- Summarize previous evidence on the effectiveness of workplace interventions for employees with neck and upper extremity (UE) disorders.
- Discuss the findings of the prospective cohort study on the effects of workplace interventions for neck/UE disorders in reducing sickness absence.
- Identify groups of employees most likely to benefit from workplace interventions for neck/UE disorders.
Musculoskeletal disorders are common in the working population and are widely recognized as causes of disability and sick leave.1–3 Among these disorders, neck pain is the second most common disorder after low-back pain. In general, neck pain is more prevalent in women than men, and the peak prevalence is at about 45 years of age,4 which means that neck pain is most prevalent during working age. Musculoskeletal disorders, such as neck pain, are among the main causes of both sick leave and disability pensions, leading to high costs for both the individual and society.4,5
The challenge of reducing sickness absence for individuals with musculoskeletal disorders has previously been investigated.6,7 Moderate evidence has been found for the positive effect of multidisciplinary rehabilitation in terms of improving disability and reducing sickness absence. The involvement of workplaces has also been suggested as beneficial.8,9 A Cochrane review investigating workplace intervention in work disability found that, for musculoskeletal disorders, workplace interventions were more effective than usual care10; however, most studies included in these reviews have focused mainly on low-back pain. A recent study on sickness absence among workers with neck or shoulder pain comparing a brief intervention with a multidisciplinary intervention, which in some cases involved the employer, found no significant difference between the two approaches.11 In another study, it was found that the involvement of the workplace, work adjustments, and supportive measures at the workplace were important factors for return to work after long-term sickness absence among women with neck pain.12 Factors at work, such as nonstrenuous work and recuperation, have both been linked to lower levels of sickness absence and higher work ability.13–15 Although there is some evidence that workplace interventions can reduce sickness absence in general for workers with neck and upper extremity (UE) disorders, there is very little research concerning for whom and when this might be true. The studies made in this area have mainly focused on return to work after sickness absence. In a synthesis of reviews, it was found that workers with high self-efficacy and lower severity of illness/injury had a positive association with return to work and that older workers and female workers had a negative association with return-to-work outcomes.16
The Work Ability Index (WAI) has shown to predict sickness absence, in that workers who report a lower level of work ability are at a higher risk for future sickness absence.17,18 When using workplace interventions with the aim of reducing sickness absence, this factor might affect the result of the intervention, that is, workers who report higher work ability might have a different response to the intervention than workers with a lower level of work ability. This might also be true for the length of sickness absence, which seem to play a role with regard to return to work. In a recent study, it was found that having longer sickness absence at baseline reduced the chance of returning to work within 12 months for workers with chronic musculoskeletal pain.19
The aim of this study was to investigate whether workplace interventions are effective in reducing sickness absence in persons with work-related neck and UE disorders in different groups, according to age, sex, sickness-absence period, and work ability score (WAS), and also to investigate whether disorder improvement after intervention reduces sickness absence and whether there is a difference between men and women with work-related neck and UE disorders, regarding cause, interventions, work ability, and sickness absence.
This was a follow-up study using material from Sweden's Work-related disorders survey from 2010, 2012, and 2014; from its Work Environment survey from 2009, 2011, and 2013; and from the “Longitudinal integration database for health insurance and labour market studies” (LISA) database from 2011 and 2013. All Swedish residents are included in the LISA database. Individuals in these databases were linked using the Swedish personal number in Statistics Sweden's (SCB) Microdata Online Access (MONA) system. Both the Work-related disorders and the Work Environment surveys are additions to the annual Swedish Labour Force Survey. The Work-related disorders survey is conducted by telephonic interview, with questions about work-related accidents and disorders. For inclusion in this study, the following question was used: “Now we would like to know about disorders caused by circumstances other than accidents at work. These might be workload, doing a job that is physically demanding or monotonous, or exposure to chemicals, noise, stress or, say, bullying. Have you at any time during the past 12 months suffered from disorders caused by circumstances of this kind at work?” Also used was a question about where the disorders were located: “During the past 12-month period, because of work, have you had problems with your neck, shoulder or arm?” If a worker answered yes to these two questions, the worker was included in this study.
