Manual Therapy 20 (2015) 666e671
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Efficacy of ‘Tailored Physical Activity’ on reducing sickness absence among health care workers: A 3-months randomised controlled trial* Lotte Nygaard Andersen a, c, *, Birgit Juul-Kristensen a, b, Kirsten Kaya Roessler c, Lene Gram Herborg d, Thomas Lund Sørensen e, Karen Søgaard a a
Institute of Sports Science and Clinical Biomechanics, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark Institute of Occupational Therapy, Physiotherapy and Radiography, Bergen University College, Bergen, Norway Department of Psychology, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark d Senior Citizen and Health Department, Social and Health Affairs, Municipality of Sonderborg, Denmark e Medical Department, Hospital of Southern Jutland, Region of Southern Denmark, Denmark b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 January 2015 Received in revised form 13 April 2015 Accepted 24 April 2015
Aim: The aim was to evaluate efficacy of “Tailored Physical Activity” (TPA) versus a reference group (REF) in reducing the number of self-reported days of sickness absence for health care workers in the Sonderborg Municipality. Methods: In this randomised controlled trial, all participants (n ¼ 54) received health guidance for 1.5 h and were randomised to TPA or REF. The primary aim was to make a comparison of participants' selfreported sickness absence due to musculoskeletal troubles measured three months after baseline. Secondary outcomes included anthropometric, health-related and physical capacity measures. Results: A TPA intervention was not significantly more effective than REF in reducing sickness absence caused by musculoskeletal troubles. However, there were significant improvements for TPA participants compared to REF in reducing pain intensity from 47.9 mm to 21.8 mm (p < .01), increasing work ability from 7.3 to 8.1 (p ¼ .04) and decreasing kinesiophobia from 26.7 to 22.5 (p < .01). A trend towards a significant improvement was seen for aerobic capacity while no effect of the intervention was found on productivity, BMI or grip strength. Conclusion: This physical activity intervention is a promising initiative for health care workers since participants achieved a substantial effect on their experience of pain, on their work ability and on their fear of physical movement relating to pain. Moreover, a difference in aerobic capacity was apparent between the sample groups. TPA however, had no significant effect in reducing sickness absence days. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Physical training Intervention Musculoskeletal Prevention Sickness absence Denmark
1. Introduction Health care workers frequently complain of musculoskeletal disorders (National Institute of Public Health, 2007; Danish Working Environment Authority, 2012; Souza and Alexandre, 2012) and make up one of the seven occupations in Denmark
* Clinical trial number: The trial is registered in the ClinicalTrials.gov, number NCT01543984. * Corresponding author. Campusvej 55, Department of Psychology, University of Southern Denmark, 5230 Odense M, Denmark. Tel.: þ45 6550 7591. E-mail addresses: [email protected] (L.N. Andersen), bjuul-kristensen@ health.sdu.dk (B. Juul-Kristensen), [email protected] (K.K. Roessler), lghe@ sonderborg.dk (L.G. Herborg), [email protected] (T.L. Sørensen), [email protected] (K. Søgaard).
http://dx.doi.org/10.1016/j.math.2015.04.017 1356-689X/© 2015 Elsevier Ltd. All rights reserved.
with the highest risk of long-term sick leave and early retirement (Danish Working Environment Authority, 2010). Health care workers are at high risk of long-term sick leave, as they are subject to considerable physical work demands that involve patient handling and manual tasks. Their work is also characterized by having prolonged periods of standing and walking and by requiring awkward postures that are potentially harmful for n et al., 1995; Waters et al., their musculoskeletal system (Torge 2006). Moreover, the demographic increase in age in the population as a whole can lead to an increased pressure on the nursing sector and may result in a parallel development of increased musculoskeletal stress among health care workers (Danish Working Environment Authority, 2010). Musculoskeletal disorders are often recurrent: they can result in chronic pain and can affect employees' functional capacity. In a
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
