mean 9.4, SD 4.2; P=.004) and daily activity impairment (mean 37.2, SD 22.8, vs mean 35.8, SD 22.4; P<.001).

    Engagement Outcomes

    The effect of race/ethnicity and gender concordance/discordance on completion rates and PT-patient interactions is presented in Figure 3 and Table 2, respectively. Both race/ethnicity and gender concordance were positively associated with the completion rate when compared to discordant dyads, particularly for those who identified as White (adjusted odds ratio [aOR] 1.11, 95% CI 1.05-1.19, P<.001), Hispanic (aOR 1.27, 95% CI 1.08-1.54, P=.009), and women (aOR 1.10, 95% CI 1.06-1.18, P<.001), as shown in Figure 3. PT-patient interactions were consistently high across the different race and gender dyads. Although there was a statistically significant difference between concordant and discordant dyads, this may not represent a clinically significant difference, as this variation ranged from 1 to 3 text interactions. An exception was observed among Asian patients, where a wider difference was observed, with discordant dyads reporting significantly higher interactions (adjusted β coefficient 5.32, 95% CI 3.28-7.36) compared to concordant dyads.

    ‎
    Figure 3. The aORs for concordance stratified by patient race/ethnicity and gender (reference: PT-patient concordance) for the following outcomes: (A) completion rate, (B) pain response rate, (C) daily activity response rate, (D) depression response rate; and (E) anxiety response rate. Statistically significant aORs are depicted in red. aOR: adjusted odds ratio.
    Table 2. Generalized linear model estimates of text interactions between patients and PTsa stratified by patient race/ethnicity and gender (reference: PT-patient concordance).
    VariableConcordance, mean (95% CI); n (%)Discordance, mean (95% CI); n (%)Adjusted β coefficient (95% CI)P-valueb
    Race/ethnicity

    		White22.76 (22.13 to 23.39); 43,523 (87.2)20.10 (19.39 to 20.80); 6397 (12.8)–2.66 (–3.03 to –2.30)<.001

    		Asian16.76 (14.65 to 18.87); 254 (3.6)22.08 (21.37 to 22.78); 6773 (96.4)5.32 (3.28 to 7.36)<.001

    		Black20.17 (19.12 to 21.23); 1061 (14.4)21.53 (20.82 to 22.24); 6284 (85.6)1.36 (0.44 to 2.28).004

    		Hispanic19.48 (18.31 to 20.65); 669 (9.7)21.20 (20.50 to 21.91); 6240 (90.3)1.72 (0.67 to 2.77).001
    Gender

    		Women23.07 (22.44 to 23.70); 33,854 (79.7)21.57 (20.89 to 22.25); 8610 (20.3)–1.50 (–1.83 to –1.17)<.001

    		Men20.60 (19.93 to 21.27); 9706 (66.2)21.44 (20.80 to 22.09); 19,031 (33.8)0.84 (0.51 to 1.17)<.001

    aPT: physical therapist.

    bAll P values were significant.

    Satisfaction

    Program satisfaction was high across different races/ethnicities (mean 8.52, 95% CI 8.27-8.77, to mean 9.37, 95% CI 9.18-9.55) regardless of concordance status (all P>.05; Table 3), except for Asian patients who were statistically more likely to report higher satisfaction when matched in discordant dyads (adjusted β coefficient 0.36, 95% CI 0.12-0.59, P=.003; reference: concordance). Men showed a statistically higher likelihood to report satisfaction when matched in discordant dyads (adjusted β coefficient 0.15, 95% CI 0.10-0.20, P<.001; reference: concordance). In both cases (Asian patients and men), differences were small and therefore might not be clinically significant.

    Table 3. Generalized linear model estimates of satisfaction level by patient race/ethnicity and gender (reference: PTa-patient concordance).
    VariableConcordance, mean (95% CI); n (%)Discordance, mean (95% CI); n (%)Adjusted β coefficient (95% CI)P-value
    Race/ethnicity

    		White8.93 (8.84 to 9.02); 43,523 (87.2)8.88 (8.78 to 8.98); 6397 (12.8)–0.04 (–0.09 to 0.01).08

    		Asian8.52 (8.27 to 8.77); 254 (3.6)8.88 (8.78 to 8.98); 6773 (96.4)0.36 (0.12 to 0.59).003b

    		Black9.17 (9.02 to 9.31); 1061 (14.4)9.19 (9.09 to 9.29); 6284 (85.6)0.02 (–0.11 to 0.15).76

    		Hispanic9.37 (9.18 to 9.55); 669 (9.7)9.24 (9.14 to 9.34); 6240 (90.3)–0.13 (–0.30 to 0.04).13
    Gender

    		Women9.06 (8.97 to 9.15); 33,854 (79.7)9.07 (8.98 to 9.17); 8610 (20.3)0.01 (–0.03 to 0.06).61

    		Men8.68 (8.58 to 8.77); 9706 (33.8)8.82 (8.73 to 8.92); 19,031 (66.2)0.15 (0.10 to 0.20)<.001b

    aPT: physical therapist.

    bSignificant P values.