The population eligible for the Labour Force Survey were all persons aged 15 to 74 years and registered in Sweden. After the questions in the Labour Force survey were completed, all individuals aged 16 to 64 years, employed and not on long-term sickness absence, were invited to answer questions either from the Work-related disorders survey or the Work Environment survey, depending on when the Labour Force survey was conducted. The Work Environment survey is conducted every 2 years during the last quarter of the year, and the Work-related disorders survey is conducted during the first quarter of the following year. Approximately 70% of those who participate in the Work Environment survey are also included in the Work-related disorders survey.
The Work-related disorders survey was used as the main source of data for this project, as it includes questions on work-related disorders, causes of the disorders, interventions, and self-reported sickness absence. The only question from the Work Environment survey that was used was the question on work ability, the WAS.
In the years 2010, 2012, and 2014, a total of 39,717 workers aged 16 to 64 years were asked whether they had had any work-related disorders during the last 12 months. Of these, 11,287 (28%) replied that they had had a work-related disorder (physical or other types) during the last 12 months. On the question about work-related neck and UE during the last 12 months, 1750 replied that they have or had had work-related neck or UE disorders during the last 12 months, and these respondents were included in this study (Table 1).
Cause of Disorder
For what factor at work had caused the disorder, the question used was “Now we wonder what you think caused the disorders.” For this question, several alternative answers were used for this study: vibrations, awkward postures, repetitive movements, manual handling, computer work, harassment, and stress. Respondents could choose one or more answers.
Workplace interventions were measured by the question “Were any of the following measures taken to alleviate the disorders, to stop them from getting worse or returning?” If the individual stated that interventions had taken place, the following question was asked: “What interventions have been done in the workplace where your disorders developed?” Several alternatives were used for this study: in the physical work environment, in the work organization, in the work methods, education for the workers, use of personal protection or aides, individual support, and other. Respondents could choose one or more answers. If the respondent answered that interventions had been made the following question was asked: “Have your disorder improved as a result of the intervention? (yes/no).”
Work ability was measured with the WAS question, using a score of 1 to 10: “Assume that your work ability at its best has a value of 10 points. How many points would you give your current work ability?” This question was asked in the Work Environment survey, which is done by telephonic interview. Studies have shown that the WAS question has good validity and reliability when compared with the total WAI. In this study, the WAS was categorized into four levels: 1 to 5 (poor), 6 to 7 (moderate), 8 to 9 (good), and 10 (excellent). This question was included, as it were hypothesized that sickness absence would be different in the different WAS categories and also that the effect of an intervention could be different. Of those who were included in this study, 884 persons had answered the WAS question (Table 1).
Sickness absence was measured with two different measures. The first was self-reported sick leave from the Work-related disorders survey. The question used to measure sickness absence was: “Have you during the last 12 months received sick-leave benefits (or similar), full- or part-time, as a result of these disorders?” If the individual stated that they had received sick-leave benefits, another question was asked: “How long in total during the last 12 months did the sickness absence or sick-leave benefits last as a result of these disorders?” The alternatives given for this question were 1 to 3 days, 4 days to 1 week, 8 days to 2 weeks, 2 to 4 weeks, 4 weeks to 2 months, 2 to 3 months, 3 to 6 months, 6 months to 1 year, the whole year or more. For this study, a median of each of these was used to calculate the total number of self-reported sickness absence days (S-Days).
The second measure of sickness absence used was obtained from the LISA database. For this study, only records for the years 2011 and 2013 were available; for those who answered the Work-related disorders survey in 2014, only self-reported sickness absence is used. The measure used was total net days (2 days on 50% sick leave measured as one net day) more than 14 days during the year following participation in the Work-related disorders survey (N-Days). The LISA database contains demographic information such as employment, education, family, sickness absence, disability pension, and so forth. In Sweden, sickness benefits are granted to those who have impaired work capacity due to disease or injury. There is no financial reimbursement for the first qualifying day. The initial 7 days of a sick-leave period are self-certified; after that, a medical certificate is required. Sickness benefits are covered by the employer for the first 14 days of a sick-leave spell. Thereafter, sickness benefits are granted by the Swedish Social Insurance Agency.
