Effect of a Work-Focused Intervention on Neck and Back Pain
Effect of a Work-Focused Intervention on Neck and Back Pain
This study was part of a large randomised controlled multicentre trial of sick-listed patients referred to the neck and back outpatient clinics at St. Olav's Hospital and Oslo University Hospital, Ullevål between August 2009 and August 2011. The primary outcome, RTW, was reported in Spine. The patients were randomised in blocks to a work-focused or control intervention using a website hosted by the medical faculty. The allocation was concealed in the data files for the researchers analysing the outcome until all analyses were run. An independent statistician generated the block size stratified by centre. The block size was concealed for all involved in the study. The first block was 20 and subsequent blocks 10.
All patients included in the study had signed an informed consent. The study was conducted in accordance with the Helsinki Declaration and the Norwegian guidelines authorised by the Data Protection for Research at Oslo University Hospital (1207–091208). The study was evaluated by the Regional Committees for Medical and Health Research Ethics in South-East Norway (S09024b 2009/1000) and registered at clinicaltrials.gov (NCT00840697).
All patients referred for diagnostic consideration or multidisciplinary treatment of neck and/or back pain were screened for eligibility at their first consultation at the outpatient clinic. The inclusion criteria were: neck and/or back pain, age 18–60 years, employed or self-employed, and duration of sick leave between 4 weeks and 12 months. Patients in need of surgical treatment were excluded from the study. Additional exclusion criteria were: cauda equina syndrome; symptomatic spinal deformity; osteoporosis with fracture; inflammatory rheumatic disease; pregnancy; legal labour dispute; insufficient Norwegian language skills; cardiac, pulmonary, or metabolic disease with functional restrictions; and DSM-V-diagnosed mental disorders.
Both work-focused and control interventions took place at the neck and back clinics of the respective hospitals, but separate teams was used for the different interventions to avoid contamination. All participants received a standard clinical examination by a physician before inclusion in the study. In this consultation relevant imaging was evaluated and patients were informed about findings and were also informed that the origin of pain is often difficult to visualise via imaging. Patients were also reassured that daily activities, physical exercise, or work would not hurt or damage their necks or backs. Emphasis was placed on removing fear avoidance, restoring activity level, and enhancing self-care and coping.
At the time of this study, the neck and back clinic at St. Olav's hospital used a comprehensive multi-disciplinary intervention based on the model described by Brox et al., whereas the neck and back clinic at Oslo University Hospital used a brief model based on the model by Indahl et al.. Both programmes were used as control interventions (Table 1). The brief intervention at Oslo University hospital consisted of the diagnostic clarification at the first visit and a session with a physiotherapist. The physiotherapist advised patients in activities and encouraged the patients to exercise. The physiotherapists also focused on reducing fear avoidance. One clarifying session with the medical specialist was also offered within 2 weeks.
The multidisciplinary intervention at St. Olav's hospital was administered by a team of medical specialists, physiotherapists and a social worker. The treatment had components of both cognitive behavioural therapy and exercise. The main focus was on reassurance, removing fear avoidance and physical conditioning.
The work-focused intervention was also a multidisciplinary intervention, and had duration of 5–6 days (Table 1). However, additional focus was placed on the RTW process and on reducing FAB of work. Patients received individual appointments with a caseworker during the first days of treatment. Work histories, family lives, and obstacles to RTW were discussed. The caseworkers contacted participants' employers by phone in most cases (unless the patient refused) to inform them of the programme and to inquire about possible temporary modifications at work. The patients created a RTW schedule together with the caseworker and the multi-disciplinary team. The patients and caseworkers also discussed relevant issues for a meeting with the employer. Additionally, the caseworkers offered the patients assistance at this meeting if requested. If sick-leave compensation was an issue, the caseworkers contacted municipal social services. The medical records and RTW schedules were sent to participants and their general practitioner, who managed the patients' sick-leave certificates.
The participants completed a comprehensive questionnaire before randomisation and at 4 and 12-months follow-up. Compliance to the treatment was assessed by the multidisciplinary team, and was defined as attending at least 50% of the treatment sessions offered.
Demographic Factors. Gender and age was recorded from patient medical records. Education was classified into four categories: up to 10 years primary school; vocational high school or general academic secondary school; college or university <4 years; and college or university ≥4 years. The two last categories were collapsed for logistic regression analyses. Occupation was manually classified using the International Standard Classification of Occupations (ISCO-88) and reported using four categories: low-skilled blue collar worker, high-skilled blue collar worker, low-skilled white collar worker, and high-skilled white collar worker.
Concurrent Treatment. Patients were asked if they had received treatment outside the hospital the last 4 months at the 4-month follow-up, and the last 8 months at 12-month follow-up. They answered 8 dichotomous questions (yes/no) about exercise by physiotherapist, other treatment by physiotherapist, manual therapy, psychomotor physiotherapy, treatment by chiropractor, alternative medicine, other rehabilitation programs or other therapy. If at least one question was answered "yes", they were considered to have had concurrent treatment.
Hospital Anxiety and Depression Scale (HADS). The level of psychological distress was assessed at baseline using the validated Norwegian version of the Hospital anxiety and depression scale (HADS). HADS has one subscale for depression (HADS-D) and another for anxiety (HADS-A). Both subscales consist of 7 items scored from 0 to 3, adding up to a sum score falling within a range of 0 to 21. High scores indicate high level of symptoms. Cases with more than one missing value in a subscale were excluded. In the case of a single missing value, the missing value was replaced with the individual mean.
Pain. Pain was measured with an 11-point numeric rating scale (NRS) ranging from 0 (no pain) to 10 (worst possible pain). Patients were asked to rate pain at rest and during activity for back/neck pain and leg/arm pain. The highest score of the four scales was used in the analysis. Changes in scores between baseline and the 12-month follow-up were computed by subtracting the scores at 12 months from the baseline scores. In some of the analyses, these scores were dichotomised using a cut-off point of 2 for the change between baseline and 12-month follow-up.
Disability. Neck and back pain-related disability was measured by the Norwegian version of the Oswestry Disability Index (ODI) for back pain patients and Neck Disability Index (NDI) for neck pain patients. Both questionnaires are composed of 10 items ranging from 0 to 5. The summed score is presented as a percentage, where 0% represents no disability and 100% represents maximum disability. In the analyses, the higher of the two scores was used if the patient had completed both questionnaires. One or two missing values were replaced with the individual mean. Scores with more than two missing values were excluded from the analysis.
Change in the ODI/NDI was calculated. In analyses with dichotomised scores, a cut-off point of 12 for the change between baseline and 12-month follow-up were applied.
Fear Avoidance. The Fear Avoidance Belief Questionnaire (FABQ) has two subscales. The physical activity subscale (FABQ-P) has four items with a possible score from 0 to 24. The work subscale (FABQ-W) has seven items with a possible score of 0 to 42. High scores indicate a high degree of fear avoidance. The minimal detectable change in the FABQ score was 9 for the physical activity subscale and 12 for the work subscale in a previous study of the Norwegian version. These values were used as the cut-off in analyses with dichotomised scores.
Return to Work. Return to work was defined as the first five-week period after randomisation that the patient did not receive sickness benefits, work assessment allowance (AAP), or disability pension from the Norwegian Labour and Welfare Administration (NAV). The five-week duration was chosen as Norwegian holidays last five weeks. Information on social compensation benefits was taken from national databases. Patients receiving partial disability pension before inclusion were considered RTW when they returned to their partly disabled status.
The sample size calculation was based on the primary outcome, RTW, and is reported elsewhere.
Comparisons of differences in baseline variables between patients who completed 12 month follow-up and patients lost to follow-up, and between the work-focused and control group were tested with t-tests for continuous variables and chi-square tests for categorical variables. Change in pain and disability from baseline to follow-up were tested using paired-sample t-tests. Differences in change scores between the work-focused and control intervention regarding pain and disability were analysed with independent samples t-tests. These analyses were carried out both with an intention to treat analyses and with available cases only (patients attending 12-month follow-up). In the intention to treat analyses multiple imputing was used to replace missing values at 12 months. Patients with missing baseline scores were not included in the analyses. The variables pain at baseline, disability at baseline, age, and return to work within one year were used to impute the missing pain scores at 12 months. The variables age, gender, education, smoking, occupation, disability at baseline, pain at baseline, and return to work within one year were used to impute disability scores at 12 months in participants with missing data at one-year follow-up. Changes in FABQ-P and FABQ-W from baseline to 4-month follow-up were calculated. Only patients with complete FABQ scores at both baseline and 4-months were included in these analyses. Subsequently, logistic regression analyses were applied to evaluate if reduction in FABQ- P and FABQ-W during the interventions influenced pain, disability and RTW within 12-month follow-up. Pain, disability and FABQ scores were dichotomised into improved and not improved in these analyses.
First, univariate logistic regression analyses, including baseline demographic variables and improvement in FABQ-P and FABQ-W from baseline to 4-month follow-up, were calculated as independent variables. Improvements in pain (≥2 points on NRS), disability (≥12 points on NDI/ODI) and RTW within 12 months were calculated as dependent variables. Subsequently, three multiple logistic regression models were built. Age, gender and intervention group were controlled for in all models and in addition we included variables with p < 0.2 from the univariate analyses. Multiple regressions controlling for baseline values of FABQ were also conducted. Correlations between independent variables were tested with Spearman's rho, and none of the variables were correlated above 0.7. Goodness of fit was tested using the Hosmer-Lemeshow test.
A two-sided significance level of p < 0.05 was used for all analyses.
All statistical analyses were performed using SPSS Statistics, version 20 (IBM corp®, Armonk NY, USA).
Methods
This study was part of a large randomised controlled multicentre trial of sick-listed patients referred to the neck and back outpatient clinics at St. Olav's Hospital and Oslo University Hospital, Ullevål between August 2009 and August 2011. The primary outcome, RTW, was reported in Spine. The patients were randomised in blocks to a work-focused or control intervention using a website hosted by the medical faculty. The allocation was concealed in the data files for the researchers analysing the outcome until all analyses were run. An independent statistician generated the block size stratified by centre. The block size was concealed for all involved in the study. The first block was 20 and subsequent blocks 10.
All patients included in the study had signed an informed consent. The study was conducted in accordance with the Helsinki Declaration and the Norwegian guidelines authorised by the Data Protection for Research at Oslo University Hospital (1207–091208). The study was evaluated by the Regional Committees for Medical and Health Research Ethics in South-East Norway (S09024b 2009/1000) and registered at clinicaltrials.gov (NCT00840697).
Participants
All patients referred for diagnostic consideration or multidisciplinary treatment of neck and/or back pain were screened for eligibility at their first consultation at the outpatient clinic. The inclusion criteria were: neck and/or back pain, age 18–60 years, employed or self-employed, and duration of sick leave between 4 weeks and 12 months. Patients in need of surgical treatment were excluded from the study. Additional exclusion criteria were: cauda equina syndrome; symptomatic spinal deformity; osteoporosis with fracture; inflammatory rheumatic disease; pregnancy; legal labour dispute; insufficient Norwegian language skills; cardiac, pulmonary, or metabolic disease with functional restrictions; and DSM-V-diagnosed mental disorders.
Procedures and Interventions
Both work-focused and control interventions took place at the neck and back clinics of the respective hospitals, but separate teams was used for the different interventions to avoid contamination. All participants received a standard clinical examination by a physician before inclusion in the study. In this consultation relevant imaging was evaluated and patients were informed about findings and were also informed that the origin of pain is often difficult to visualise via imaging. Patients were also reassured that daily activities, physical exercise, or work would not hurt or damage their necks or backs. Emphasis was placed on removing fear avoidance, restoring activity level, and enhancing self-care and coping.
At the time of this study, the neck and back clinic at St. Olav's hospital used a comprehensive multi-disciplinary intervention based on the model described by Brox et al., whereas the neck and back clinic at Oslo University Hospital used a brief model based on the model by Indahl et al.. Both programmes were used as control interventions (Table 1). The brief intervention at Oslo University hospital consisted of the diagnostic clarification at the first visit and a session with a physiotherapist. The physiotherapist advised patients in activities and encouraged the patients to exercise. The physiotherapists also focused on reducing fear avoidance. One clarifying session with the medical specialist was also offered within 2 weeks.
The multidisciplinary intervention at St. Olav's hospital was administered by a team of medical specialists, physiotherapists and a social worker. The treatment had components of both cognitive behavioural therapy and exercise. The main focus was on reassurance, removing fear avoidance and physical conditioning.
The work-focused intervention was also a multidisciplinary intervention, and had duration of 5–6 days (Table 1). However, additional focus was placed on the RTW process and on reducing FAB of work. Patients received individual appointments with a caseworker during the first days of treatment. Work histories, family lives, and obstacles to RTW were discussed. The caseworkers contacted participants' employers by phone in most cases (unless the patient refused) to inform them of the programme and to inquire about possible temporary modifications at work. The patients created a RTW schedule together with the caseworker and the multi-disciplinary team. The patients and caseworkers also discussed relevant issues for a meeting with the employer. Additionally, the caseworkers offered the patients assistance at this meeting if requested. If sick-leave compensation was an issue, the caseworkers contacted municipal social services. The medical records and RTW schedules were sent to participants and their general practitioner, who managed the patients' sick-leave certificates.
Data Collection
The participants completed a comprehensive questionnaire before randomisation and at 4 and 12-months follow-up. Compliance to the treatment was assessed by the multidisciplinary team, and was defined as attending at least 50% of the treatment sessions offered.
Demographic Factors. Gender and age was recorded from patient medical records. Education was classified into four categories: up to 10 years primary school; vocational high school or general academic secondary school; college or university <4 years; and college or university ≥4 years. The two last categories were collapsed for logistic regression analyses. Occupation was manually classified using the International Standard Classification of Occupations (ISCO-88) and reported using four categories: low-skilled blue collar worker, high-skilled blue collar worker, low-skilled white collar worker, and high-skilled white collar worker.
Concurrent Treatment. Patients were asked if they had received treatment outside the hospital the last 4 months at the 4-month follow-up, and the last 8 months at 12-month follow-up. They answered 8 dichotomous questions (yes/no) about exercise by physiotherapist, other treatment by physiotherapist, manual therapy, psychomotor physiotherapy, treatment by chiropractor, alternative medicine, other rehabilitation programs or other therapy. If at least one question was answered "yes", they were considered to have had concurrent treatment.
Hospital Anxiety and Depression Scale (HADS). The level of psychological distress was assessed at baseline using the validated Norwegian version of the Hospital anxiety and depression scale (HADS). HADS has one subscale for depression (HADS-D) and another for anxiety (HADS-A). Both subscales consist of 7 items scored from 0 to 3, adding up to a sum score falling within a range of 0 to 21. High scores indicate high level of symptoms. Cases with more than one missing value in a subscale were excluded. In the case of a single missing value, the missing value was replaced with the individual mean.
Pain. Pain was measured with an 11-point numeric rating scale (NRS) ranging from 0 (no pain) to 10 (worst possible pain). Patients were asked to rate pain at rest and during activity for back/neck pain and leg/arm pain. The highest score of the four scales was used in the analysis. Changes in scores between baseline and the 12-month follow-up were computed by subtracting the scores at 12 months from the baseline scores. In some of the analyses, these scores were dichotomised using a cut-off point of 2 for the change between baseline and 12-month follow-up.
Disability. Neck and back pain-related disability was measured by the Norwegian version of the Oswestry Disability Index (ODI) for back pain patients and Neck Disability Index (NDI) for neck pain patients. Both questionnaires are composed of 10 items ranging from 0 to 5. The summed score is presented as a percentage, where 0% represents no disability and 100% represents maximum disability. In the analyses, the higher of the two scores was used if the patient had completed both questionnaires. One or two missing values were replaced with the individual mean. Scores with more than two missing values were excluded from the analysis.
Change in the ODI/NDI was calculated. In analyses with dichotomised scores, a cut-off point of 12 for the change between baseline and 12-month follow-up were applied.
Fear Avoidance. The Fear Avoidance Belief Questionnaire (FABQ) has two subscales. The physical activity subscale (FABQ-P) has four items with a possible score from 0 to 24. The work subscale (FABQ-W) has seven items with a possible score of 0 to 42. High scores indicate a high degree of fear avoidance. The minimal detectable change in the FABQ score was 9 for the physical activity subscale and 12 for the work subscale in a previous study of the Norwegian version. These values were used as the cut-off in analyses with dichotomised scores.
Return to Work. Return to work was defined as the first five-week period after randomisation that the patient did not receive sickness benefits, work assessment allowance (AAP), or disability pension from the Norwegian Labour and Welfare Administration (NAV). The five-week duration was chosen as Norwegian holidays last five weeks. Information on social compensation benefits was taken from national databases. Patients receiving partial disability pension before inclusion were considered RTW when they returned to their partly disabled status.
Sample Size
The sample size calculation was based on the primary outcome, RTW, and is reported elsewhere.
Statistical Methods
Comparisons of differences in baseline variables between patients who completed 12 month follow-up and patients lost to follow-up, and between the work-focused and control group were tested with t-tests for continuous variables and chi-square tests for categorical variables. Change in pain and disability from baseline to follow-up were tested using paired-sample t-tests. Differences in change scores between the work-focused and control intervention regarding pain and disability were analysed with independent samples t-tests. These analyses were carried out both with an intention to treat analyses and with available cases only (patients attending 12-month follow-up). In the intention to treat analyses multiple imputing was used to replace missing values at 12 months. Patients with missing baseline scores were not included in the analyses. The variables pain at baseline, disability at baseline, age, and return to work within one year were used to impute the missing pain scores at 12 months. The variables age, gender, education, smoking, occupation, disability at baseline, pain at baseline, and return to work within one year were used to impute disability scores at 12 months in participants with missing data at one-year follow-up. Changes in FABQ-P and FABQ-W from baseline to 4-month follow-up were calculated. Only patients with complete FABQ scores at both baseline and 4-months were included in these analyses. Subsequently, logistic regression analyses were applied to evaluate if reduction in FABQ- P and FABQ-W during the interventions influenced pain, disability and RTW within 12-month follow-up. Pain, disability and FABQ scores were dichotomised into improved and not improved in these analyses.
First, univariate logistic regression analyses, including baseline demographic variables and improvement in FABQ-P and FABQ-W from baseline to 4-month follow-up, were calculated as independent variables. Improvements in pain (≥2 points on NRS), disability (≥12 points on NDI/ODI) and RTW within 12 months were calculated as dependent variables. Subsequently, three multiple logistic regression models were built. Age, gender and intervention group were controlled for in all models and in addition we included variables with p < 0.2 from the univariate analyses. Multiple regressions controlling for baseline values of FABQ were also conducted. Correlations between independent variables were tested with Spearman's rho, and none of the variables were correlated above 0.7. Goodness of fit was tested using the Hosmer-Lemeshow test.
A two-sided significance level of p < 0.05 was used for all analyses.
All statistical analyses were performed using SPSS Statistics, version 20 (IBM corp®, Armonk NY, USA).