physically demanding job it is particularly important to maintain good musculoskeletal health, not least to ensure the workers' continuing ability to perform their task (Pohjonen, 2001). Alongside depression, pain-related work interruption and work-related pain had the largest total effect on the duration of absences from work (Franche et al., 2011). This means that focussing on pain to prevent or reduce work absence is important in any intervention aimed at preserving or enhancing work ability (Andersen et al., 2012a). Physical activity interventions involving exercises to strengthen muscles have been tested among various occupational groups and have proved to be effective in reducing pain and improving muscular strength (Andersen et al., 2008; Zebis et al., 2011). However, there is a lack of evidence on preventive activities among employees whose work demands heavy physical work, and this study will contribute knowledge on preventive activities for a specific employment group with high physical work demands. The strategy we pursue is to improve the physical capacity of the health care worker in order to prevent musculoskeletal disorders and n et al., 1995, 1999). preserve or enhance work ability (Torge The aim of this study was to evaluate “Tailored Physical Activity” (TPA) against a reference group (REF) with regard to its efficacy in reducing the number of self-reported days with sick leave and improving secondary outcomes related to pain, physical capacity, work ability and kinesiophobia. The intervention in the study arm was carefully chosen based on the results in previous evidencebased studies that have shown themselves to be effective in corresponding occupational groups (Andersen et al., 2008, 2010; Jorgensen et al., 2011; Zebis et al., 2011; Andersen et al., 2012b) and takes the participants' pain history into consideration (Horneij et al., 2004; Andersen et al., 2013). 2. Methods and design
667
The inclusion criteria were that they should: (1) be health care workers performing manual work and (2) have a history of selfreported work-related musculoskeletal pain in back or upper body. Participants who were excluded or eligible participants who did not wish to participate were registered as is recommended by the CONSORT statement (Schulz et al., 2011). 2.3. Procedure for recruitment, randomisation and allocation Workers from the Social and Health Affairs Department of Sønderborg Municipality were informed about the project through internal correspondence. Eligible participants who volunteered received written information including an invitation to an information meeting. After the meeting, written informed consent for participation in the study was provided by those participants who fulfilled inclusion criteria. The participants who were recruited were randomised in permuted blocks of 2 and 4 according to computer-generated random numbers, to participate in either TPA or REF. The study utilised an allocation concealment procedure ensuring that group allocation was not known until the participant was enrolled in the study. Concealment of the assigned intervention was ensured by having a secretary in the administration of Social and Health Affairs in Sonderborg Municipality obtain the opaque, sealed envelope containing the participant's assigned intervention in the period after the participants received health guidance and just before the intervention was initiated. Neither the investigator nor health personnel at the Health Care Centre had any other role in the sequence generation or subsequent allocation concealment. Health care professionals and participants were aware of the allocation arm but blinded to the results of baseline assessment. Health care professionals who were outcome assessors were blinded to participants' allocation.
2.1. Study design 2.4. Interventions The study is a parallel, randomised, single-blind controlled trial designed to evaluate the efficacy of TPA, including general aerobic training and specific strength training, against REF in relation to the participants' self-reported number of days of sick-leave after three months. Secondary outcomes related to pain, physical capacity, work ability, productivity and kinesiophobia. The study was conducted in Sonderborg Municipality, Denmark, between January 2012 and April 2014. In addition to the interventions, pre- and post-intervention tests and assessments were performed at the Health Care Centre in Sonderborg meaning that participants may have up to 30 min transport time from their different work places. Outcome evaluations will be performed three months after baseline and 12 months after baseline. In the present paper three months results are reported. The protocol was approved by The Regional Scientific Ethics Committee for Southern Denmark (project-ID S-20110040), The Danish Data Protection Agency, registered in the ClinicalTrials.gov, number NCT01356784 and was also published (Andersen et al., 2013). 2.2. Study population The study population consisted of health care workers e.g. nurses and nurses' aids with varying levels of health care education. They were all employed by the Social and Health Affair Department of the Municipality of Sonderborg and included participants working in municipal health care, e.g. nursing homes or home care (n ¼ 560).