Wilcoxon Rank Sum Test was performed to estimate the association between workplace intervention and sickness absence, both S-Days and N-Days, and between improvement after intervention and sickness absence, both S-Days and N-Days. The analyses were stratified for sex, age group, and WAS.
Quantile regression is a semi-parametric statistical method that investigates the association of workplace intervention and several percentiles of sickness absence.20
A quantile regression analysis was used to estimate the association between receiving a workplace intervention, improvement after intervention, and sickness absence (N-Days) in the 90th to 99th percentiles. This analysis was also performed with an adjustment for sex. Coefficients from Quantile regression are interpreted similarly to coefficients of ordinary linear regression except that a quantile regression coefficient indicates the change in the value at the given percentile, not the mean, of the outcome variable.
For all analyses in this study, SAS version 9.3 (SAS Institute, Cary, North Carolina) was used. Descriptive data on the neck and UE disorders group and on the no neck or UE disorders group were derived through frequency analyses.
There were a total of 11,287 persons who reported that they have or had had work-related disorders during the last 12 months. Of these, 1750 reported neck or UE disorders. In this group, there was a larger proportion who reported “poor” or “moderate” work ability compared with the group with work-related disorders but not a work-related neck or UE disorder.
The most commonly reported causes of the work-related neck or UE disorder were, for women, work stress (38%), and among men, heavy manual work (27%) (Table 2).
A total of 622 persons (35%) reported that they had received workplace interventions, 37% of the women and 31% of the men. The most common intervention was in the physical work environment for both men and women (Table 3), with 66% of the women who received an intervention reporting that the disorder improved as a result of the intervention, and 69% of the men reporting improvement.
Among those who reported interventions, the largest proportion were found in the age group 30 to 49 years; 38% of those in that age group reported that they had received interventions. Among those who reported a WAS and reported that they had received interventions, the highest proportion was among those who reported excellent work ability (42%) (Table 3).
The mean number of self-reported median sickness absence days during the last 12 months (S-Days) was 65 days for female workers and 49 days for male workers. The mean number of registered net days more than 14 days (N-Days) during the following year was 17 for female workers and 9 for male workers (Table 4). Among the workers who reported WAS, the highest number of sickness absence days, both S-Days and N-Days, was found in the group reporting poor work ability (WAS 1-5) (Table 4). The median number of N-Days for the workers reporting neck or UE disorders was 0, for the 75th percentile was 0, for the 90th percentile was 26.5 days, and for the 95th percentile was 89.25 days.
Workplace Intervention and Sickness Absence
There was no statistical difference in either S-Days or N-Days between those who reported receiving a workplace intervention and those who did not (Tables 5 and 6). There was a statistical difference found between the groups who reported improvement of work-related neck or UE disorders after workplace intervention both in S-Days and N-Days and those who did not report improvement after intervention. For S-Days, the difference was found in the total group, and for N-Days, in the total group, among female workers and among those reporting good work ability (WAS 8-9) (Tables 7 and 8).
The quantile regression analysis showed a statistically significantly lower number of sickness absence N-Days for those workers who reported improvement after intervention on the 91st, 92nd, and 95th percentiles. The differences in number of N-Days were for the 91st percentile 54 days (15 to 69 days, P = 0.012), for the 92nd percentile 54 days (17 to 71 days, P = 0.019), and for the 95th percentile 101 days (39 to 140 days, P = 0.018). When adjusted for sex, the patterns were as above. There was also a significant difference in the sickness absence days between men and women in the 91st and 92nd percentiles (P = 0.042 and P = 0.017, respectively).
The main findings in this study suggest that workplace intervention can reduce sickness absence for workers with neck or UE disorders, among certain groups, if the intervention improves the disorder. Significantly lower sickness absence was found among those who reported improvement of their symptoms after an intervention, and within this group, a significantly lower sickness absence was found among women and those who reported good work ability (WAS 8-9).
