Self perceptions as predictors for return to work 2 years after rehabilitation in orthopedic trauma inpatients

[Quality of Life and Return to Work Following Rehabilitation - Results of the icfPROreha Study]

BACKGROUND

Not only the severity of an injury, but also bio-psycho-social factors affect health-related quality of life and participation in social life after severe musculoskeletal injuries.

METHODS

Multicentre prospective longitudinal study with follow-up up to 78 weeks after discharge from inpatient trauma rehabilitation. Data were collected using a comprehensive assessment tool. Quality of life was assessed using the EQ-5D-5L, return to work by patients' self-reports and routine data of health insurances. Analyses of the association between quality of life and return to work, change over time in quality of life compared to the general German population and multivariate analyses to predict quality of life were conducted.

RESULT

Of 612 study participants (444 men (72.5%); M=48.5 years; SD 12.0), 502 (82.0%) returned to work 78 weeks after discharge from inpatient rehabilitation. Quality of life improved during rehabilitation treatment from 50.18 to 64.50 (mean of visual analogue scale of EQ-5D-5L) and slightly to 69.38 78 weeks after discharge from inpatient trauma rehabilitation. EQ-5D index was below the values of the general population. In total, 18 factors were selected to predict quality of life 78 weeks after discharge from inpatient trauma rehabilitation. Among others, pain at rest and suspected anxiety disorder at admission had a very strong effect on quality of life. Contextual factors such as therapies after acute care and self-efficacy also had an effect on quality of life 78 weeks after discharge from inpatient rehabilitation.

CONCLUSION

Bio-psycho-social factors affect long-term quality of life of patients with musculoskeletal injuries. Already at the time of discharge from acute treatment and even more at the beginning of inpatient rehabilitation, decisions can be made in order to achieve the best possible quality of life for those affected.

Depression symptoms reducing return to work: a meta-analysis of prospective studies

Objectives. This study was conducted to estimate the risk of return to work for people who experience symptoms of depression based on the pool of prospective data. Methods. All online articles in PubMed and Scopus which were accessible before November 2019 were searched. The odds ratios of each of the studies were pooled together to obtain an overall odds ratio. The pool of studies was with random effects. The analysis was performed based on the depression symptoms scale, type of disease and duration of follow-up. Two other aspects were examined in the analysis, one being the bias in the publication of studies and the other being the level of heterogeneity that was examined. Results. Thirty-five studies were selected for the meta-analysis. The pooled odds ratio indicates that the odds of return to work in people with depressive symptoms is 31% lower than in those without depressive symptoms. The funnel plot shows that there is asymmetry. The Egger test result was significant (p < 0.001) and there is publication bias. Conclusion. Depression symptoms after sick leave due to physical illness is a risk factor for not returning to work.

ICF-based prediction of return to work after trauma rehabilitation: Results of the icfPROreha study in patients with severe musculoskeletal injuries

Background

Physical aspects such as the type and severity of an injury are not the only factors contributing to whether or not a person can return to work (RTW) after a serious injury. A more comprehensive, biopsychosocial approach is needed to understand the complexity of RTW fully. The study aims to identify predictors of RTW 78 weeks after discharge from initial inpatient trauma rehabilitation in patients with severe musculoskeletal injuries using a biopsychosocial perspective.

Methods

This is a prospective multicenter longitudinal study with a follow-up of up to 78 weeks after discharge from trauma rehabilitation. Data on potential predictors were collected at admission to rehabilitation using a comprehensive assessment tool. The status of RTW (yes vs. no) was assessed 78 weeks after discharge from rehabilitation. The data were randomly divided into a training and a validation data set in a ratio of 9:1. On the training data, we performed bivariate and multiple logistic regression analyses on the association of RTW and potential predictors. The final logit model was selected via stepwise variable selection based on the Akaike information criterion. The final model was validated for the training and the validation data.

Results

Data from 761 patients (n = 561 male, 73.7%; mean age: 47.5 years, SD 12.3), primarily suffering from severe injuries to large joints and complex fractures of the large tubular bones, could be considered for analyses. At 78 weeks after discharge, 618 patients (81.2%) had returned to work. Eleven predictors remained in the final logit model: general health, current state of health, sensation of pain, limitations and restrictions in activities and participation (disability), professional sector, ongoing legal disputes, financial concerns (assets), personality traits, life satisfaction preaccident, attitude to life, and demand for pension claim. A predicted probability for RTW based on the multiple logistic regression model of 76.3% was revealed as the optimal cut-off score based on the ROC curve.