All randomised participants received health guidance for 1.5 h from a trained supervisor. While the REF group received the health guidance only, the intervention group was offered TPA as an add-on health promotion activity. TPA started within one week of baseline measurements, health guidance and randomisation and was additional to any existing life style activity. Health guidance was a 1.5 h dialogue between the participant and health supervisor, based on the participant's lifestyle, motivation, resources and power to act. During the conversation participants had the opportunity to prepare a goal-oriented health plan identifying the means at their disposal to achieve the changes that they wanted and needed. The health supervisor provided input and encouraged participants to take an active part in optimising their own lives by focussing on well-being in everyday life, physical activity and weight loss. 2.5. Tailored physical activity-group (TPA) TPA sessions were performed in teams of up to 10 participants, were supervised by physiotherapists and included a standardised combination of aerobic fitness and strength training for 50 min three times per week over ten weeks during work hours. The participants were referred to one of three training programmes based on their primary region of musculoskeletal problems (neck and shoulder pain, arm and/or hand pain, lower back pain). The three standardised training programmes all started with a 5 min warm-up, during which the participants gradually increased their heart rate (HR) followed by aerobic fitness training for 20 min,
668
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
at intensities ranging from 50% with a progression up to 80% heart rate reserve. During the following weeks, progression of training was tailored to the participant's current training status and pain problems (Hofmann and Tschakert, 2011). For the warm-up and the aerobic fitness training, the participants could choose between ergometer cycling, rowing, stepping or cross training. The choice was taken after consultation with the physiotherapist and taking into consideration the participant's current musculoskeletal problems and general health. The relative workload was estimated on the basis of the known relationship between HR and oxygen uptake. HR at rest was set at 70 beats per minute and maximum HR was estimated (Tanaka et al., 2001). HR was monitored during each training session to ensure training intensity. Participants who had pain relating to the upper body and the neck were referred to high-intensity strength training programmes modified from earlier studies (Andersen et al., 2008, 2010; Zebis et al., 2011; Andersen et al., 2012b). The programme for neck and shoulder pain contained five different dumbbell exercises; one-arm row, shoulder abduction, shoulder elevation, reverse flies and upright row. The programme for pain primarily located in the arm and/or hand region contained five different dumbbell exercises: front raise, shoulder abduction, reverse flies, shoulder elevation and wrist extension. During the intervention period, the training load progressively increased from 15 repetitions maximum (~70% of maximal intensity) at the beginning of the training period to 8e12 repetitions maximum (~75e85% of maximal intensity) during the later phase. The strengthening exercises were performed in a conventional manner using consecutive concentric and eccentric muscle contractions. Three of the five different exercises with three sets per exercise were performed during each training session in an alternating manner, with shoulder elevation being the only exercise that was performed during each session (Andersen et al., 2008). Participants who had symptoms in the lower back were referred to specific strength training, corresponding to the exercises for the upper body. In addition, they performed a standardised coordination exercise programme for the lower spine composed of stabilising exercises activating the rectus abdominis, erector spinae and oblique externus muscles at intensities of more than 60% of their maximal voluntary contraction strength (Jorgensen et al., 2010, 2011). The rate of progression of all the exercises was dependent upon strength gains. To ensure standardised guidance only physiotherapists trained in accordance with the manuals for the training concepts took part in the project. The physiotherapists were encouraged to use their professional judgement to tailor each participant's programme, so that the physical demands of the programme matched the musculoskeletal condition. Training activity was recorded in a diary.
collected from a questionnaire. In addition, secondary outcome measurement were performed at baseline and repeated after three months (Andersen et al., 2013). Aerobic capacity was estimated using the Aastrand-Rhyming Test, and the result was adjusted for age and gender and normalised to body weight (Aastrand et al., 2003). Aerobic capacity testing was performed by trained physiotherapists. Hand-grip strength was measured using a digital hand-held dynamometer (American Society of Hand Therapist, 1992). The rationale for measuring hand grip strength is that it is among others correlated with mortality and development of disability (Bohannon, 2008). Additionally, responses to questions on general pain, work ability, productivity and kinesiophobia are reported. Questionnaires were completed twice, at baseline and at the end of the three months intervention. The questionnaire was constructed for use in the present project but consisted mainly of standardised and validated scales that had previously been used (Andersen et al., 2013).
3. Outcomes
4. Results
The primary aim was to make a comparison af the participants' self-reported sickness absence due to musculoskeletal troubles after three months (Andersen et al., 2013). It was evaluated using a modified question from the Nordic Musculoskeletal Questionnaire “How many days in total have you been on sick leave because of musculoskeletal trouble (such as ache, pain, discomfort) during the last 3 months?” (0 days, 1e7 days, 8e30 days, >30 days) (Kuorinka et al., 1987). At baseline, the question refers to the three months previous to the trial, while the primary endpoint measurement reflects the three months intervention period. Height and body weight were measured without shoes and wearing light clothes. Characteristics of the participants were
In the Sonderborg Municipality, Denmark, a total of 54 health care workers were included in this randomised controlled trial. The flow of participants' progress of the study is shown in Fig. 1. Baseline characteristics of the participants for TPA and REF are given in Table 1. No differences between the groups were found at baseline.