    Clinical Outcomes

    The impact of race/ethnicity and gender on clinical outcomes is presented in Figure 3 and Table S3 in Multimedia Appendix 1.

    Significant and similar pain response rates (adjusted: 62.16%, 95% CI 52.45-70.99, to 68.88%, 95% CI 63.10-74.12) were observed across different races/ethnicities. Regarding gender, men reported a slightly higher response rate when supported by women PTs (discordant dyads) in relation to concordant dyads (aOR 1.08, 95% CI 1.01-1.15, P=.03). Alongside pain recovery, daily activity impairment followed a similar trend, with high and significant response rates across all races/ethnicities and genders (69.03%, 95% CI 56.94-78.98 to 74.46%, 95% CI 70.81-77.81) regardless of concordance (all P>.05; Figure 3). Anxiety was similarly not impacted by concordance. In contrast, White patients had a higher depression response rate when supported by PTs from other races/ethnicities (aOR 1.20, 95% CI 1.02-1.39; reference: concordance), while men supported by men PTs were more likely to report significant depression recovery than discordant dyads (aOR 1.23, 95% CI 1.05-1.47, P=.01; reference: discordance).

    Effect of Covariates on Outcomes

    The effect of covariates on outcomes among concordant/discordant dyads is depicted in Figure S2 in Multimedia Appendix 1 for race/ethnicity and Figure S3 in Multimedia Appendix 1 for gender.

    Several covariates had an effect on the completion rate in both analyses (race/ethnicity concordance and gender concordance). A higher BMI, pain, depression and anxiety levels, and lower education levels were associated with a lower completion rate. In contrast, older age and not being employed (unemployed or retired) were associated with a higher completion rate. A lower pain response rate was observed across those with a higher BMI, lower education, and higher mental health scores. Chronic conditions and not being employed were also associated with smaller improvements in clinical outcomes.


    Principal Findings

    The findings of this study demonstrated that being concordant on race/ethnicity and gender did not translate into higher satisfaction or better clinical outcomes in asynchronous digital care, despite the positive effect on treatment completion. Concordant dyads were associated with higher completion rates when compared to their counterparts, particularly for White participants (aOR 1.11, 95% CI 1.05-1.19, P<.001), Hispanic participants (aOR 1.27, 95% CI 1.08-1.54, P=.009), and women (aOR 1.10, 95% CI 1.06-1.18, P<.001). Race and gender dyads had frequent interactions, ranging from 16.8 to 23.01, with a minimal difference between groups. An exception was observed among Asian discordant dyads, which exchanged an adjusted mean of 5.32 more texts (95% CI 3.28-7.36, P<.001) than concordant dyads.

    Program satisfaction was high regardless of concordance status (>8.52, 95% CI 8.27-8.77). Race/ethnicity concordance did not impact clinical outcomes for pain, anxiety, and daily activity impairment, as all improved similarly (eg, pain response rates ranged from 62.16%, 95% CI 52.45-70.99, to 68.88%, 95% CI 63.10-74.12, across different race/ethnicity dyads). An exception was observed for depression, in which White patients experienced a higher success rate when matched with PTs from other races/ethnicities (aOR 1.20, 95% CI 1.02-1.39; reference: concordance). In terms of gender concordance, men had a slightly higher pain response rate in discordant dyads (aOR 1.08, 95% CI 1.01-1.15, P=.03; reference: concordance) and a higher depression response rate in concordant dyads (aOR 1.23, 95% CI 1.05-1.47, P=.01; reference: discordance).

    Comparison With the Literature

    Interactions in a digital context have distinct characteristics as they are primarily facilitated through video calls and text messaging. These forms of communication have been shown to reduce fears of embarrassment and foster a sense of support [41,42], as providers are readily accessible. However, it remains unclear whether the demographic characteristics of providers, and their concordance with patients, significantly influence intervention outcomes. Studies evaluating the effect of race/ethnicity or gender concordance on outcomes have been mainly conducted in primary/outpatient care [7-9,43,44] and behavioral health [45-47], making this study the first of its kind in the context of remote digital health. Herein, we explored this association in a large cohort whose demographic characteristics matched the US general population [48], particularly those reporting MSK pain [49]. Analyses were adjusted for key baseline characteristics that differed statistically between concordant and discordant groups and were previously reported to influence outcomes [37-39]. This approach reduces bias and permits a clearer understanding of the effect of racial/ethnic and gender concordance on outcomes. Our findings showed that concordance does not have a uniformly significant impact across all race/ethnicity, gender, and outcomes, as previously reported [15,50]. Importantly, the differences noted in some variables did not form a sustained pattern that can be operationalized in clinical practice.