Conclusion

A holistic biopsychosocial approach is needed to address RTW and strengthen person-centered treatment and rehabilitation. Patients at risk for no RTW in the long term can already be identified at the onset of rehabilitation.

Systematic Review of Biopsychosocial Prognostic Factors for Return to Work After Acute Orthopedic Trauma: A 2020 Update

OBJECTIVE

To provide updated evidence on prognostic factors for return to work (RTW) in the early and late phases after acute orthopedic trauma from a biopsychosocial perspective.

METHODS

A systematic review of articles indexed in the MEDLINE, CINAHL, and Embase databases between 2010 and 2020 was performed. The inclusion criteria were cohort studies of employed populations sustaining acute orthopedic trauma with follow-up data on RTW. Biopsychosocial prognostic factors for RTW must be reported in the multiple regression models and divided into early (≤ 6 months) and late phases (> 6 months) postinjury. Two reviewers performed study selection, assessed the risk of bias and quality using the Quality in Prognosis Studies (QUIPS) tool and the Newcastle-Ottawa Scale (NOS), and extracted data independently.

RESULTS

Thirty articles were included with a follow-up period of 1-58 months. Based on the QUIPS tool, 7 studies (23%) were considered to have a low risk of bias, and 21 studies (70%) were considered to have a moderate risk of bias. Based on the NOS, the quality was high in 87% of the included studies. The RTW rates ranged from 22% to 74% in the early phase and from 44% to 94% in the late phase. In the early phase, strong evidence was found for injury severity. In the late phase, strong evidence was found for age, injury severity, level of pain, self-efficacy, educational level, blue-collar work, and compensation status; moderate evidence was found for recovery expectations and physical workload. There was limited or inconsistent evidence for the other factors.

CONCLUSION

Based on the levels of evidence, injury severity should be considered as one of the key barriers to RTW in the early and late phases postinjury. This finding underlines the need for serious injury prevention efforts. Our results also emphasize the multifaceted actions of the biopsychosocial model to facilitate RTW: promoting policies for older injured workers, improving access to medical and rehabilitation facilities, and adapting physical workload. Multiple other factors are likely important but require additional high-quality studies to assess their role in the RTW process.

Patient experiences of rehabilitation following traumatic complex musculoskeletal injury – A mixed methods pilot study

Introduction

The aim of this pilot study was to measure patient reported outcomes and evaluate their rehabilitation experience following traumatic complex musculoskeletal (CMSK) injury.

Methods

A mixed methods prospective observational study was undertaken between December 2015 and March 2018 to explore patient reported outcomes following CMSK injury sustained as a result of major trauma and perception of their rehabilitation and its impact on their recovery. The participants were asked to complete a series of outcome measures at 18-months post-injury. The data was anonymised and analysed by the lead researcher.

Results

Thirty patients were recruited into the study (19 males, 11 females) between 23 and 76 years of age (median 52 years). Their injury profile was split between open fractures 30%, pelvic fractures 23%, multiple fractures 27% and polytrauma 20%. The majority (60%) reported moderate disability at 18-months post injury with 50% returning to full time employment. Patients with multiple or open fractures reported the worst outcomes. There was no relationship between frequency, quantity or duration of physiotherapy and outcome. However, 77% reported supplementing their NHS rehabilitation with other interventions such as gym membership, hydrotherapy and psychological therapies. The thematic analysis showed that patients considered intensity, quality and coordinated timely access to rehabilitation as the most important factors.

Conclusion

From a patient perspective, the current NHS rehabilitation provision does not appear to meet their complex needs. In addition to physical and vocational rehabilitation, patients wanted effective pain management and psychological support. Strengthening current therapy services and involving other sectors (e.g. Citizen’s Advice) could help achieve this. Early access to an intensive multidisciplinary rehabilitation programme was perceived to have positive benefits and improve outcomes.

Creating Value by Prioritizing Mental and Social Health After Injury

Given the strong influence of mental and social health on symptom intensity and magnitude of limitations, attempts to increase value in orthopedic trauma must attend to emotional and social recovery. Low value and potentially harmful interventions after trauma such as excessive reliance on medication, low value surgeries for "delayed healing" or "symptomatic implants," repeated visits with a physical therapist, and other biomedical interventions often reflect misdiagnosis and mismanagement of social and mental health. A better approach is to anticipate emotional and social recovery; to get social and mental health specialists involved immediately after injury; and to develop strategies that set firm limits on biomedical tests and treatments that are unlikely to contribute to health and risk reinforcing stress, distress, and less effective coping strategies.

Cross-Cultural Adaptation of the Work Rehabilitation Questionnaire (WORQ) to French: A Valid and Reliable Instrument to Assess Work Functioning

Purpose The Work Rehabilitation Questionnaire (WORQ) is a patient-reported instrument to assess work related functioning in vocational rehabilitation (VR) and work, based on the International Classification of Functioning, Disability and Health (ICF) core set for VR. The objective of this study was to cross-culturally adapt WORQ to French and to evaluate its psychometric properties. Methods The cross-cultural adaptation followed a dual-panel approach. Psychometrics was examined in one VR-centre in the French speaking part of Switzerland. Test-retest reliability was analyzed with intraclass correlation coefficient (ICC), internal consistency was assessed with Cronbach's alpha. Construct validity was determined by convergence to the self-reported general functioning scale and the Hospital Anxiety and Depression Scales (HADS). The association of patient's return-to-work expectation within 6-months and having a case manager was examined. The usability of WORQ-French was tested in 10 VR patients. Results Eighty-nine patients with musculoskeletal injuries were included. WORQ-French showed excellent internal consistency (0.968) and a high test-retest reliability (0.935). WORQ-French was positively associated with self-reported general functioning (r = 0.662) and both HADS scales (r = 0.56-0.57). Neither the patient's return-to-work expectation nor having a case manager were significantly correlated with WORQ-French. Usability in terms of understandability of questions and response options was found to be good. Seven patients rated the length of WORQ-French as good, while two found the instrument a little too long and one found it too long. Conclusions WORQ French is a valid, reliable, and easy to administer instrument to assess self-reported work functioning given our study setting and sample characteristics.

The Nature of Trauma Pain and Its Association with Catastrophizing and Sleep

BACKGROUND

Nearly 2.8 million people are hospitalized in the USA annually for traumatic injuries, which include orthopedic and internal organ injuries. Early post-injury pain is predictive of poor outcomes, including inability to eventually return to work, and long-term psychological distress. The goal of the present study was to improve our scientific understanding of trauma-related pain by examining (1) the nature and frequency of inpatient trauma pain and (2) the associations between inpatient trauma pain, education, opioid analgesic equivalent use, pain catastrophizing, and sleep quality.

METHOD

The study included 120 patients hospitalized at a major level I regional trauma center for the care of (1) closed long bone or calcaneus fractures and/or (2) an intraabdominal injury caused by blunt force trauma and requiring surgical repair (i.e., laparotomy). Medical records were reviewed to obtain demographic information and information about opioid use during hospitalization. In addition, participants were administered measures of average pain intensity, pain catastrophizing, and sleep quality.

RESULTS

Education, opioid analgesic equivalents, catastrophizing, and poor sleep quality together accounted for 28% of the variance of average pain intensity over a 24-h period (p < .001), with each variable making a significant independent association.

CONCLUSION

Two of the factors associated with pain intensity in the study sample-catastrophizing and sleep quality-are modifiable. It is therefore possible that interventions that target these variables in patients who are hospitalized for trauma could potentially result in better long-term outcomes, including a reduced risk for developing chronic pain. Research to evaluate this possibility is warranted.

Impact of a non-return-to-work prognostic model (WORRK) on allocation to rehabilitation clinical pathways: A single centre parallel group randomised trial

INTRODUCTION

Stratified medicine might allow improvement of patient outcomes while keeping costs stable or even diminishing them. Our objective was to measure if a prediction model, developed to predict non-return to work (nRTW) after orthopaedic trauma, improves the allocation to various vocational pathways for use in clinical practice.

MATERIAL AND METHODS

Randomised-controlled trial on vocational inpatients after orthopaedic trauma (n = 280). In the intervention group, nRTW risk (estimated using the WORRK tool) was given to the clinician team before allocation of vocational pathways, while in the control group it was not. Three pathways were available: simple, coaching and evaluation (EP). Accompanying indications for interpretation of the nRTW risk were given. The primary outcome was the proportion of patients allocated to the EP. The secondary outcome was patients' and clinicians' satisfaction.

RESULTS

450 patients were assessed for eligibility, 280 included, 139 randomized to the control group (mean age 42.3years) and 141 to the intervention group (43.2years). The two groups had a similar risk profile. The patients in the intervention group were more often referred to the EP compared to the control group, but not statistically significantly more (risk ratio 1.31 [95% CI 0.70-2.46]). The number needed to treat was 30. When considering patients transferred to different pathways during rehabilitation, more patients from the intervention group were transferred to the EP over the course of the rehabilitation, increasing the risk ratio to 1.57 [95% CI 0.89 to 2.74].

DISCUSSION

The knowledge of the risk of nRTW has an influence, that is not however statistically significant and is without clinical importance as previously defined by our own power calculations (based on a 15% increase in referral to EP in the intervention group compared to the control group), on clinical decision making with regards to the allocation of patients to different physical and vocational rehabilitation programs after orthopaedic trauma. This influence is less than what was expected, possibly due to insufficient directive guidelines accompanying the WORRK model, or because clinicians associate less hours of therapy (as with certain rehabilitation programs) to disadvantaging the patient. These findings do, however, support the multi-factorial aspect of clinician decision-making.

Validation of the Readiness for Return-To-Work Scale in Outpatient Occupational Rehabilitation in Canada

Purpose To examine construct and concurrent validity of the Readiness for Return-To-Work (RRTW) Scale with injured workers participating in an outpatient occupational rehabilitation program. Methods Lost-time claimants (n = 389) with sub-acute or chronic musculoskeletal disorders completed the RRTW Scale on their first day of their occupational rehabilitation program. Statistical analysis included exploratory and confirmatory factor analyses of the readiness items, reliability analyses, and correlation with related scales and questionnaires. Results For claimants in the non-job attached/not working group (n = 165), three factors were found (1) Contemplation (2) Prepared for Action-Self-evaluative and (3) Prepared for Action-Behavioural. The precontemplation stage was not identified within this sample of injured workers. For claimants who were job attached/working group in some capacity (n = 224), two factors were identified (1) Uncertain Maintenance and (2) Proactive Maintenance. Expected relationships and statistically significant differences were found among the identified Return-To-Work (RTW) readiness factors and related constructs of pain, physical and mental health and RTW expectations. Conclusion Construct and concurrent validity of the RRTW Scale were supported in this study. The results of this study indicate the construct of readiness for RTW can vary by disability duration and occupational category. Physical health appears to be a significant barrier to RRTW for the job attached/working group while mental health significantly compromises RRTW with the non-job attached/not working group.

Assessing Psycho-social Barriers to Rehabilitation in Injured Workers with Chronic Musculoskeletal Pain: Development and Item Properties of the Yellow Flag Questionnaire (YFQ)

Purpose To develop a multidimensional scale to asses psychosocial beliefs-the Yellow Flag Questionnaire (YFQ)-aimed at guiding interventions for workers with chronic musculoskeletal (MSK) pain. Methods Phase 1 consisted of item selection based on literature search, item development and expert consensus rounds. In phase 2, items were reduced with calculating a quality-score per item, using structure equation modeling and confirmatory factor analysis on data from 666 workers. In phase 3, Cronbach's α, and Pearson correlations coefficients were computed to compare YFQ with disability, anxiety, depression and self-efficacy and the YFQ score based on data from 253 injured workers. Regressions of YFQ total score on disability, anxiety, depression and self-efficacy were calculated. Results After phase 1, the YFQ included 116 items and 15 domains. Further reductions of items in phase 2 by applying the item quality criteria reduced the total to 48 items. Phase factor analysis with structural equation modeling confirmed 32 items in seven domains: activity, work, emotions, harm & blame, diagnosis beliefs, co-morbidity and control. Cronbach α was 0.91 for the total score, between 0.49 and 0.81 for the 7 distinct scores of each domain, respectively. Correlations between YFQ total score ranged with disability, anxiety, depression and self-efficacy was .58, .66, .73, -.51, respectively. After controlling for age and gender the YFQ total score explained between R2 27% and R2 53% variance of disability, anxiety, depression and self-efficacy. Conclusions The YFQ, a multidimensional screening scale is recommended for use to assess psychosocial beliefs of workers with chronic MSK pain. Further evaluation of the measurement properties such as the test-retest reliability, responsiveness and prognostic validity is warranted.

Person-related factors associated with work participation in employees with health problems: a systematic review

PURPOSE

The objective of this systematic review was to explore and provide systematically assessed information about the association between person-related factors and work participation of people with health problems. The research question was: what is the association between selected person-related factors and work participation of workers with health problems?

METHODS

A systematic review was carried out in PubMed and PsycINFO to search for original papers published between January 2007 and February 2017. The risk of bias of the studies included was assessed using quality assessment tools from the Joanna Briggs Institute. The quality of evidence was assessed using the GRADE framework for prognostic studies.

RESULTS

In total, 113 studies were included, all of which addressed the association between person-related factors and work participation. The factors positively associated with work participation were positive expectations regarding recovery or return to work, optimism, self-efficacy, motivation, feelings of control, and perceived health. The factors negatively associated with work participation were fear-avoidance beliefs, perceived work-relatedness of the health problem, and catastrophizing. Different coping strategies had a negative or a positive relationship with work participation.

CONCLUSIONS

The results of this review provide more insight into the associations between different cognitions and perceptions and work participation. The results of this study suggest that person-related factors should be considered by occupational- and insurance physicians when they diagnose, evaluate or provide treatment to employees. Further research is required to determine how these physicians could obtain and apply such information and whether its application leads to a better quality of care.

A Return-to-Work Prognostic Model for Orthopaedic Trauma Patients (WORRK) Updated for Use at 3, 12 and 24 Months

Purpose Updating the Wallis Occupational Rehabilitation Risk (WORRK) model formula, predicting non-return to work (nRTW) at different time points (3 and 12 months) than in the validation study (2 years). Methods Secondary analysis of two samples was carried out (following orthopaedic trauma), including work status, the first at 3 months (428 patients) and the second at 12 months (431 patients) after discharge from rehabilitation. We used calibration (agreement between predicted probabilities and observed frequencies) and discrimination (area under the receiver operating characteristics curve) to assess performance of the model after fitting it in the new sample, then calculated the probabilities of nRTW based on the coefficients from the 2-year prediction. Finally, the intercepts were updated for both 3- and 12-month prediction models (re-calibration was necessary for the adjustment of these probabilities) and performance re-evaluated. Results Patient characteristics were similar in all samples (mean age 43 in both groups; 86% male at 3 months, 84% male at 12 months). The proportion of nRTW at 3 months was 63.8% and 53.4% at 12 months (50.36% at 2 years). Performance of the original WORRK for both 3- and 12-month prediction showed an AUC of 0.73, while statistically significant miscalibration was found for both time points (p < 0.001). After the updating of the intercept, calibration was improved and did not show significant miscalibration (p = 0.458 and 0.341). The AUC stayed at 0.73. Conclusion The WORRK model was successfully adapted by changing the intercept for 3- and 12-month prediction of nRTW, now available for use in clinical practice.

Predictors of non-return to work 2 years post-injury in road traffic crash survivors: Results from the UQ SuPPORT study

PURPOSE

Individuals who have sustained an injury from a road traffic crash (RTC) are at increased risk for long lasting health problems and non-return to work (NRTW). Determining the predictors of NRTW is necessary to develop screening tools to identify at-risk individuals and to provide early targeted intervention for successful return to work (RTW). The aim of this study was to identify factors that can predict which individuals will not RTW following minor or moderate injuries sustained from a RTC.

METHOD

Participants were 194 claimants (63.4% female) within a common-law "fault-based" system from the UQ SuPPORT cohort who were working prior to their RTC. Participants were assessed at 6 months on a variety of physical and mental health measures and RTW status was determined at 2 years post-RTC. RTW rate was 78.4%.

RESULTS

Univariate predictors of NRTW included being the driver or passenger, having a prior psychiatric diagnosis, high disability level, low mental or physical quality of life, predicted non-recovery, high pain, low function, high expectations of pain persistency, low expectations about RTW, having a psychiatric diagnosis, elevated depression or anxiety. The final multivariable logistic regression model included only two variables: disability level and expectations about RTW. Seventy-five percent of individuals who will not RTW by 2 years can be identified accurately at an early stage, using only these two predictors.

CONCLUSION

The results are promising, because they suggest that having information about two factors, which are easily obtainable, can predict with accuracy those who will require additional support to facilitate RTW.

