Low Back Pain Treatment by Athletic Trainers and Athletic Therapists: Biomedical or Biopsychosocial Orientation?

Low back pain (LBP) continues to be highly prevalent worldwide and was identified in the latest Global Burden of Disease Study¹ as the leading cause of disability globally. In the United States, LBP has been the leading noncommunicable condition contributing to years lived with disability since 1990.² In Canada, 50.9% of chronic pain sufferers identified the upper or lower back as the location of their pain.³ The huge personal, societal, and economic costs of LBP have led to decades of research attempting to uncover its underlying mechanisms to improve treatment and recovery; however, the concerning prevalence of LBP may stem from the biomedical lens that once dominated early treatment practices and remains influential today.

The biomedical/biomechanical model of disease focuses on physical injury or tissue damage as the cause of pain and disability; therefore, related treatment practices focus on identifying the injury or tissue damage and treating that as the source of pain.^4,5 In the last 20 years, the biomedical model has increasingly been questioned because the relationship between pain and injury has been exceedingly variable in laboratory and clinical studies. Inconsistencies among tissue injury, pain, and dysfunction versus the medical management of LBP have shifted researchers toward a different approach to treatment.

In contrast, biopsychosocial concepts emphasize the importance of addressing patient-oriented beliefs and attitudes, such as fear of pain and catastrophizing, which could lead to maladaptive behaviors that contribute to disability and chronic pain.^6,7 Promoted as the new theoretical framework for LBP treatment, the biopsychosocial model emphasizes understanding human behavior through interactions among the individual's biological, social, and psychological aspects.⁸ Treatment practices encourage education on pain, elements of cognitive behavioral therapy, and graded exercise programs. Statistically and clinically significant improvements in patient outcomes have been observed with these methods.⁹ Furthermore, guidelines for LBP treatment now promote an approach that includes biopsychosocial principles to aid health care providers (HCPs) in addressing these psychological responses and do not recommend biomedical/biomechanical advice and instruction.¹⁰

Health care providers' orientations toward LBP can influence rehabilitation and alter patient outcomes.^11–13 A biomedical orientation tends to focus on finding and treating pain as a consequence of physical injury, whereas a biopsychosocial orientation also addresses the psychological and social states of the patient.^14,15 The choice of treatment and advice given to the patient regarding activity and work restrictions are ways in which the treatment orientation has been documented as influencing rehabilitation.^16,17 Measuring the treatment orientation of HCPs is therefore important, and the Pain Attitudes and Beliefs Scale for Physiotherapists (PABS-PT) is a tool that was developed for this purpose.

The PABS-PT is a 2-factor scale.¹⁵ The first 10-item factor addresses a biomedical orientation (the biomedical subscale), and the second 9-item factor describes a biopsychosocial orientation (the biopsychosocial subscale).¹⁵ Using the PABS-PT, investigators¹¹ reported that patients who received treatment from biopsychosocial-oriented physiotherapists (PTs) described better disability and pain outcomes than patients who received treatment from biomedical-oriented counterparts. In another study,¹⁸ PTs and general practitioners (GPs) who scored high on the biomedical subscale of the PABS-PT were more likely to give work and activity advice that was not in line with LBP treatment guidelines. These findings suggested that the treatment orientation of an HCP, regardless of exercise prescription, influenced the use of LBP guidelines and, ultimately, rehabilitation outcomes. The existing literature has focused on PTs, GPs, and other rehabilitation therapists, but we were unable to find data on athletic trainers (ATs) in the United States and athletic therapists in Canada (C-ATs), even though ATs and C-ATs treat patients with LBP from both the general and athletic populations. Therefore, the purpose of our study was to measure the treatment orientations of ATs and C-ATs regarding LBP. In addition, we investigated how these beliefs varied based on the volume of LBP patients and the experience of the AT or C-AT.

METHODS

The National Athletic Trainers' Association (NATA) generated a random list of 2075 active members who were willing to be notified about research studies. These members were contacted via e-mail, which contained a link to our survey. The Canadian Athletic Therapists Association (CATA), with a significantly smaller membership than that of the NATA, e-mailed all members (including students) with a link to the survey. Respondents were included in this analysis if they were active members of the NATA or CATA, whereas student respondents were excluded. Concordia University's Human Ethics Committee approved the survey (certificate of ethical responsibility #300006431), and participants were asked to consent at the beginning of the survey (if they disagreed, they were disqualified).

The survey asked demographic questions about sex, age, education level, years of experience, work setting, postprofessional training, specialization, number of LBP patients per year, personal episodes of LBP, and client base; questions were open response or multiple choice. We used the PABS-PT¹⁵ to assess treatment orientation, modifying the name of the survey from “…for Physiotherapists” to “…for Athletic Trainers” or “…for Athletic Therapists” according to our sample (Table 1). The 19 items of the PABS-AT/C-AT were rated using a 6-point Likert scale ranging from totally disagree to totally agree. Scores on the PABS-AT/C-AT were calculated by summing the indicated number (1–6) for every item in each subgroup. Therefore, the biomedical subscale (10 items) had a potential score of 10 to 60, and the biopsychosocial subscale (9 items) had a potential score of 9 to 54. The reliability of the PABS-PT was previously found¹⁵ to be satisfactory for both the biomedical (Cronbach α = 0.80) and the biopsychosocial (Cronbach α = 0.68) subscales. Also, validity was determined by associations between the PABS-PT and measures of similar constructs, including the Health Care Providers' Pain and Impairment Relationship Scale and the Tampa Scale for Kinesiophobia, whereby the PABS-PT predicted a judgment of harmfulness for certain daily activities.¹⁵ The 19-item version of the PABS with modification for provider type has been used to study PTs in New Zealand, Sweden, The Netherlands, Quebec, Canada, the United Kingdom, and the United States; GPs in the United Kingdom; and chiropractors in Australia.^{11,15,17–21} The survey was hosted on SurveyMonkey.com (San Mateo, CA). Responses were matched to an Internet protocol address and were time stamped but remained anonymous. Response data were downloaded as a Microsoft Excel (version 2011 for Mac; Redmond, WA) spreadsheet.

Table 1 Pain Attitudes and Beliefs Scale for Athletic Trainers and Athletic Therapists^a

View Fullscreen

We completed the data analysis using SPSS (version 24.0; IBM Corp, Armonk, NY) statistical software; the α level was set a priori at .05 for statistical significance. Descriptive data were generated for demographic, educational, and work characteristics; mean biomedical and biopsychosocial scores were calculated for all ATs and C-ATs; and t tests and analyses of variance were used to identify significant differences between groups. When an analyses of variance identified a significant difference, a Tukey post hoc analysis was used to confirm where the significant differences occurred between groups. Spearman correlations were used to assess relationships between the biomedical and biopsychosocial scores and the AT or C-AT characteristics, including age, number of LBP patients per year, and years of experience.

RESULTS

Of the 2075 NATA-registered ATs invited to participate in our survey, 23 (1.1%) surveys were returned to the sender with error messages indicating incorrect addresses. Of the 2052 delivered e-mails, 355 members responded (response rate = 17.3%). We then excluded 82 respondents (9 were missing >1 response on each subscale of the PABS-AT, and 73 did not start or complete the survey). The total sample of ATs analyzed after exclusion (n = 273) was 13.3% of the total sample contacted.

The CATA contacted all approximately 2500 members to participate in our survey, and 484 responded (response rate = 19.4%). We excluded 102 respondents (9 were missing >1 response on each subscale of the PABS–C-AT, 76 did not complete the survey, and 17 were students). The total sample of C-ATs analyzed after exclusion (n = 382) was 15.3% of the total sample contacted.

When a respondent had only 1 missing response on each respective subscale of the PABS–AT/C-AT, we averaged the rest of the responses on the subscale missing the response and substituted the value as the missing response. Although no method has been published for dealing with missing data on the PABS-PT, our approach is an accepted solution to missing self-reported quantitative data.^18,22 During the analysis of demographic groups, we removed respondents who did not provide an answer or whose answers were not numeric (eg, “enough” in the category “low back pain patients per year”).

Most ATs were female (56%) and had a master's degree (70%; Table 2). Because of the overall older population of ATs who responded compared with C-ATs, it is important to report the median values of interest. The majority of ATs who worked primarily in secondary schools treated a median (interquartile range) of 10 (7.5–23) LBP patients per year. As well, most of the ATs who identified high school athletes as their primary client base treated a median (interquartile range) of 11.5 (5–24.5) LBP patients per year, whereas the 34% of those who mostly treated the general public treated a median (interquartile range) of 50 (12–100) LBP patients per year. The youngest AT was age 22 and the oldest was age 75; the average age was 39.8 years.

Table 2 Demographic Characteristics and PABS-AT/C-AT Results for ATs and C-ATs

View Fullscreen

Most C-ATs were female (66%), had a bachelor's degree (64%), and treated a client base of the general public (57%; means reported in Table 2). (Again because the data were skewed toward the younger population, we report the median values of interest.) The majority of C-ATs who worked primarily in private clinics treated a median (interquartile range) of 50 (25–150) LBP patients per year. Also, most of the C-ATs who identified the general public as their primary client base treated a median (interquartile range) of 50 (20–100) LBP patients per year. Interestingly, 13% (mean) of C-ATs chose other for their education, which included diplomas and postgraduate studies. One of the most frequently attended accredited athletic therapy institutions in Canada is Sheridan College. For a while, Sheridan College graduates earned a diploma for sports injury management, which explains the large number of other responses in this category. The youngest C-AT was age 21 and the oldest was age 69; the average age was 35.6 years.

Athletic Trainers

Biomedical scores differed based on the number of LBP patients treated per year (F = 4.1, P = .007). Post hoc comparisons using the Tukey honestly significant difference (HSD) test revealed that ATs who treated 9 to 15 LBP patients per year had a higher biomedical score (35.0 ± 5.7) than ATs who treated 16 to 34 (31.9 ± 5.5, P = .039) or >34 LBP patients per year (31.7 ± 8.6, P = .018; Table 2). The biomedical and biopsychosocial scores did not differ based on age, years of experience, or client base. A negative correlation was present between age (r = −0.147, P = .015) and the biomedical subscale score of the PABS-AT, and a positive correlation was observed between the average number of LBP patients treated per year and the biopsychosocial subscale score of the PABS-AT (r = 0.132, P = .032; Table 3).

Table 3 Spearman Correlations Between Age, Years of Experience, and Number of Low Back Pain Patients Treated per Year and the PABS-AT/C-AT Subscales (Biomedical and Biopsychosocial)

View Fullscreen

DISCUSSION

Previous researchers^{11,15,17–21} investigated the pain-oriented beliefs of PTs, GPs, chiropractors, and other clinicians using the 19-item PABS. We aimed to characterize the biomedical and biopsychosocial beliefs and orientations of ATs and C-ATs. To our knowledge, these are the first data to measure treatment orientations among ATs and C-ATs. In other populations of HCPs, treatment orientation has been associated with changes in LBP patient outcomes. Because ATs and C-ATs treat LBP patients in the active and general populations, it is important for us to understand their orientations toward LBP treatment.

The ATs and C-ATs who treated more LBP patients per year had lower biomedical scores. Also, C-ATs who were older and had more years of experience displayed higher biopsychosocial scores, and C-ATs who primarily treated the general public had lower biomedical scores.

Although cross-sectional studies do not allow for causal inferences, many potential explanations are possible for these findings. We speculate that after treating more people with chronic LBP, ATs and C-ATs may naturally come to realize the poor correlation between the lesion site of LBP and patient function. This concept is similar to the biopsychosocial approach to LBP rehabilitation, which is noticeably different from the standard athletic training or health profession educational approach. Athletic training education, as well as education for almost all other HCPs, is based on the biomedical approach, which includes a systematic assessment process (from history to special tests) for identifying the lesion site of an injury and then treating it accordingly. Although this strategy works for most injuries, it is becoming accepted that the biomedical approach does not adequately explain the clinical nuances of chronic LBP, which ATs and C-ATs might be learning with experience and exposure over time. Another possible explanation may be postgraduate training. We found no relationship between education or specialized training and pain-oriented beliefs, but previous authors¹⁷ reported lower biomedical scores and higher biopsychosocial scores among physiotherapists with postprofessional training in chronic pain management. Cultural aspects have been suggested as being influential as well.²³ At a minimum, we may need to increase exposure to biopsychosocial evidence-based studies to guide appropriate treatment orientations in future ATs and C-ATs.

Our sample of ATs responding to this survey described primarily treating high school and collegiate or university athletes. These practice settings may have predisposed them to see more patients with acute versus chronic injuries^24,25 and consequently believe that biopsychosocial principles were less effective with their practice population, a belief that is not supported by the literature.^26,27 Maladaptive pain beliefs, such as risk avoidance, delay recovery during the acute and subacute phases as well. Therefore, ATs should be able to appropriately incorporate biopsychosocial strategies with acutely injured patients to address psychological distress before chronic pain develops²⁸—for example, earlier detection of when high levels of fear and avoidance are delaying activity progression after an acute injury and being able to adjust the treatment approach to directly alleviate the fear.

Studies from Canada,¹⁷ New Zealand,²⁰ and the United Kingdom¹⁸ indicate discouraging associations between treatment orientation and treatment behavior. Similar mean biomedical and biopsychosocial scores were associated with other HCPs' recommendations against (including delaying return to work and activity)^17,18 and disuse of the treatment guidelines (Table 4).²⁰ The reasons for misuse or disuse of treatment guidelines are concerning in that such behavior can have a negative effect on patient outcomes. For example, researchers¹¹ measuring the effect of biopsychosocial training on PTs described positive associations between treatment orientation and patient disability and pain outcomes. Specifically, posttraining biopsychosocial scores >40 were associated with less disability and pain in patients receiving treatment for LBP.¹¹

Table 4 Summary of Existing International Studies Using 19 Items as Compared With ATs' and C-ATs' Pain Attitudes and Beliefs Scale Results

View Fullscreen

FUTURE DIRECTIONS

This area of research would benefit from further measurement of factors influencing ATs' and C-ATs' treatment orientations and of the influence of orientation on patient outcomes. One potential step is to pursue reducing PABS–AT/C-AT biomedical scores in athletic training and athletic therapy populations, which has not yet been undertaken. Both the NATA and CATA have required certain educational competencies to be taught and evaluated in their training programs, including psychosocial principles, but neither organization provides specific and standardized guidelines regarding the actual instruction and dissemination of the material.^29,30 This is important because studies^29,31,32 published around the time of release of the current NATA educational competencies (2011) showed that ATs did not implement the most effective and appropriate psychosocial techniques, expressed low satisfaction with taught psychosocial interventions and referral content, and wished to learn more about psychosocial strategies. Future educational interventions, therefore, may be delivered to practicing ATs who lack crucial components of psychosocial education through continuing education courses, workshops, or conferences, but it may be more effective to deliver multiple biopsychosocial courses throughout a typical athletic training education program.²⁹ This is an important distinction because, although postgraduate training has been demonstrated as effective in influencing treatment orientation,^11,33 for a truly psychologically informed practice, therapists need to be able to incorporate elements of cognitive behavioral therapy to enhance their usual treatment practices and patient management.³⁴ These skills require practice and feedback from a mentor, an educational approach that is not typically offered in traditional continuing education courses.³⁵ Consequently, future researchers should aim to critically examine the acquisition of a biopsychosocial treatment orientation through educational bodies and how to enhance this orientation in practicing ATs and C-ATs.

LIMITATIONS

This study had several limitations. The open-ended design of some questions forced us to eliminate many respondents' answers: for example, those who chose to report a percentage instead of an approximate number of LBP patients seen per year. In a similar vein, we excluded a large number of respondents who did not complete the survey. One reason for noncompletion may have been that respondents found the survey too extensive, even though we informed them of the estimated 10-minute completion time in the first e-mail. The generalizability of our results may be hindered by the lower response rates from our populations. In addition, only a sample of the NATA population was contacted (a little more than 2000 ATs), whereas the entire CATA population was contacted (approximately 2500 people). Some of the e-mailed invitations sent to ATs may have been construed as spam because they were from the created e-mail “universitylowbackpain@gmail.com” rather than 1 of our official university-affiliated e-mails. If we had been able to send an e-mail from the NATA directly, we might have had a better response rate. Notably, our CATA e-mails were sent directly from a CATA representative, and the response rate was better. Our decision to include only 1 submission (the first) per Internet protocol address (to ensure that respondents completed the survey only once) may have also limited our total response rate (eg, in work environments where employees share computers). Many previous investigators included patient vignettes¹⁷ or described patient treatments¹¹ to measure changes in treatment behavior and advice. We did not include these measures, so we cannot report on correlations between the observed treatment orientation and actual treatment behavior or advice. We selected age comparisons a priori that would come close to balancing the samples; however, this created different age categories between ATs and C-ATs that are difficult to compare. We also did not gather information on patient load, nor did we separate patients according to initial evaluations or follow-ups.

CONCLUSIONS

Athletic trainers and C-ATs may apply a more biomedical-oriented approach to LBP treatment, and such an orientation may represent a disconnect between what was learned during training and what is required in practice. Specifically, our survey findings suggest that gaps remain regarding the implementation of biopsychosocial training in the current educational standards. These results do not indicate that ATs and C-ATs should leave their biomedical orientation behind completely, but instead, they should adopt a psychologically informed approach to better address patients with LBP.