3.1. Statistical analysis A power calculation was made based on the assumption that 15% or less of the health care workers in the TPA group and 50% of the participants in the REF group reported that they have had sickness absence within the last three months, a sample size of 22 individuals in each group was required to achieve greater than 80% statistical power (one-sided, alpha ¼ 0.05). Differences in baseline characteristics between TPA and REF were tested with Pearson's c2 for categorical variables, while continuous variables were tested with an independent samples ttest. The primary analysis compared the proportion of self-reported sickness absence from work due to musculoskeletal complaints at three months follow-up between TPA and REF using a Pearson's c2. The participants' self-reported sickness absence was analysed as a dichotomous outcome indicating whether the participant had no sickness absence (0 days) because of musculoskeletal troubles, or had sickness absence during the last three months (1e7 days, 8e30 days, >30 days). When comparing continuous secondary outcomes, a linear mixed model was used for analysis of group differences, and a paired samples t-test was used for comparison of baseline and follow-up within groups. All analyses were performed in accordance to the intention-totreat-principle, i.e. all randomised participants were included in the analyses, missing values being substituted by carried forward or backward measured values leaving a delta-value of null. All results are given as mean (95% CI) or numbers (%). Similarly, mean and change values are reported with 95% CI, with p 0.05 considered statistically significant. IBM SPSS Statistics for Windows Version 21.0. (Armonk, NY:IBM Corp) was used for data analysis.
4.1. Sickness absence At follow-up, more participants from TPA than REF reported that they had no sickness absence because of musculoskeletal troubles within the last three months (Table 2), however, the difference was not significant (p ¼ .40).
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
669
Table 2 Sickness absence days. TPA, n ¼ 27
0 days (no (%)) 1e7 days (no (%)) 8e30 days (no (%)) >30 days (no (%))
REF, n ¼ 27
Baseline
Follow-up
Baseline
Follow-up
13 10 3 1
18 6 3 0
13 12 1 1
15 10 0 2
(48.1) (37.0) (11.1) (3.7)
(66.7) (22.2) (11.1) (.0)
(48.1) (44.4) (3.7) (3.7)
(55.6) (37.0) (.0) (7.4)
P ¼ .4
(Table 3). Participants in TPA group maintained hand-grip strength, while those in the REF group decreased significantly in REF from baseline to follow-up. This resulted in a significant difference between groups for left-hand but not the right-hand grip strength. No significant difference from baseline until follow-up was shown between TPA and REF for BMI. 5. Discussion
Fig. 1. Flow diagram of the progress through the study.
4.2. Pain intensity, work ability, productivity and kinesiophobia There was a significant reduction in pain intensity in TPA compared to REF from baseline to follow-up (p .01) (Table 3). Also a significant improvement in work ability (p ¼ .04) and kinesiophobia (p .01) were seen in TPA versus REF from baseline to follow-up, while no significant between-group difference was found in productivity. 4.3. Aerobic capacity, grip strength and BMI Aerobic capacity showed a strong trend toward a significant improvement in TPA compared to REF from baseline until follow-up
Table 1 Baseline characteristics.
Female/male sex e no (%) Age e mean, yr Civil status e no (%) Married/cohabitant Other Education e no (%) No and low education Short education Medium education Smoking e no (%) Yes No, but previous smoker No, never smoked Use of pain killers the last three months Daily basis One or more times a week One or more times a month Infrequent or never Physical activity status e no (%) Almost completely passive or light physical activity less than 2 h/week Light physical activity 2e4 h/week Light physical activity more than 4 h/week or more strenuous activity 2e4 h/week Strenous activity more than 4 h/week a
n ¼ 26.
TPA, n ¼ 27
REF, n ¼ 27
25/2 (92.6/7.4) 45.1 (41.4e48.8)
26/1 (96.3/3.7) 47.1 (43.6e50.6)
20 (74.1) 7 (25.9)
22 (81.5) 5 (18.5)
12 (44.4) 6 (22.2) 9 (33.3)
12 (44.4)a 4 (14.8) 10 (37.0)
6 (22.2) 9 (33.3) 12 (44.4)
5 (18.5)a 9 (33.3) 12 (44.4)
5 10 8 4
(18.5) (37.0) (29.6) (14.8)
5 6 9 6
(18.5) (22.2) (33.3) (22.2)
4 (14.8)
4 (14.8)
12 (44.4) 10 (37.0)
12 (44.4) 8 (29.6)
1 (3.7)
3 (11.1)
The present RCT on a standardised tailored physical activity intervention did not show any statistical significant effect on selfreported sickness absence days. This is in line with the systematic review from Proper et al. (2002), which overall also showed limited evidence for reduction in sickness absence. However, this is due primarily to the lack of high-quality studies. Important findings in the present study were that it showed significant improvements for TPA participants compared to REF in pain intensity, work ability and kinesiophobia and additionally a strong trend towards improved aerobic capacity. However, in spite of these positive effects TPA was not more effective than REF in reducing sickness absence days. This could probably be explained by the fact that the outcome measure was not sensitive enough to measure the changes or by the fact that sickness absence as a multifaceted outcome is difficult to improve. Another explanation of the lack of change could be recall bias, as the participants