    Herein, concordance seemed to support higher completion rates, particularly among White and Hispanic groups and among women. Albeit in a different setting, a previous study reported that race concordance is important for retention in psychology interventions in adolescents [47]. Points of similarity may help enhance therapeutic relationships, resulting in a more trusting patient, willing to commit to the therapeutic plan developed collaboratively [51,52]. For example, women patients may find it easier to trust women PTs, who in turn may be more likely to empathize with them, given their common backgrounds [53].

    High interaction levels were observed across all race/ethnicity and gender dyads, which are suggestive of therapeutic alliance establishment regardless of concordance. Further investigation will allow us to fully understand which factors of asynchronous communication may contribute more significantly to rapport building and therapeutic alliance establishment.

    Similarly, high satisfaction was reported across dyads but was not generally impacted by concordance. Asian patients also reported slightly higher satisfaction when paired in discordant dyads, although this difference was small (adjusted β coefficient 0.36, 95% CI 0.12-0.59, P=.003), suggesting that it may lack clinical significance. The reason underlying these findings is unclear but may be related to unmeasured confounders. One can speculate that the diversity within the U.S. Census “Asian” category, encompassing groups ranging from Koreans to Asian Indians, may not reflect the true cultural or linguistic heterogeneity of the group [50,54]. The literature evaluating the effect of racial concordance on satisfaction has also shown inconclusive and contradictory findings [15].

    Race/ethnicity concordance was not associated with significant improvements in most clinical outcomes (pain, daily activity impairment, and anxiety). An exception was depression, in which PTs’ demographic characteristics may have helped foster better outcomes. White discordant dyads reported better depression response rates, suggesting that although shared identity may facilitate a trustworthy relationship and engagement with the treatment, it is not the prime determining factor in sustaining a significant clinical outcome [15]. Additional factors, in addition to provider background, can carry significant weight in determining outcomes. In studies examining outcome predictors across nearly all pain conditions and therapies, the risk factors for treatment failure almost entirely comprise patient-related variables, including psychiatric morbidities, socioeconomic factors, or other clinical characteristics (eg, secondary gain, obesity, and smoking) [55-57]. Although some of these variables are modifiable, many are not, and even for those which are, addressing them may be time-consuming, prohibitively expensive, and fraught with other roadblocks.

    Herein, the clinical team consisted of experienced clinicians, many with subspecialty training, all trained in equity, cultural humility, and unconscious bias, which may have reduced any subtle effects of concordance to the point of being undetectable with our study design [58,59]. Following Centers for Disease Control and Prevention (CDC) recommendations, training in cultural safety and humility is vital for empowering clinicians to adapt to diverse cultural groups, mitigating implicit racial bias and imbalances in patient-provider relationships, and promoting patient-centered care [44,60,61]. This is especially important in the light of the scarcity of diversity among providers compared to the general population, which is reflected in the physical therapy workforce [62,63].

    Regarding the effect of gender concordance on clinical outcomes, only depression was influenced by concordance, with men showing a greater recovery trajectory when matched with a man PT. The reasons behind this finding are unknown, and the literature evaluating the effect of gender concordance on depression and other psychiatric outcomes is scarce. Depression is often deeply rooted in emotional beliefs and experiences [64], where societal and cultural aspects might play a significant role and influence patient-provider interactions [52]. Men might have felt more open and less stigmatized to share and discuss mental health issues when treated by a PT sharing the same gender identity. Interestingly, in the context of psychotherapy, men seem to exhibit no preference for therapist gender [45]. This evidence suggests that preferences observed in one clinical field may not necessarily apply to others and that caution is needed when generalizing findings across different outcomes, settings, or specialties, as the effects of concordance may be unique to each outcome and specific context.

    Strengths and Limitations

    This study contains several important strengths: (1) the topic itself—the effect of race/ethnicity and gender concordance on engagement and clinical outcomes in the context of digital health using a fully remote DCP, laying the foundation for future research on this issue in the context of digital health; (2) the large sample size and proportional representation of patients from minority backgrounds, addressing a limitation identified previously [7,65]; and (3) the inclusion of validated outcome measures, previously lacking in comparable literature [50].