To reduce the burden of work-related sickness absence, for both workers and society, it is important to identify who is at risk and what factors can be targeted via interventions.21 As has been shown previously and in this study, a measure of the individual's health, such as the WAS, could provide important information regarding for whom an intervention might be effective. As the lowest number of sickness absence days (both N-Days and S-Days) were found in the group who reported excellent work ability (WAS 10) and the highest in the group with poor work ability (WAS 1-5), this somewhat confirms that the measure of the WAS can provide useful information on the risk of sickness absence among workers. As we wanted to investigate for who and when intervention might affect sickness absence, we included this measure. If there would have been little difference in sickness absence between the WAS groups, then this measure would not be useful in predicting sickness absence.
A recent study on intervention in an industry setting using a participatory approach found no effect on health outcomes.22 The authors stated that one reason for this might be that the implemented action does not sufficiently reflect on the outcome. In the present study, it was shown that interventions did not significantly reduce sickness absence, unless the work-related disorder improved as a cause of the intervention. This affirms the statement that the intervention needs to sufficiently affect the health of the worker to be effective. This implies that follow-up of the workers is important to monitor the intervention for the desired effect.
In a recent review on prevention of new episodes of neck pain,23 it was found that exercise was more effective in preventing new episodes of neck pain than were ergonomic interventions. If, as the present study suggests, it is improvement of the disorder that is most important, then it could be true that interventions aimed at improving workplace ergonomics alone might not be powerful enough to reduce symptoms. As the concept of work ability can, in a simple form, be defined as a balance of demands at work and the individual's capacity to perform the work, a strategy that both strengthens the individual's capacity and also considers the demands at work would appear to be a better approach than one that only uses interventions aimed at either. The effects of work place interventions found in this study in the different subgroups were not conclusive. That only a few differences were found within the subgroups could be due to that there were few subjects in these groups, which results in low power.
Self-reported sickness absence has been shown to have a good correlation with recorded sickness absence.24,25 In this study, we used both self-reported (S-Days) and recorded (N-Days) sickness absence measures and found significant association between improvement of symptoms after intervention for both self-reported and recorded sickness absence days. This was despite the fact that the self-reported sickness absence measure only measured sickness absence due to work-related disorders, and the recorded measure used all-cause sickness absence of more than 14 days.
The measure of neck or UE disorder consisted only of a question regarding whether the worker had had work-related disorders in the neck or UE during the last 12 months that had made it difficult to perform work either at the workplace or in the home. No measurements of the intensity of the neck pain were made. This is a limitation in this study, as it is known that intensity of the neck pain is a predictor for long-term sick leave.26 The study by Holtermann et al26 used a 0 to 9 scale to describe pain intensity, with 0 being no pain and 9 being the worst pain possible. In that study, it was found that among workers with a pain intensity score at least 7, 23% had long-term sickness episodes compared with 15% among those who reported a score of 4 on pain intensity. In this study, we used the WAS as a health measure, which showed a similar result, in that individuals who reported a low WAS also had a higher mean number of sickness absence days, both S-Days and N-Days.
In this study group, which consisted of workers with work-related neck or UE disorder, 69% were women. Women also reported that work stress was the single most common cause of their work-related symptoms, with 38%. Men reported that heavy manual work was the most common cause of their neck or UE disorder. When looking at what type of intervention was performed, the most common for both men and women were physical interventions, about 35%. In this study, we stratified the study population into age and WAS groups, as it is known that both age and work ability are linked to sickness absence.16–18 In the different WAS groups, the highest proportion of interventions was found in the excellent (WAS 10) group. As a higher level of work ability are protective for sickness absence, the interventions in this group could have a lower chance of reducing sickness absence than the groups with lower work ability. Similar reasoning can be made with regards to the different age groups where interventions made in the youngest age group might not have the same effect as in the older age group with regards to the interventions effect on sickness absence. From the available data, we cannot in detail describe the types of interventions that were made or the decision-making process for selecting a certain intervention. Previous studies have suggested that multidisciplinary approaches can be beneficial, especially if the workplace is involved.8–10
Quantile regression analysis was used to not only investigate the effect of workplace interventions on the whole group but also to investigate when they were effective. The quantile analysis showed that the intervention had an effect on sickness absence when sickness absence was not short and not long. The analysis showed an effect of the intervention on sickness absence between 24 and 90 N-Days, but not for fewer or additional days. A previous study divided sickness absence into three time periods, early (≤14 days), medium–late (15 to 90 days), and late (≥90 days) for return to work, using the same data source on registered sickness absence as in this study.27 This can be interpreted to mean that workplace interventions used in this study were effective in reducing medium long (medium–late) sickness absence but not short (early) or longer (late) periods of sickness absence.