were expected to remember the number of sickness absence days during the last three months. On the other hand, strategies in the workplace aiming at enhancing physical capacity and the ability to cope with musculoskeletal pain have been successful (Andersen et al., 2010, 2012b), but mainly among sedentary workers. In the present study, where participants are required to perform more heavy work, the physical activity intervention aimed at enhancing participants' physical activity, and therefore aerobic capacity was used as an indicator of progression. Our results on aerobic capacity for health care workers are in line with previous results, and the improvement in aerobic capacity is an important indicator of the ability to prevent decline in work ability and also of the prevention of age-related deterioration of physical capacity (Pohjonen, 2001). The result on work ability strengthens the results on the health care workers' physical capacity as a highly significant improvement for TPA compared to REF. This indicates that the improvements of aerobic capacity positively affected the work ability of the health care workers as was also shown in a previous study (Pohjonen and Ranta, 2001). Furthermore, productivity improved for TPA, although not at a significant level compared to REF. The improvements in work ability are generally important for maintaining participants' attachment to the labour market. Although improvements in work ability in the present study may seem relatively small, even small changes may be of importance as indicated by a study of Sell et al. (Sell et al., 2009) showing that a one point decrease in work ability is associated with 15% increased risk of long-term sickness absence. Health care workers have previously been shown to have a high prevalence of overweight and obesity (Christensen et al., 2012) and
670
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
Table 3 Secondary outcomes. TPA, n ¼ 27 Baseline Pain VAS 0e100 (mm (CI)) Work ability, 0e10 (score (CI)) Productivity, 0e10 (score (CI)) Kinesiophobia, TSK-17, 17e68 (score(CI) BMI (kg/m2 (CI)) Aerobic capacity (ml/min/kg (CI)) Grip strength, right (kg (CI)) Grip strength, left (kg (CI))
REF, n ¼ 27 Follow-up
p
Baseline
Follow-up
p
p-Value for change
47.9 7.3 7.5 26.7
(37.9e57.8) (6.9e7.8) (6.9e8.1) (23.7e29.7)
21.8 8.1 8.1 22.5
(10.5e33.1) (7.4e8.9) (7.4e8.9) (20.5e24.5)
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Efficacy of ‘Tailored Physical Activity’ on reducing sickness absence among health care workers: A 3-months randomised controlled trial* Lotte Nygaard Andersen a, c, *, Birgit Juul-Kristensen a, b, Kirsten Kaya Roessler c, Lene Gram Herborg d, Thomas Lund Sørensen e, Karen Søgaard a a
Institute of Sports Science and Clinical Biomechanics, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark Institute of Occupational Therapy, Physiotherapy and Radiography, Bergen University College, Bergen, Norway Department of Psychology, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark d Senior Citizen and Health Department, Social and Health Affairs, Municipality of Sonderborg, Denmark e Medical Department, Hospital of Southern Jutland, Region of Southern Denmark, Denmark b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 January 2015 Received in revised form 13 April 2015 Accepted 24 April 2015
Aim: The aim was to evaluate efficacy of “Tailored Physical Activity” (TPA) versus a reference group (REF) in reducing the number of self-reported days of sickness absence for health care workers in the Sonderborg Municipality. Methods: In this randomised controlled trial, all participants (n ¼ 54) received health guidance for 1.5 h and were randomised to TPA or REF. The primary aim was to make a comparison of participants' selfreported sickness absence due to musculoskeletal troubles measured three months after baseline. Secondary outcomes included anthropometric, health-related and physical capacity measures. Results: A TPA intervention was not significantly more effective than REF in reducing sickness absence caused by musculoskeletal troubles. However, there were significant improvements for TPA participants compared to REF in reducing pain intensity from 47.9 mm to 21.8 mm (p < .01), increasing work ability from 7.3 to 8.1 (p ¼ .04) and decreasing kinesiophobia from 26.7 to 22.5 (p < .01). A trend towards a significant improvement was seen for aerobic capacity while no effect of the intervention was found on productivity, BMI or grip strength. Conclusion: This physical activity intervention is a promising initiative for health care workers since participants achieved a substantial effect on their experience of pain, on their work ability and on their fear of physical movement relating to pain. Moreover, a difference in aerobic capacity was apparent between the sample groups. TPA however, had no significant effect in reducing sickness absence days. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Physical training Intervention Musculoskeletal Prevention Sickness absence Denmark
1. Introduction Health care workers frequently complain of musculoskeletal disorders (National Institute of Public Health, 2007; Danish Working Environment Authority, 2012; Souza and Alexandre, 2012) and make up one of the seven occupations in Denmark
* Clinical trial number: The trial is registered in the ClinicalTrials.gov, number NCT01543984. * Corresponding author. