A practical tool to improve outcomes in Work Injury Management

Abstract

BACKGROUND:

Psychosocial factors have the greatest influence on work rehabilitation outcomes yet effective and efficient systems to manage these factors are not widely utilized in Work Injury Management.

OBJECTIVE:

To report on the development and validation of a comprehensive assessment tool with practical utility in identifying and measuring the biopsychosocial factors which are barriers to return to work and community.

METHODS:

Literature search identified many instruments designed to identify the presence of psychosocial factors impacting recovery and return to work following musculoskeletal injury. This research aimed to match assessment with intervention. In clinical practice over a 5 year period, this led to development of a composite questionnaire, the Abilita Rehabilitation Index, which was trialed with 43 adults participating in Occupational Rehabilitation. The validation data are based on the results of 957 case records of Occupational Rehabilitation participants.

RESULTS:

Examination of Cronbach coefficient alpha of the instrument indicates strong internal consistency (0.90) and factor analysis supports satisfactory construct validity of the domains (subscales) with factor loading scores ranging from 0.73 to 0.90.

CONCLUSION:

There is evidence that the proprietary Abilita Rehabilitation Index is a valid and reliable instrument to identify and measure the influence of psychosocial factors impacting an individual’s recovery and return to work. Automated reports generated from this tool provide an evidence-based resource to identify risk and support rehabilitation planning in Work Injury Management.

Keywords

Return to work musculoskeletal disorders psychosocial factors pain biopsychosocial

1 Introduction

Effective Work Injury Management requires the use of appropriate activities and procedures that ensure a timely, safe and durable return to suitable work for injured workers. It includes psychosocial approaches in addition to traditional bio mechanical approaches [1]. The development of the Flags Model [2] provided the first widely accepted biopsychosocial assessment framework for musculoskeletal injuries. It introduced the wide range of biological, psychological and social (work, environmental and cultural) factors that are both predictive of risk of work absence and influential in development of prolonged pain behaviour and disability. In practice this model resulted in classification of injured workers as ‘low risk’ or ‘high risk’ but did not provide specific information to aid intervention planning.

The biopsychosocial model has been applied with varying protocols over many years and this may have weakened its acceptability and perceived value. In 2012, an international forum considered its success over the previous 25 years, and concluded that the biopsychosocial model is sound however failures have occurred in how it has been understood and applied [3]. Multiple physical, psychological and work interventions have been developed and researched but rarely have all three components been integrated into one intervention. In 2005 Loisel et al. [4] described the multiple legal, administrative, social, political and cultural obstacles that contribute to the complexity of disability-related work absence. These researchers reasoned that a holistic system-wide reform is necessary to achieve the paradigm shift of adding biopsychosocial to an essentially biomedical approach; to include the fundamental evidence-based practices of early diagnostic triage, identification of potential psychosocial and workplace barriers to return to work, good self-management coping, and early return to safe work [4].

In 2006, the South Australian Government commissioned a study of Facilitators and Barriers to Return to Work [5] in order to inform ongoing research requirements. This study also concluded that work disability and return-to-work are multi-determined outcomes and that they cannot be accurately predicted just from knowledge of the medical or physical condition. This review recommended that to achieve optimal return-to-work outcomes, reform is required within systems and via reduced use of unreliable or invalid interventions.

In 2009 in response to this accumulation of evidence, the biopsychosocial approach was promoted by the Australian and New Zealand Heads of Workers Compensation Authorities and Heads of Compulsory Third Party Insurance Schemes [6]. Biopsychosocial approach guidelines were prepared, advising that negative psychosocial factors are associated with poor prognosis, and recommending the use of tools to screen for “yellow flags” 1 [7]. Guides are published in most jurisdictions in Australia and New Zealand, for example NZ Acute Low Back Pain Guidelines [8] WorkCover NSW Soft Tissue Injury [9], Clinical Framework for the Delivery of Health Services (Vic) [10].

The lack of improvement in health and work outcomes is visible in SafeWork Australia’s [11] most recent key information report in which Workers Compensation claim costs (2011/12) across Australia equaled $7.8 billion. Their Return to Work Survey [12] reports 2013/14 Current Return to Work Proportion as averaging 77% across Australian jurisdictions and New Zealand, and the 3-month stable 2 RTW proportion was 61% in both Australia and New Zealand. There are indications of some jurisdictions trending improvement on previous years, however the overall results and lack of durability in return to work, highlight the need for further consideration of current practice.

The model of rehabilitation practice applied in Australia is dependent on scheme policy and the interpretation of that policy by insurers, employers and rehabilitation service providers [13, 14]. Rehabilitation may be managed by employer injury management personnel or by referral to external rehabilitation and health providers. The Nationally Consistent Approval Framework for Workplace Rehabilitation Providers (WRP) (Australian and New Zealand) [14] stipulates that in their role, Rehabilitation Provider staff will identify and address the psychosocial barriers, risks and strengths influencing return to work, and their decisions are to be evidence-based with assessments that demonstrate the need for recommended services. Modifiable worker factors with consistent evidence across two or more health conditions are; emotional distress, negative enduring psychology/personality factors, negative health and disability perception, decreased physical activity, lack of family support, poor general health, increased functional disability, increased reported pain and fatigue, and lack of motivation to return to work [15]. Self-report questionnaires are considered the gold standard for the proper measurement of perceived pain [16] and influential beliefs, attitudes and expectations. Dunstan et al. [17] confirmed that injured worker’s beliefs and attitudes, assessed using a Theory of Planned Behaviour model, accurately identified influences on future work expectation and outcomes (79.2% sensitivity, 86.4% specificity).

The early identification of the presence and extent of influence of psychosocial factors is essential for their effective management. Casey [18] reviewed over 9,000 de-identified WRP case records specifically referred for return to work assistance and concluded that duration of disability is a significant determinant in positive RTW outcomes. Regardless, in this cohort there was a mean of 22 months (SD = 35) time from injury to referral.

Internal locus of control (ILC) is significantly linked to perceived bodily pain, and interventions in vocational rehabilitation designed to provide participants with a ‘toolbox’ to help them take responsibility for their own well-being and health, result in increased ILC [19]. There is robust evidence to support calls for a paradigm shift in workers compensation systems to include the combination of early intervention protocols with the provision of coping skills training and information toclaimants [20].

Given the evidence supporting this approach, it is necessary to consider the reasons that Work Injury Management systems have not achieved consistent, effective biopsychosocial assessment and intervention. The barriers to the implementation of a best-practice model are multiple [4] and despite numerous evidence-based guidelines there is little indication of adherence to these guidelines. Primary challenges include the commercial reality of time and resources required to administer questionnaires [35], lack of incentive to implement change due to financial limitation of service requests [13], and imprecise intervention recommendations [4]. Importantly, Waddell, Burton and Main emphasized that screening is not an end in itself and is only of value if it is linked to the delivery of effective rehabilitation and work-focusedinterventions [22].

The lack of specific work-injury-management training for Rehabilitation Consultants is a compounding factor. Appropriate clinical training is a necessary requirement for early graduate occupational therapists and physiotherapists to move into the field of work injury management and prevention [23]. Sullivan et al. [24] reported that practice guidelines are rarely employed by health professionals, concluding; “The development of techniques to maximize adoption and adherence to new treatment protocols will be as important as the development of new intervention approaches themselves. Unless this final objective can be achieved, patients with musculoskeletal conditions will not benefit fully from new knowledge gained through clinical science.”(p. 486).

The expectation of different stakeholders will influence any injury management approach [25] as will the effectiveness of their co-operation and communication. Employer representatives and external rehabilitation providers often hold differences in work-related beliefs [26]. Potentially, the application of structured biopsychosocial protocols delivering standardized assessment reports and coaching interventions would mitigate the impact of differences in stakeholder rehabilitation expectations.

The evidence outlined in this introduction highlights the need for better quality resources to support Work Injury Management personnel to achieve early identification of injured workers at risk of prolonged disability, and to ensure those workers receive information and coping strategies appropriate to their individual requirements. These resources will need to include protocol and training strategies [24]. The Abilita Rehabilitation Index (ARI) was developed in response to this need, and is supported by training resources and a coaching course with which Rehabilitation Consultants assist their clients to learn strategies to overcome personal and social barriers to recovery and return to work. The purpose of the current study is to examine the psychometric properties of the ARI and its capacity to support the requirements of this new model for Work Injury Management.

2 Materials and methods

2.1 Abilita Rehabilitation Index (ARI)

The ARI was developed as a single, comprehensive assessment tool designed to meet specific criteria. That is, a measure that was:

Appropriate for a population whose primary characteristics are pain behaviour and work disability as a consequence of musculoskeletal injury or illness.

Based on reliable evidence that the included factors either predict or influence the onset of persistent pain, disability or delayed return to work.

Constructed from items or whole scales with established psychometric strength.

To include domains that represent constructs which are evidence-based, meaningful to the respondent population and inform the preparation of a self-help coaching plan with the goal of durable return to work.

Supported by software to deliver automated reports with calculated scores and risk ratings and the collation of items into domains.

Repeatable, and with subsequent reports comparing scores and ratings with those from the initial report.

The developers of the ARI drew on the Initiatives on Methods, Measurement and Pain Assessment in Clinical Trials (IMMPACT) [27, 28] who publish consensus recommendations on outcomes assessment in clinical trials for chronic pain. It was of particular importance to define and construct domains that fitted with psychosocial-factor evidence for the target population and that primarily contain factors amenable to change. The core outcome domains recommended by IMMPACT are: (1) ratings of pain intensity, (2), physical functioning, (3) emotional functioning, (4) participant ratings of improvement and satisfaction with treatment, (5) symptoms and adverse events, (6) participant disposition. The first three assessment domains were selected to be included in the ARI. The fourth domain is included in the ARI outcome measure, and the remaining two domains are addressed during user training.

Six attitudinal constructs were judged by the developers to be relevant to the ARI; these are Pain, Function, Emotions, Coping, Confidence and Work Perceptions. A seventh domain, Occupational Factors, was created to include a variety of work-related risk factors that are primarily not amenable to change. As both modifiable and non-modifiable factors are predictive of risk of prolonged pain and of work disability, both components are included in the ARI, to gain an accurate measure of risk. However as the primary purpose of this questionnaire is to inform self-management skills coaching, the majority of questions and domain content comprise factors that are amenable to change. These influential psychosocial factors are surveyed by the questionnaire, and represented in the ARI domains as summarized in Table 1.

Evidence supporting these constructs was identified through comprehensive research, including studies cited in the following domain evidence summaries.

Pain

A pain rating is not an objective measure, it is simply a measure of the extent to which a particular person reports his or her pain, and it is open to subjective influences including the level of distress [29]. A higher level of reported pain is associated with catastrophizing and higher catastrophizing 3 is positively associated with fear-avoidance beliefs [31]. Pain-related fear is more disabling than pain itself [32] and fear of pain is the strongest predictor of activity avoidance and disability [33]. A person’s understanding of pain strongly influences their coping capacity and recovery [34].

Function

Initial levels of pain ratings and perceived functional limitations are well known predictors of work disability [24]. Fear of movement has been shown to be a predictor of disability levels [35]. There is strong evidence that continuing usual activities as normally as possible (including work activities) despite pain is associated with better outcomes than is traditional medical treatment and rest [36].

Emotions

Depression increases the risk of prolonged work disability associated with musculoskeletal conditions [37]. Psychological constructs are established predictors of poor rehabilitation outcome, and the combining of important psychological elements is more useful than isolating them, in both identifying high-risk respondents and in informing tailored intervention [38].

Coping

Passive coping has been found to be a strong, independent risk factor for disabling neck and/or back pain [39]. Helplessness has been significantly associated with poorer outcomes in Whiplash Associated Disorders [40]. After early screening for disability risk factors, intervention strategies can be improved by focusing on coping with job factors, pain coping strategies, and expectations for recovery [41].

Confidence

Self-efficacy is an important determinant of clinical [42] and occupational outcome of back pain [43, 44]. Improved self-efficacy predicts successful RTW at 6 months post carpal tunnel surgery [45].

Work Perceptions

The strongest workplace predictors of low back pain work disability appear to be worker perceptions, including of physical work demands, organisational support, and fears of re-injury [46]. Workplace conditions and individual perceptions about work could be as important as anxiety, depression, pain intensity or functional disability in predicting and preventing work disability [47].

Occupational factors

This domain does not represent a single attitudinal construct; it comprises multiple worker and workplace characteristics that are not necessarily related, are primarily not amenable to change and are known to increase the risk of prolonged work disability, including type of work, claim history, age and education of worker. [5, 48]

2.2 Instrument development

A sequential, stepped approach [49] was followed to develop the assessment instrument and pilot test it with an appropriate population sample before offering it to Occupational Rehabilitation Providers. This study has utilized the pre-existing set of de-identified ARI records which have been entered and securely stored as electronic data.

From 2001 to 2008, the first author of this study coordinated research within a Rehabilitation Provider practice located in the Northern Territory of Australia, to develop a structured, evidence-based and work-focused pain self-management coaching course suitable for delivery within Occupational Rehabilitation. That research included the utilization of validated self-report questionnaires for the purpose of both initial assessment, and measurement of change following the coaching course, in reported pain, coping, function and self-management approach.

As a result of literature review and practice-based research, it was concluded that the assessment objectives could be achieved using a single comprehensive questionnaire which incorporated the Örebro Musculoskeletal Pain Screening Questionnaire (ÖMPSQ) [50] and the Pain Self-Efficacy Questionnaire (PSEQ) [44] in their entirety. In addition, twenty six items were selected from other valid and reliable questionnaires in order to meet the twin objectives of providing an initial measure of risk of long term pain and disability, and provide information from which to chart an appropriate intervention. The goal in selection of questions to inform the Abilita domains was not to achieve a compact and highly homogenous set but more to ensure that the domains were content valid to meet the construct purpose. There is evidence to suggest that the pursuit of high levels of intra-correlation within the domains may be evidence of question redundancy rather thanutility [51].

The questionnaire, the Abilita Rehabilitation Index is presented in four sections. Section 1: About You; contains 16 items including demographic, education, lifestyle and employment questions, and two “readiness to work” questions derived from Motivational Interviewing research [52]. Most Section 1 items are in Verbal Rating Scale format with responses assigned a numerical value. Section 2: Managing Your Injury is comprised of 14 Likert scale items which have been used in various forms in different questionnaires. Section 3: Örebro Musculoskeletal Pain Screening Questionnaire and Section 4: Pain Self Efficacy Questionnaire. Both ÖMPSQ and PSEQ are included in their entirety with the kind permission of their authors.

The ARI incorporates items originally scored on varying scales and with differing scoring orientation, for example ÖMPSQ rates question responses on an 11 point Likert scale, on which for some questions, a high score represents a favorable condition and with others the high score is unfavourable. Responses to the PSEQ and Section 2 questions are rated on a 7 point Likert scale. With all PSEQ questions a high score is favorable. Higher aggregate ARI score indicates higher risk of persistent pain and disability.

A notable feature of the ARI is that software delivers the questionnaire online, stores all inputs in a secure database, and transforms the raw data to calculate the individual domain ratings, the ARI aggregate score and total scores for the ÖMPSQ and PSEQ. These outputs and relevant case data are incorporated in the ARI Initial Report which is immediately available online to the consultant administering the assessment. A second report delivered concurrently, the ARI Profile Report, includes detailed characteristics within each domain for the specific purpose of aiding the Rehabilitation Consultant to prepare a plan for self-management coaching tailored to individual requirements.

At completion of the coaching component of the injured worker’s rehabilitation program, the respondent completes the questionnaire online once again, to measure change. This results in generation of the ARI Impact Report providing comparison with initial scores. An additional four questions gain the participant’s rating of benefit and satisfaction with the intervention; recommended as the fourth of the IMMPACT core domains [49]. The ARI questionnaire may be completed twice more resulting in the generation of the Review1 and Review2 Reports.

2.3 Participants

During 2006 to 2008, the ARI questionnaire was trialled in a pilot study within an Occupational Rehabilitation practice. Respondents (N = 43) had sustained a musculoskeletal injury and were receiving rehabilitation and compensation under workers compensation, disability, motor accident or military compensation schemes. The mean duration for the 43 cases from date of injury to date of initial assessment was 271 weeks, with median of 185 weeks and ranged from 17 to 1565 weeks. Respondents ages ranged from 20 to 59 years, with a mean of 41.3 (SD:10.48). Females comprised 44% of the trial participants. At the time of initial assessment, 49% were not at work, 28% were working part time and 23% were working full time.

Having established that the tool had satisfactory face validity properties for use within regular rehabilitation service provision it was made available to willing parties in other jurisdictions in Australia and New Zealand from 2009.

Data utilized in this study are from the Initial questionnaire database derived from 957 respondents, accumulated between January 2009 and October 2014. This population was comprised of working age people with musculoskeletal injury receiving rehabilitation and compensation under worker’s compensation, disability, motor accident or military compensation schemes. Mean duration of claim from date of injury to date of initial assessment was 109 weeks, with median of 44 weeks and range of 2 to 1181 weeks. Respondent ages ranged from 18 to 71 years, with mean age of 43.9 (SD = 11.12), 58% were female. At the time of initial assessment, 47% were not at work, 28% were working part time and 25% were working full time.

Records indicate that participants had experienced a variety of musculoskeletal injuries. Respondents reported their experience of pain according to the ÖMPSQ question: Where do you have pain? 36% reported multiple pain sites, 10% lower back/lower limb, 17% other two sites combined, and those with a single pain site reported; upper limb 10%, lower back 11%, lower limb 10%.

2.4 Procedure

Specific training was mandatory for Rehabilitation Consultants in order for them to access the Abilita Assessment website to ensure consistency in introduction of the online self-report questionnaire to participants. Consultant adherence to this process was facilitated by the web-based assessment questionnaire and automated, standardized reports, minimizing time required and negating both observer error and bias. Participant error was minimized by the consistent web-based presentation of the questionnaire and by introduction to the assessment by a trained Consultant. Respondent bias was minimized by the varied question orientation, and the scattered location throughout the instrument of the multiple questions informing each domain. Respondents knew that their actual responses remained confidential and this provided additional protection against participant misinformation.

2.5 Acceptability

Acceptability was assessed through a survey of Rehabilitation Consultants who had administered the instrument tool to evaluate if it was viewed, by themselves and respondents, as useful to their purposes. This survey was conducted using online survey facility “Survey Monkey” responses to 7 statements were received from 35 active Abilita Consultants. The statements were: Most clients found the questions relevant to them; Most clients found the online format acceptable; Most clients found the domain chart provided an accurate representation of the impact of their injury; I find the assessment format (online and automatic reports) easy to administer, in return for the amount of information collected; I have found the assessment process (questionnaire and discussion of Initial report) helps to engender clients’ trust and engagement in my rehabilitation proposal; I have found the Abilita Reports useful in planning rehabilitation; and I have found the Abilita reports valuable in communicating with treatment providers, employers and insurers in relation psychosocial influences. Responses to the survey were rank ordered in the form: Disagree, Mostly disagree, Mostly agree, Agree.

2.6 Data analysis

Reliability and validity analysis have been conducted using IBM SPSS Statistics V21 software. Reliability was assessed by examination of internal consistency (Cronbach’s alpha coefficient). Concurrent validity was assessed by examining correlations between the ARI and the ÖMPSQ and PSEQ. These are two established measures of attitudes that influence pain, coping and work disability [44, 50]. Where the sample was of sufficient size (Validation data), construct validity was assessed through analysis of the ARI’s factor structure (factor analysis). This study was approved by the Faculty Human Ethic Committee at La Trobe University.

3 Results

The primary outputs for the Abilita Rehabilitation Index (ARI) are the ARI scores, the 7 domain ratings and the aggregate scores for the ÖMPSQ and PSEQ. Rating for the domain Occupational Factors was not included in the data analysis because this domain is designed to record a set of categorical variables known to increase the risk of persistent work disability [5] that are not necessarily related. The raw responses from the other six domains have been transformed to standardize orientation and weight. These scores and ratings are the variables included in data analysis for this study.

The responses to questions that inform the domains are on a Likert scale and therefore constitute ordinal data. The raw score of each of the questions in the domains is transformed by the software to have equal weight and the same orientation. The domain scores are sums of the questions which have been transformed and factored to a rating from 0 to 4, and can therefore be treated as continuousvariables.

3.1 Trial data

The measure of internal consistency, Cronbach Alpha coefficient, was calculated as 0.86 for the overall questionnaire. Cronbach Alpha coefficient calculations for the individual domains are shown in Table 2.

In this small sample (N = 43), the domain Cronbach’s alpha scores ranged from 0.61 to 0.82.

Results of correlations with both ÖMPSQ and PSEQ were high; ÖMPSQ (r = 0.88, p = 0.01) and PSEQ (r = –0.88, p = 0.01). This result provided evidence of concurrent validity of the ARI.

3.2 Validation data

The descriptive statistics for the 957 responses are in Table 3. These continuous variables are described in terms of mean, median, range and standard deviation.

Factor loading within each of the six domains was established using principal component analysis with orthogonal rotation and subject to Eigenvalue >1. The purpose of this analysis was not to seek reduction in the number of items informing each domain as all items were selected to provide particular information to aid rehabilitation intervention. The aim of this analysis was to establish that each question loads onto a principal component for that domain. A mean factor score is presented in Table 4 in lieu of publication of the identity and number of questions informing each domain as these constitute part of the intellectual property claimed in this instrument. That information can be made available for genuine research purposes; interested researchers should contact the first author directly.

The ARI score was assessed for Construct Validity via Factor Analysis and Internal Consistency (Cronbach’s alpha). The Cronbach score for the total ARI was 0.90. Individual domain factor scores are set out in Table 5.

The factor scores were obtained by principal component analysis with orthogonal rotation, and subject to Eigenvalue >1, accounting for 67.5% of the total variance. The purpose of analysis was to confirm that the domain variable all load onto the principal component. Item/adjusted total correlations (which excluded the relevant domain from the total for each correlation) are also represented in Table 5. These correlations varied from 0.63 to 0.84 with two domain correlations below 0.7, although still moderately high at 0.63 (Pain) and 0.66 (Emotions).

As expected, the results of correlations with ÖMPSQ and PSEQ were high; ÖMPSQ (r = 0.93, p = 0.01) and PSEQ (r = –0.88, p = 0.01). This confirmation of concurrent validity also provides initial predictive validity of the ARI as these two instruments have strong capacity to predict pain-related disability [44, 50].

3.3 Acceptability survey

Responses to the Acceptability Survey were rank ordered in the form: Disagree, Mostly disagree, Mostly agree, Agree. Results indicate a high level of acceptability to both participants and users with 99% of respondents selecting ‘Agree’ or ‘Mostly Agree’ to all seven acceptability attributes of the questionnaire.

4 Discussion

Accurate assessment of psychosocial factors associated with pain and disability can assist in the design of individually-tailored rehabilitation services [53]. The ARI was developed to provide both assessment of psychosocial risk indicative of persistent pain and disability, and sufficient reporting to inform an Occupational Rehabilitation intervention to reduce that risk. The current research aimed to assess the reliability and validity claims of the ARI. Results from this study indicate that this instrument does have acceptable psychometric characteristics and therefore the capacity to support its purpose.

The contribution of psychosocial factors in the persistence of pain and disability, and the superior results attainable from evidence-based biopsychosocial approach to rehabilitation, are well documented [24]. This paper brings much of that evidence to focus on a measurement approach that serves the dual purpose of establishing a baseline risk rating for each participant and for providing a chart for the design of individually-tailored rehabilitation intervention. The key to implementation of this new Occupational Rehabilitation model is the development of a valid set of domains that add dimension to a biopsychosocial assessment risk rating. An important strength of the ARI domains definition and content is that they were refined with selection, trial and review of items, in Occupational Rehabilitation practice.

Results from the current study, using data collected over a 5 year period, have provided psychometric support for the scale. The ARI was developed through selection of appropriate questions to inform the domains and statistical testing followed a correlational design. Once sufficient data were available, validity was assessed through analysis of the ARI factor structure and by examination of the ARI’s relationships with validated measures of attitudes that influence pain, coping and work disability. The ARI’s concurrent validity was reflected in high correlations (in expected directions) with both ÖMPSQ and PSEQ.

The questions that inform each domain are varied in number with each cluster of questions demonstrating satisfactory, above 0.7, mean factor scores for all other than the question clusters for Function (0.69) and Coping (0.59). The Coping score may be a result of the breadth of contributing questions, which are included to capture the wide range of factors influencing this domain. Internal consistency of each domain was demonstrated through satisfactory (above 0.7) Cronbach’s alpha results for all domains other than Work Perceptions (0.64). The coefficient range for domains was from 0.64 to 0.87. The domain Work Perceptions was retained because it provided essential information for intervention design, and because the overall Cronbach’s Alpha result for the entire tool, including that domain, was strong. These domains inform the ARI, and all domains had strong factor loading, ranging from 0.73 to 0.90. These results indicate satisfactory construct validity both for the ARI and within each domain. The sample size of 957 adds strength to the stability and results fromthis study.

The initial Abilita assessment delivers the Rehabilitation Consultant immediate scored reports of practical value to communications with client, customer and treating practitioners. The ARI Impact Report, generated from repeat assessment after coaching, enables a comparison with the initial scores and domain ratings. After an appropriate intervention [19], it is no longer the baseline psychosocial factors that predict RTW outcome but rather the reduction in these factors [21]. The ARI Impact Report records both the change in influential factors and change in work status. A separate, comprehensive study by the current researchers is focusing on the results from Abilita Coaching, including the relationship between change in ARI score and work outcomes. This research will provide further opportunity to evaluate the predictive validity of the ARI.

5 Conclusion

The composite assessment instrument utilized in this study, the Abilita Rehabilitation Index, has high utility within the closely-regulated sphere of Occupational Rehabilitation. It provides guidance for the development of best-practice biopsychosocial rehabilitation intervention with the potential to contribute to a reduction in persistent pain, distress and work disability as a consequence of musculoskeletal injury. This model, set in the context of return to work rehabilitation, is ideally placed to deliver earlier work readiness and more durable return to work outcomes.

Further research is now required to more extensively test the utility of this approach and its capacity to consistently facilitate improved and sustained return to work post injury.

Conflict of interest

This article reports part of the Master’s thesis in Applied Science of P. Garton. P. Garton is the Managing Director of Abilita Services Pty Ltd which holds the intellectual property rights of the Abilita Rehabilitation Index. Professor G. Murphy and Dr P. O’Halloran declare they have no conflict ofinterest.

Footnotes

Yellow flags are personal psychosocial factors associated with unfavourable clinical outcomes and the transition to persistent pain and disability.

“3-month stable” is defined as the proportion of injured workers who were working, either part-time or full-time, at the time of the survey and had been back at work for at least 13 weeks on a regular basis [].

Catastrophizing is defined as “a particular response to pain symptoms that includes elements of rumination, magnification and helplessness” (p.9) and predicts pain intensity and medication dependency [].

Acknowledgments

The authors would like to acknowledge the professionalism of the Rehabilitation Providers and Injury Management personnel in both public and private employers who have implemented the Abilita Program for the benefit of their clients, and thereby contributed to this research dataset. Thanks also to Ben Creswick, Software Engineer who has contributed meticulous attention in development of software supporting this instrument; to Louise Bilato, Rehabilitation Counsellor who participated in early concept and program development; and Peter Garton for his support, contribution to development of the Program, and data analysis.

References

James

. Work Injury Management. Work 2014;48(4):555–556.

Kendall

, Linton

, Main

. Guide to assessing psychosocial yellow flags in acute low back pain: Risk factors for long-term disability and work loss. Wellington, NZ: ACC 1997;31.

Pincus

, Kent

, Bronfort

, Loisel

, Pransky

, Hartvigsen

. Twenty-five years with the biopsychosocial model of low back pain-is it time to celebrate?. A report from the twelfth international forum for primary care research on low back pain. Spine 2013;38(24):2118–2123.

Loisel

, Buchbinder

, Hazard

, Keller

, Scheel

, van Tulder

, et al. Prevention of work disability due to musculoskeletal disorders: The challenge of implementing evidence. J Occup Rehabil 2005;15(4):507–524.

Foreman

, Murphy

, Swerissen

. Facilitators and barriers to return to work: A Literature Review for South Australia Workcover. Melbourne: Australian Institute for Primary Care, La Trobe University 2006.

Heads of Workers Compensation Authorities (AU). Biopsychosocial Injury Management HWCA [Internet], Melbourne, AU. 2011. Available from http://www.hwca.org.au/projects.php.

Koes

, van Tulder

, Lin

, Macedo

, McAuley

, Maher

. An updated overview of clinical guidelines for the management of non-specific low back pain in primary care. Eur Spine J 2010;19(12):2075–2094.

Accident Compensation Corporation (NZ). New Zealand Acute LowBack PainGuidelinesACC[Internet],NZ. 2004. Available from http://www.acc.co.nz/about-acc/research/WPC096922.

WorkCover NSW (AU). Improving outcomes: Integrated, active management of workers with soft tissue injury. Sydney: WorkCover NSW, 2008.

10.

Health Services Group. Clinical Framework For the Delivery of Health Services.WorkSafe Victoria,Melbourne, AU. 2012. Available from https://www.worksafe.vic.gov.au/forms-and-publications/forms-and-publications/clinical-framework-for-the-delivery-of-health-services.

11.

SafeWork Australia. Key Workers’ Compensation Information Australia. SafeWork Australia, Canberra, AU. 2015. Available from http://www.safeworkaustralia.gov.au/sites/swa/about/publications/pages/key-wc-information-2014.

12.

Social Research Centre. Return to Work Survey 2013/14 Summary Report (Australia and New Zealand). Safe Work Australia, Canberra, AU. 2014. Available from http://www.safeworkaustralia.gov.au/sites/swa/workers-compensation/rtw/pages/rtw.

13.

Wales

, Matthews

, Donelly

. Medically unexplained chronic pain in Australia: Difficulties for rehabilitation providers and workers in pain. Work 2010;36(2):167–179.

14.

Heads of Workers Compensation Authorities(AU). Guide Nationally Consistent Approval Framework for Workplace Rehabilitation Providers. HWCA, Melbourne, AU. 2012. Available from http://www.hwca.org.au/projects.php.

15.

Wagner

, White

, Schultz

, Murray

, Bradley

, Hsu

, et al. Modifiable worker risk factors contributing to workplace absence: A stakeholder-centred best-evidence synthesis of systematic reviews. Work 2014;49(4):541–558.

16.

Strong

, Unruh

, Wright

, Baxter

. Pain: A textbook for therapists. Edinburgh,UK: Churchill Livingstone, 2002.

17.

Dunstan

, Covic

, Tyson

. What leads to the expectation to return to work? Insights from a Theory of Planned Behavior (TPB) model of future work outcomes. Work 2013;46(1):25–37.

18.

Casey

, Guy

, Cameron

. Determining return to work in a compensation setting: A review of New South Wales workplace rehabilitation service provider referrals over 5 years Work 2014;48(1):11–20.

19.

Selander

, Marnetoft

, Åsell

, Selander

. Internal locus of control and vocational rehabilitation. Work 2008;30, 149–155.

20.

Aurbach

. Breaking the web of needless disability. Work 2014;48(4):591–607.

21.

Wideman

, Adams

, Sullivan

. A prospective sequential analysis of the fear-avoidance model of pain. Pain 2009;145, 45–51.

22.

Waddell

, Burton

, Main

. Identifying People at Risk of Long-term Incapacity for Work. London, UK: Royal Society of Medicine Press Ltd, 2003.

23.

Adam

, Strong

, Chipchase

. Readiness for work injury management and prevention: Important attributes for early graduate occupational therapists and physiotherapists. Work 2014;48(4):567–578.

24.

Sullivan

, Feuerstein

, Gatchel

, Linton

, Pransky

. Integrating psychosocial and behavioral interventions to achieve optimal rehabilitation outcomes. J Occup Rehabil 2005;15(4):475–489.

25.

Franche

, Baril

, Shaw

, Nicholas

, Loisel

. Workplace-based return-to-work interventions: Optimizing the role of stakeholders in implementation and research. J Occup Rehabil 2005;15(4):525–542.

26.

Murphy

, Foreman

, Young

. Differences in the organizational behaviour beliefs held by Australian employer representatives and health professionals involved in occupational rehabilitation: Implications for workplace disability management. Intern J Hum Resour Manag 1997;8(1):18–28.

27.

Dworkin

, Turk

, Farrar

, Haythornthwaite

, Jensen

, Katz

, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005;113(1-2):9–19.

28.

Dworkin

, Turk

, Peirce-Sandner

, Baron

, Bellamy

, Burke

, et al. Research design considerations for confirmatory chronic pain clinical trials: IMMPACT recommendations. Pain 2010;149(2):177–193.

29.

Waddell

. The Back Pain Revolution. 2nd ed. Edinburgh, UK: Churchill Livingstone 2004.

30.

Sullivan

, Adams

, Rhodenizer

, Stanish

. A psychosocial risk factor-targeted intervention for the prevention of chronic pain and disability following whiplash injury. Phys Ther 2006;86, 8–18.

31.

Besen

, Young

, Shaw

. Returning to work following low back pain: Towards a model of individual psychosocial factors J Occup Rehabil 2014;25(1):25–37.

32.

Crombez

, Vlayen

, Heuts

, Lysens

. Pain-related fear is more disabling than pain itself: Evidence on the role of pain-related fear in chronic back pain disability. Pain 1999;80:329–339.

33.

Vlaeyen

, Linton

. Fear-avoidance model of chronic musculoskeletal pain: 12 years on. Pain 2012;153(6):1144–1147.

34.

Moseley

. Unravelling the barriers to reconceptulising of the problem in chronic pain: The actual and perceived ability of patients and health profesionals to understand the neurophysiology Pain 2003;4(4):184–189.

35.

Wideman

, Hill

, Main

, Lewis

, Sullivan

, Hay

. Comparing the responsiveness of a brief, multidimensional risk screening tool for back pain to its unidimensional reference standards: The whole is greater than the sum of its parts. Pain 2012;153(11):2182–2191.

36.

Waddell

, Burton

. Occupational health guidelines for the management of low back pain at work: Evidence review. Occup Med (Lond) 2001;51.

37.

Sullivan

, Adams

, Thibault

, Corbiere

, Stanish

. Initial depression severity and the trajectory of recovery following cognitive-behavioral intervention for work disability. J Occup Rehabil 2006;16(1):63–74.

38.

Campbell

, Bishop

, Dunn

, Main

, Thomas

, Foster

. Conceptual overlap of psychological constructs in low back pain. Pain 2013;154(9):1783–1791.

39.

Mercado

, Carroll

, Cassidy

, Cote

. Passive coping is a risk factor for disabling neck or low back pain. Pain 2005;117.

40.

Casey

, Feyer

, Cameron

. Identifying predictors of early non-recovery in a compensation setting: The Whiplash Outcome Study. Injury 2011;42(1):25–32.

41.

Shaw

, Pransky

, Patterson

, Winters

. Early disability risk factors for low back pain assessed at outpatient occupational health clinics. Spine 2005;30(5):572–580.

42.

Foster

, Thomas

, Bishop

, Dunn

, Main

. Distinctiveness of psychological obstacles to recovery in low back pain patients in primary care. Pain 2010;148(3):398–406.

43.

Richard

, Dionne

, Nouwen

. Self-efficacy and health locus of control: Relationship to occupational disability among workers with back pain. J Occup Rehabil 2011;21(3):421–430.

44.

Nicholas

. The pain self-efficacy questionnaire: Taking pain into account. Eur J Pain 2007;11(2):153–163.

45.

Amick ii

, Habeck

, Ossmann

, Fossel

, Keller

, Katz

. Predictors of successful work role functioning after carpal tunnel release surgery. J Occup Environ Med 2004;46(5):490–500.

46.

Shaw

, van der Windt

, Main

, Loisel

, Linton

. Early patient screening and intervention to address individual-level occupational factors (“blue flags”) in back disability. J Occup Rehabil 2009;19(1):64–80.

47.

Truchon

, Schmouth

, Cote

, Fillion

, Rossignol

, Durand

. Absenteeism screening questionnaire (ASQ): A new tool for predicting long-term absenteeism among workers with low back pain. J Occup Rehabil 2012;22(1):27–50.

48.

Pransky

, Verma

, Okurowski

, Webster

. Length of disability prognosis in acute occupational low back pain. Spine 2006;31(6):690–697.

49.

Turk

, Dworkin

, Burke

, Gershon

, Rothman

, Scott

, et al. Developing patient-reported outcome measures for pain clinical trials: IMMPACT recommendations. Pain 2006;125(3):208–215.

50.

Linton

, Boersma

. Early Identification of Patients at Risk of Developing a Persistent Back Problem: The Predictive Validity of The Örebro Musculoskeletal Pain Questionnaire. Clin J Pain 2003;19, 80–86.

51.

Boyle

. Does item homogeneity indicate internal consistency or item redundancy in psychometric scales? Personality and Individual Differences 1991;12:291–294.

52.

Rollnick

, Miller

, Butler

. Motivational Interviewing in Health Care: Helping Patients Change Behaviour. London: The Guilford Press, 2008.

53.

Kendall

, Burton

, Main

, Watson

. Tackling musculoskeletal problems: A guide for the clinic and workplace - identifying obstacles using the psychosocial flags framework. London: The Stationery Office, 2013.