Strengths, Limitations, and Methodological Considerations
This study has several strengths. It is prospective, it is based on a representative sample of the Swedish working population, and as we used both self-reported and sickness absence from official registries, nearly all episodes of sickness absence are included. Some weaknesses of the study include the use of self-reporting by means of telephonic interview. As there is no objective measure (except sickness absence), we cannot appraise the seriousness of the disorders, whether interventions have been made or not, or whether the workers are exposed to work factors which are considered risk factors for neck and UE disorders. Also, there are some concerns with using administrative data, as there is no information on reliability and no control variables are used.28 In this study, we performed statistical analysis of several stratified subgroups. As this leads to multiple tests, there is a possibility that the findings were due to chance. However, all tests were based on the pre-planned hypothesis and no post-hoc test was made.29 Another methodological aspect of this study, like all studies of sickness absence, is that it took place in a national context. Very few studies of sickness absence have investigated whether the national context plays a role in the results.30 It should be stressed that the results of this study are found in the Swedish setting and might differ from other countries, as compensation for sickness absence varies between countries. The registry measure of sickness absence from the LISA registry covers all-cause sickness absence, and in this study, we cannot distinguish between different causes. The use of the self-reported measure on work-related sickness absence make it possible, to some extent, to compare the self-reported sickness absence and the registered all-cause sickness absence. In future research, studies investigating how workplace intervention can be used to reduce neck and UE disorders are needed and also, how measures of work ability can be used in determining for who these interventions might be effective in preventing sickness absence.
The findings in this study suggest that workplace intervention can reduce sickness absence for workers with neck or UE disorders only if the intervention improves the disorder. The interventions were most effective in reducing medium long sickness absence periods. This information can be useful for occupational health professionals when recommending work place interventions for reducing the risk of sickness absence.
1. Holtermann A, Hansen JV, Burr H, Søgaard K. Prognostic factors for long-term sickness absence among employees with neck-shoulder and low-back pain. Scand J Work Environ Health
2. Hogg-Johnson S, van der Velde G, Carroll LJ, et al. The burden and determinants of neck pain in the general population: results of the Bone and Joint Decade 2000–2010 Task Force on Neck Pain and Its Associated Disorders. Spine (Phila Pa 1976)
2008; 33 (4 suppl):S39–S51.
3. Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet
4. Hoy D, March L, Woolf A, et al. The global burden of neck pain: estimates from the global burden of disease 2010 study. Ann Rheum Dis
5. Van Rijn RM, Robroek SJW, Brouwer S, Burdorf A. Influence of poor health on exit from paid employment: a systematic review. Occup Environ Med
6. Kamper SJ, Apeldoorn AT, Chiarotto A, et al. Multidisciplinary biopsychosocial rehabilitation for chronic low back pain: Cochrane systematic review and meta-analysis. BMJ