Development of a Modified Version of the Spinal Function Sort (M-SFS): A Mixed Method Approach

Purpose To develop a modified version of the spinal function sort (M-SFS) by measuring work-related self-efficacy beliefs in patients with chronic low back pain. Methods A mixed method design consisting of three different methods (M1-3) was performed. In semi-structured interviews participants were asked how often they performed the activities of the 50 SFS items in 1 week, and which spinal postures and movements were associated with their back pain (M1). Quantitative analysis of previously obtained SFS data investigated internal consistency, unidimensionality, item response, and floor and ceiling effect (M2). Experts rated the SFS items based on their relevance (M3). The findings from these methods were used within a final scoring system for item reduction. Results From semi-structured interviews with 17 participants, eight new items emerged (M1). Quantitative analysis of 565 data sets (M2) revealed very high internal consistency of all items (Cronbach's alpha = 0.98) indicating item redundancy; unidimensionality of the SFS was supported by principal component analysis; good item response was confirmed by Rasch analysis; and a floor effect of four items depicting very heavy material handling was found. Experts agreed on 8 out of the 50 SFS as relevant (M3). From the original SFS, 12 items met the predefined summary score of 9. Conclusions A modified version of the SFS with 20 items has been developed. Feasibility, reliability and validity of this modified version must be tested before it can be used in clinical practice.

Care related pain in rehabilitation after orthopedic trauma: An exploratory study with qualitative data

OBJECTIVE

Care related pain (CRP) is generally under-estimated and rarely studied in rehabilitation as well as in general medecine. Beliefs about pain influence psychological distress, adjustment to pain and physical disability. In this sense, perceptions of CRP could limit recovery. This exploratory study aims to understand patients' and caregivers' subjective perceptions and beliefs about CRP.

PATIENTS AND METHODS

Questionnaires about CRP were submitted to members of the interdisciplinary team of a rehabilitation hospital and to patients with musculoskeletal complaints (cross-sectional design). Twenty patients were also individually interviewed (qualitative data). Four topics were addressed: frequency of CRP, situations and procedures causing CRP, beliefs about CRP and means used to deal with CRP.

RESULTS

Seventy-five caregivers and 50 patients replied to the questionnaire. CRP is a very common experience in rehabilitation and it is recognized by both groups. Generally, the situations causing CRP reflect the specificity of rehabilitation (mobilization…) and are similarly perceived by patients and caregivers, with patients considering them as more painful. Beliefs about CRP are clearly different from those usually associated with pain. Both groups point out the utilitarian and the inevitable character of CRP. They differ on that, that patients had a more positive view about CRP. They associate it more often with progress and see it as acceptable at least until a certain limit. They are also able to perceive the richness of means used by physiotherapists to help them coping with CRP.

CONCLUSION

Our data may suggest new keys to motivate patient to be active in rehabilitation for example in choosing carefully arguments or words which may fit theirs' beliefs about CRP, or in using various means to manage CRP. Promoting the use of relational competences with chronic pain patients and of a patient-centred approach may also be a concern in training caregivers.

Predicting non return to work after orthopaedic trauma: the Wallis Occupational Rehabilitation RisK (WORRK) model

Background

Workers with persistent disabilities after orthopaedic trauma may need occupational rehabilitation. Despite various risk profiles for non-return-to-work (non-RTW), there is no available predictive model. Moreover, injured workers may have various origins (immigrant workers), which may either affect their return to work or their eligibility for research purposes. The aim of this study was to develop and validate a predictive model that estimates the likelihood of non-RTW after occupational rehabilitation using predictors which do not rely on the worker’s background.

Methods

Prospective cohort study (3177 participants, native (51%) and immigrant workers (49%)) with two samples: a) Development sample with patients from 2004 to 2007 with Full and Reduced Models, b) External validation of the Reduced Model with patients from 2008 to March 2010. We collected patients’ data and biopsychosocial complexity with an observer rated interview (INTERMED). Non-RTW was assessed two years after discharge from the rehabilitation. Discrimination was assessed by the area under the receiver operating curve (AUC) and calibration was evaluated with a calibration plot. The model was reduced with random forests.

Results

At 2 years, the non-RTW status was known for 2462 patients (77.5% of the total sample). The prevalence of non-RTW was 50%. The full model (36 items) and the reduced model (19 items) had acceptable discrimination performance (AUC 0.75, 95% CI 0.72 to 0.78 and 0.74, 95% CI 0.71 to 0.76, respectively) and good calibration. For the validation model, the discrimination performance was acceptable (AUC 0.73; 95% CI 0.70 to 0.77) and calibration was also adequate.

Conclusions

Non-RTW may be predicted with a simple model constructed with variables independent of the patient’s education and language fluency. This model is useful for all kinds of trauma in order to adjust for case mix and it is applicable to vulnerable populations like immigrant workers.