    This study also displays limitations that warrant discussion. Considering that this DCP was mostly asynchronous, the results may not be generalizable for typical in-person or synchronous telerehabilitation contexts, where the face-to-face interaction between PTs and members is expected to be higher. Although the analysis included a wide range of patient demographic and clinical characteristics, it did not account for other variables that influence a patient’s experience with the provider (eg, years of experience, subspecialty training, and personality traits and compatibility). Additional subjective measures, such as empathy and cultural competency, which are difficult to capture, may have also impacted outcomes. Another limitation is the heterogeneity within race/ethnicity categories, particularly among Asian patients (which includes widely diverse groups). Although we considered adjusting for the random effects of the PT, the high variability of patients allocated to each PT precluded this analysis and should be addressed in future studies. Additionally, although potential bias from missing data cannot be ruled out, using the last observation carried forward provides the most conservative results.

    Future research should focus on the impact of communication channels, cadence, and duration on the therapeutic alliance, particularly when comparing different delivery models (eg, in-person physical therapy). Additional research should address the impact of training (cultural, biased, or otherwise) and other PT characteristics as potential predictors of recovery outcomes in physical therapy. Another area for investigation is the concept of intersectionality in concordance (eg, multiracial and lesbian, gay, bisexual, transgender, queer or questioning, intersex, asexual, and other [LGBTQIA+] identity). Understanding these aspects is crucial, given the current demographic trends in the US and world populations, and could lead to more personalized and effective care strategies.

    Conclusion

    Race/ethnicity and gender concordance between patients and PTs translated into higher treatment completion but did not impact satisfaction. For some clinical outcomes, discordant dyads reported greater improvements, suggesting that addressing provider diversity shortages requires more than just matching race/ethnicity or gender. Yet, the findings of this study did not identify a consistent pattern, suggesting that PTs’ characteristics, namely cultural safety training, experience, subspecialty training, and linguistic ability, may significantly impact engagement and outcomes in diverse populations. Future interventions should focus on these aspects to better understand their potential in optimizing patient experience and outcomes for MSK conditions.

    Acknowledgments

    The authors extend their gratitude to the team of physical therapists for their diligent participant management. The authors also acknowledge the contributions of João Tiago Silva, Margarida Morais, and Guilherme Freches (Sword Health, Inc) to data validation and Ashley Bass (former Sword Health, Inc, employee) for diversity, equity, and inclusion insights. This work was supported by Sword Health, Inc, and also developed within the scope of project number 62 (“Responsible AI”), financed by European Union funds, namely the Recovery and Resilience Plan—“Componente 5: Agendas Mobilizadoras para a Inovação Empresarial,” included in the NextGenerationEU funding program.

    Data Availability

    The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to privacy restrictions.

    Authors' Contributions

    Conceptualization was managed by FC; data curation by ACA and MM; formal analysis by ACA; writing—original draft preparation by ACA, DD, and FC; funding by VB; and interpretation of data and writing—reviewing and editing by all authors. All authors have read and agreed to the published version of the manuscript.

    Conflicts of Interest

    ACA, DJ, MM, CM, VY, FDC, and FC were employed by the company Sword Health, Inc, the sponsor of this study. VB, VY, and FDC also hold equity in Sword Health, Inc, and VB is the chief executive officer of the same company. SPC is an independent scientific and clinical consultant who received adviser honorarium from Sword Health, Inc. The study sponsor, Sword Health, Inc, was involved in the study design, data collection and interpretation, and writing of the manuscript.

    Multimedia Appendix 1

    Table S1: STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting checklist. Supplementary. Table S2: Baseline characteristics for the overall cohort and stratified by race/ethnicity concordance. Figure S1: Comparison of distribution of physical therapists between the study clinical team and the American Physical Therapy Association (APTA) in terms of (A) race/ethnicity and (B) gender. The study clinical team demographics were obtained from employer records. Table S3: Generalized linear model analysis of clinical outcome estimates by patient race and gender (reference: physical therapist-patient concordance). Figure S2: Effect of covariates on the outcome within the race concordance/discordance analysis. Figure S3. Effect of covariates on the outcome within the gender concordance/discordance analysis.

    PDF File (Adobe PDF File), 604 KB
    1. Lane E, Barnes C, Fritz JM. Differences in pain experience among different racial and ethnic groups. Phys Ther. Jan 05, 2024:pzae001. [CrossRef] [Medline]
    2. Odonkor C, Esparza R, Flores L, Verduzco-Gutierrez M, Escalon M, Solinsky R, et al. Disparities in Health Care for Black Patients in Physical Medicine and Rehabilitation in the United States: A Narrative Review. PM&R. Oct 08, 2020;13(2):180-203. [CrossRef] [Medline]
    3. Richter RR, Chrusciel T, Salsich G, Austin T, Scherrer JF. Disparities exist in physical therapy utilization and time to utilization between black and white patients with musculoskeletal pain. Phys Ther. Oct 06, 2022;102(10):pzac095. [CrossRef] [Medline]
    4. Vespa J, Medina L, Armstrong D. Demographic turning points for the United States: population projections for 2020 to 2060. U.S. Government Publishing Office. 2018. URL: https://www.govinfo.gov/app/details/GOVPUB-C3-PURL-gpo93743 [accessed 2024-07-16]
    5. Salsberg E, Richwine C, Westergaard S, Portela Martinez M, Oyeyemi T, Vichare A, et al. Estimation and Comparison of Current and Future Racial/Ethnic Representation in the US Health Care Workforce. JAMA Netw Open. Mar 01, 2021;4(3):e213789. [FREE Full text] [CrossRef] [Medline]
    6. Xierali IM, Nivet MA. The racial and ethnic composition and distribution of primary care physicians. J Health Care Poor Underserved. Jan 01, 2018;29(1):556-570. [FREE Full text] [CrossRef] [Medline]
    7. Takeshita J, Wang S, Loren AW, Mitra N, Shults J, Shin DB, et al. Association of Racial/Ethnic and Gender Concordance Between Patients and Physicians With Patient Experience Ratings. JAMA Netw Open. Nov 02, 2020;3(11):e2024583. [FREE Full text] [CrossRef] [Medline]
    8. Rogo-Gupta LJ, Haunschild C, Altamirano J, Maldonado YA, Fassiotto M. Physician gender is associated with Press Ganey patient satisfaction scores in outpatient gynecology. Womens Health Issues. May 2018;28(3):281-285. [CrossRef] [Medline]
    9. Ku L, Vichare A. The association of racial and ethnic concordance in primary care with patient satisfaction and experience of care. J Gen Intern Med. Jun 10, 2022;38(3):727-732. [FREE Full text] [CrossRef] [Medline]
    10. Alexander J, Hearld L, Mittler JN. Patient–physician role relationships and patient activation: the moderating effects of race and ethnicity. Med Care Res Rev. Jul 15, 2014;71(5):472-495. [CrossRef] [Medline]
    11. Bhati KS. Effect of client-therapist gender match on the therapeutic relationship: an exploratory analysis. Psychol Rep. Oct 01, 2014;115(2):565-583. [CrossRef] [Medline]
    12. Jerant A, Bertakis KD, Fenton JJ, Tancredi DJ, Franks P. Patient-provider sex and race/ethnicity concordance: a national study of healthcare and outcomes. Med Care. Nov 2011;49(11):1012-1020. [CrossRef] [Medline]
    13. Behn A, Davanzo A, Errázuriz P. Client and therapist match on gender, age, and income: does match within the therapeutic dyad predict early growth in the therapeutic alliance? J Clin Psychol. Mar 24, 2018;74(9):1403-1421. [CrossRef] [Medline]
    14. Sweeney CF, Zinner D, Rust G, Fryer GE. Race/ethnicity and health care communication: does patient-provider concordance matter. Med Care. Nov 2016;54(11):1005-1009. [CrossRef] [Medline]
    15. Otte SV. Improved patient experience and outcomes: is patient-provider concordance the key? J Patient Exp. May 29, 2022;9:23743735221103033. [FREE Full text] [CrossRef] [Medline]
    16. Cui D, Janela D, Costa F, Molinos M, Areias AC, Moulder RG, et al. Randomized-controlled trial assessing a digital care program versus conventional physiotherapy for chronic low back pain. NPJ Digit Med. Jul 07, 2023;6(1):121. [FREE Full text] [CrossRef] [Medline]
    17. Areias AC, Costa F, Janela D, Molinos M, Moulder RG, Lains J, et al. Long-term clinical outcomes of a remote digital musculoskeletal program: an ad hoc analysis from a longitudinal study with a non-participant comparison group. Healthcare. Nov 23, 2022;10(12):2349. [FREE Full text] [CrossRef] [Medline]
    18. Janela D, Costa F, Weiss B, Areias AC, Molinos M, Scheer JK, et al. Effectiveness of biofeedback-assisted asynchronous telerehabilitation in musculoskeletal care: a systematic review. Digit Health. Jun 09, 2023;9:20552076231176696. [FREE Full text] [CrossRef] [Medline]
    19. Scheer J, Costa F, Molinos M, Areias A, Janela D, Moulder RG, et al. Racial and ethnic differences in outcomes of a 12-week digital rehabilitation program for musculoskeletal pain: prospective longitudinal cohort study. J Med Internet Res. Oct 31, 2022;24(10):e41306. [FREE Full text] [CrossRef] [Medline]
    20. Areias AC, Molinos M, Moulder RG, Janela D, Scheer JK, Bento V, et al. The potential of a multimodal digital care program in addressing healthcare inequities in musculoskeletal pain management. npj Digit Med. Oct 10, 2023;6(1):188. [FREE Full text] [CrossRef] [Medline]
    21. Costa F, Janela D, Molinos M, Lains J, Francisco GE, Bento V, et al. Telerehabilitation of acute musculoskeletal multi-disorders: prospective, single-arm, interventional study. BMC Musculoskelet Disord. Jan 04, 2022;23(1):29. [FREE Full text] [CrossRef] [Medline]
    22. Janela D, Costa F, Areias A, Molinos M, Moulder R, Lains J, et al. Digital care programs for chronic hip pain: a prospective longitudinal cohort study. Healthcare. Aug 22, 2022;10(8):1595. [CrossRef] [Medline]
    23. Areias AC, Costa F, Janela D, Molinos M, Moulder RG, Lains J, et al. Impact on productivity impairment of a digital care program for chronic low back pain: a prospective longitudinal cohort study. Musculoskelet Sci Pract. Feb 2023;63:102709. [FREE Full text] [CrossRef] [Medline]
    24. Areias AC, Janela D, Moulder RG, Molinos M, Bento V, Moreira C, et al. Applying AI to Safely and Effectively Scale Care to Address Chronic MSK Conditions. J Clin Med. Jul 26, 2024;13(15):4366. [FREE Full text] [CrossRef] [Medline]
    25. Measuring racial and ethnic diversity for the 2020 census. United States Census Bureau. 2021. URL: https:/​/www.​census.gov/​newsroom/​blogs/​random-samplings/​2021/​08/​measuring-racial-ethnic-diversity-2020-census.​html [accessed 2024-07-16]
    26. Kolasinski SL, Neogi T, Hochberg MC, Oatis C, Guyatt G, Block J, et al. 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee. Arthritis Care Res. Jan 06, 2020;72(2):149-162. [CrossRef] [Medline]
    27. National Guideline Centre (UK). Low back pain and sciatica in over 16s: assessment and management. National Institute for Health and Care Excellence (NICE). London, United Kingdom.; 2016. URL: https:/​/www.​nice.org.uk/​guidance/​ng59/​resources/​low-back-pain-and-sciatica-in-over-16s-assessment-and-management-pdf-1837521693637 [accessed 2023-11-27]
    28. Qaseem A, Wilt TJ, McLean RM, Forciea MA. Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians. Ann Intern Med. Feb 14, 2017;166(7):514-530. [CrossRef] [Medline]
    29. Williams ACDC, Fisher E, Hearn L, Eccleston C. Psychological therapies for the management of chronic pain (excluding headache) in adults. Cochrane Database Syst Rev. Aug 12, 2020;8:CD007407. [CrossRef] [Medline]
    30. Dworkin RH, Turk DC, Wyrwich KW, Beaton D, Cleeland CS, Farrar JT, et al. Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations. J Pain. Feb 2008;9(2):105-121. [CrossRef] [Medline]
    31. Ospina MB, Dennett L, Waye A, Jacobs P, Thompson AH. A systematic review of measurement properties of instruments assessing presenteeism. Am J Manag Care. Feb 01, 2015;21(2):e171-e185. [FREE Full text] [Medline]
    32. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. May 22, 2006;166(10):1092-1097. [CrossRef] [Medline]
    33. Bijker L, Sleijser-Koehorst MLS, Coppieters MW, Cuijpers P, Scholten-Peeters GGM. Preferred self-administered questionnaires to assess depression, anxiety and somatization in people with musculoskeletal pain - a modified Delphi study. J Pain. Mar 2020;21(3-4):409-417. [FREE Full text] [CrossRef] [Medline]
    34. Toussaint A, Hüsing P, Gumz A, Wingenfeld K, Härter M, Schramm E, et al. Sensitivity to change and minimal clinically important difference of the 7-item Generalized Anxiety Disorder questionnaire (GAD-7). J Affect Disord. Mar 15, 2020;265:395-401. [CrossRef] [Medline]
    35. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. Sep 2001;16(9):606-613. [FREE Full text] [CrossRef] [Medline]
    36. Löwe B, Unützer J, Callahan CM, Perkins AJ, Kroenke K. Monitoring depression treatment outcomes with the Patient Health Questionnaire-9. Med Care. Dec 2004;42(12):1194-1201. [CrossRef] [Medline]
    37. Artus M, Campbell P, Mallen CD, Dunn KM, van der Windt DAW. Generic prognostic factors for musculoskeletal pain in primary care: a systematic review. BMJ Open. Jan 17, 2017;7(1):e012901. [FREE Full text] [CrossRef] [Medline]
    38. Verkerk K, Luijsterburg PAJ, Heymans MW, Ronchetti I, Pool‐Goudzwaard A, Miedema HS, et al. Prognosis and course of pain in patients with chronic non‐specific low back pain: a 1‐year follow‐up cohort study. Eur J Pain. Jan 06, 2015;19(8):1101-1110. [CrossRef] [Medline]
    39. Nieminen LK, Pyysalo LM, Kankaanpää MJ. Prognostic factors for pain chronicity in low back pain: a systematic review. Pain Rep. Jan 01, 2021;6(1):e919. [FREE Full text] [CrossRef] [Medline]
    40. Treede R-D, Rief W, Barke A, Aziz Q, Bennett MI, Benoliel R, et al. A classification of chronic pain for ICD-11. Pain. Jun 2015;156(6):1003-1007. [FREE Full text] [CrossRef] [Medline]
    41. Lewis E, Hassmén P, Pumpa KL. Participant perspectives of a telehealth trial investigating the use of telephone and text message support in obesity management: a qualitative evaluation. BMC Health Serv Res. Jul 09, 2021;21(1):675. [FREE Full text] [CrossRef] [Medline]
    42. Huxley CJ, Atherton H, Watkins JA, Griffiths F. Digital communication between clinician and patient and the impact on marginalised groups: a realist review in general practice. Br J Gen Pract. Nov 29, 2015;65(641):e813-e821. [CrossRef] [Medline]
    43. Walker RJ, Dawson AZ, Thorgerson A, Campbell JA, Engel S, Kastner M, et al. Relationship between satisfaction scores and racial/ethnic and sex concordance in primary care. Healthcare. Aug 12, 2023;11(16):2276. [FREE Full text] [CrossRef] [Medline]
    44. Lin OM, Reid HW, Fabbro RL, Johnson KS, Batch BC, Olsen MK, et al. Association of provider perspectives on race and racial health care disparities with patient perceptions of care and health outcomes. Health Equity. Jul 01, 2021;5(1):466-475. [FREE Full text] [CrossRef] [Medline]
    45. Ilagan GS, Heatherington L. Advancing the understanding of factors that influence client preferences for race and gender matching in psychotherapy. Couns Psychol Q. Jul 29, 2021;35(3):694-717. [CrossRef]
    46. Cheng AW, Nakash O, Cruz-Gonzalez M, Fillbrunn MK, Alegría M. The association between patient–provider racial/ethnic concordance, working alliance, and length of treatment in behavioral health settings. Psychol Serv. 2023;20(Suppl 1):145-156. [CrossRef] [Medline]
    47. Wintersteen MB, Mensinger JL, Diamond GS. Do gender and racial differences between patient and therapist affect therapeutic alliance and treatment retention in adolescents? Prof Psychol Res Pract. Aug 2005;36(4):400-408. [CrossRef]
    48. S0201 selected population profile in the United States. United States Census Bureau. 2022. URL: https://data.census.gov/table?t=-00:-01:Race%20and%20Ethnicity [accessed 2024-07-31]
    49. Burden of musculoskeletal diseases in the United States (BMUS). United States Bone Joint Initiative. 3rd Edition. 2015. URL: https:/​/www.​usbji.org/​index.​php?option=com_content&view=article&id=50:burden-of-musculoskeletal-diseases-in-the-united-states&catid=20:site-content&Itemid=148
    50. Miller AN, Duvuuri VNS, Vishanagra K, Damarla A, Hsiao D, Todd A, et al. The relationship of race/ethnicity concordance to physician-patient communication: a mixed-methods systematic review. Health Commun. Jun 20, 2023;39(8):1543-1557. [CrossRef] [Medline]
    51. Cruwys T, Lee G, Robertson A, Haslam C, Sterling N, Platow M, et al. Therapists who foster social identification build stronger therapeutic working alliance and have better client outcomes. Compr Psychiatry. Jul 01, 2023;124:152394. [FREE Full text] [CrossRef] [Medline]
    52. Eken HN, Dee EC, Powers AR, Jordan A. Racial and ethnic differences in perception of provider cultural competence among patients with depression and anxiety symptoms: a retrospective, population-based, cross-sectional analysis. Lancet Psychiatry. Nov 2021;8(11):957-968. [FREE Full text] [CrossRef] [Medline]
    53. Hall J, Roter D. Do patients talk differently to male and female physicians? A meta-analytic review. Patient Educ Couns. Dec 01, 2002;48(3):217-224. [CrossRef] [Medline]
    54. Gordon NP, Lin TY, Rau J, Lo JC. Aggregation of Asian-American subgroups masks meaningful differences in health and health risks among Asian ethnicities: an electronic health record based cohort study. BMC Public Health. Nov 25, 2019;19(1):1551. [FREE Full text] [CrossRef] [Medline]
    55. Cohen SP, Doshi TL, Kurihara C, Reece D, Dolomisiewicz E, Phillips CR, et al. Multicenter study evaluating factors associated with treatment outcome for low back pain injections. Reg Anesth Pain Med. Dec 08, 2021;47(2):89-99. [FREE Full text] [CrossRef] [Medline]
    56. Wilkens P, Scheel IB, Grundnes O, Hellum C, Storheim K. Prognostic factors of prolonged disability in patients with chronic low back pain and lumbar degeneration in primary care: a cohort study. Spine. Jan 01, 2013;38(1):65-74. [CrossRef] [Medline]
    57. Linton SJ, Shaw WS. Impact of psychological factors in the experience of pain. Phys Ther. May 2011;91(5):700-711. [CrossRef] [Medline]
    58. Dickson T, Wilkinson T, Hegedus E, Gisselman A. Transforming the patient experience: moving beyond cultural competence to cultural safety. Phys Ther. Jun 03, 2022;102(6):pzac034. [CrossRef] [Medline]
    59. Lekas H, Pahl K, Fuller Lewis C. Rethinking cultural competence: shifting to cultural humility. Health Serv Insights. Dec 20, 2020;13:1178632920970580. [FREE Full text] [CrossRef] [Medline]
    60. Cultural competence in health and human services. Centers for Disease Control and Prevention (CDC). 2024. URL: https://npin.cdc.gov/pages/cultural-competence-health-and-human-services#what [accessed 2024-07-30]
    61. Cultural competence. American Physical Therapy Association. 2020. URL: https://www.apta.org/siteassets/pdfs/policies/cultural-competence.pdf [accessed 2024-07-30]
    62. Physical therapy worforce analysis: a report from the American Physical Therapy Association. American Physical Therapy Association. Dec 2020. URL: https://www.apta.org/contentassets/5997bfa5c8504df789fe4f1c01a717eb/apta-workforce-analysis-2020.pdf [accessed 2024-07-30]
    63. Yeowell G. ‘Isn’t it all whites?’ Ethnic diversity and the physiotherapy profession. Physiotherapy. Dec 2013;99(4):341-346. [CrossRef] [Medline]
    64. Remes O, Mendes JF, Templeton P. Biological, psychological, and social determinants of depression: a review of recent literature. Brain Sci. Dec 10, 2021;11(12):1633. [FREE Full text] [CrossRef] [Medline]
    65. Schoenthaler A, Ravenell J. Understanding the Patient Experience Through the Lenses of Racial/Ethnic and Gender Patient-Physician Concordance. JAMA Netw Open. Nov 02, 2020;3(11):e2025349. [FREE Full text] [CrossRef] [Medline]

    ‎
    aOR: adjusted odds ratio
    BMI: body mass index
    CBT: cognitive behavioral therapy
    DCP: digital care program
    GAD-7: Generalized Anxiety Disorder 7-item scale
    MCIC: minimum clinically important change
    MSK: musculoskeletal
    NPRS: Numerical Pain Rating Scale
    PHQ-9: Patient Health Questionnaire 9-item
    PT: physical therapist
    WPAI: Work Productivity and Activity Impairment

    Edited by G Eysenbach, A Schwartz; submitted 13.08.24; peer-reviewed by A Bove, S Rousset; comments to author 19.09.24; revised version received 25.09.24; accepted 08.10.24; published 29.10.24.

    Copyright

    ©Anabela C Areias, Dora Janela, Maria Molinos, Virgílio Bento, Carolina Moreira, Vijay Yanamadala, Steven P Cohen, Fernando Dias Correia, Fabíola Costa. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 29.10.2024.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research (ISSN 1438-8871), is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.