7. Waddell G, BA, Kendall NAS. Vocational Rehabilitation. What Works, for Whom, and When? London: TSO; 2008.
8. Rinaldo U, Selander J. Return to work after vocational rehabilitation for sick-listed workers with long-term back, neck and shoulder problems: a follow-up study of factors involved. Work
9. Cullen KL, Irvin E, Collie A, et al. Effectiveness of workplace interventions in return-to-work for musculoskeletal, pain-related and mental health conditions: an update of the evidence and messages for practitioners. J Occup Rehabil
10. van Vilsteren M, van Oostrom SH, de Vet HC, Franche RL, Boot CR, Anema JR. Workplace interventions to prevent work disability in workers on sick leave. Cochrane Database Syst Rev
11. Moll LT, Jensen OK, Schiøttz-Christensen B, et al. Return to work in employees on sick leave due to neck or shoulder pain: a randomized clinical trial comparing multidisciplinary and brief intervention with one-year register-based follow-up. J Occup Rehabil
12. Ahlstrom L, Hagberg M, Dellve L. Workplace rehabilitation and supportive conditions at work: a prospective study. J Occup Rehabil
13. Lindberg P, Vingård E, Josephson M, Alfredsson L. Retaining the ability to work-associated factors at work. Eur J Public Health
14. Boschman JS, Noor A, Sluiter JK, Hagberg M. The mediating role of recovery opportunities on future sickness absence from a gender- and age-sensitive perspective. PLoS One
15. Oliv S, Noor A, Gustafsson E, Hagberg M. A lower level of physically demanding work is associated with excellent work ability in men and women with neck pain in different age groups. Saf Health Work
16. Cancelliere C, Donovan J, Stochkendahl MJ, Biscardi M, Ammendolia C, Myburgh C, et al. Factors affecting return to work after injury or illness: best evidence synthesis of systematic reviews. Chiropr Man Therap
17. Roelen CA, Heymans MW, Twisk JW, van der Klink JJ, Groothoff JW, van Rhenen W. Work ability index
as tool to identify workers at risk of premature work exit. J Occup Rehabil
18. Ahlstrom L, Grimby-Ekman A, Hagberg M, Dellve L. The work ability index
and single-item question: associations with sick leave, symptoms, and health: a prospective study of women on long-term sick leave. Scand J Work Environ Health
19. Brendbekken R, Vaktskjold A, Harris A, Tangen T. Predictors of return-to-work in patients with chronic musculoskeletal pain: a randomized clinical trial. J Rehabil Med
20. Koenker R. Reference: Quantile Regression. New York, NY: Cambridge University Press; 2005.
21. Steenstra IA, Verbeek JH, Heymans MW, Bongers PM. Prognostic factors for duration of sick leave in patients sick listed with acute low back pain: a systematic review of the literature. Occup Environ Med
22. Gupta N, Wåhlin-Jacobsen CD, Abildgaard JS, Henriksen LN, Nielsen K, Holtermann A. Effectiveness of a participatory physical and psychosocial intervention to balance the demands and resources of industrial workers: a cluster-randomized controlled trial. Scand J Work Environ Health
23. de Campos TF, Maher CG, Steffens D, Fuller JT, Hancock MJ. Exercise programs may be effective in preventing a new episode of neck pain: a systematic review and meta-analysis. J Physiother
24. Voss M, Stark S, Alfredsson L, Vingard E, Josephson M. Comparisons of self-reported and register data on sickness absence among public employees in Sweden. Occup Environ Med
25. Ferrie JE, Kivimaki M, Head J, Shipley MJ, Vahtera J, Marmot MG. A comparison of self-reported sickness absence with absences recorded in employers’ registers: evidence from the Whitehall II study. Occup Environ Med
26. Holtermann AP, Hansen JVP, Burr HP, Søgaard KP. Prognostic factors for long-term sickness absence among employees with neck-shoulder and low-back pain. Scand J Work Environ Health
27. Hensing G, Bertilsson M, Ahlborg G, Waern M, Vaez M. Self-assessed mental health problems and work capacity as determinants of return to work: a prospective general population-based study of individuals with all-cause sickness absence. BMC Psychiatry
28. Timofte D, Pantea Stoian A, Razvan H, et al. A review on the advantages and disadvantages of using administrative data in surgery outcome studies. J Surg
29. Armstrong RA. When to use the Bonferroni correction. Ophthalmic Physiol Opt
30. Schuring M, Burdorf L, Kunst A, Mackenbach J. The effects of ill health on entering and maintaining paid employment: evidence in European countries. J Epidemiol Community Health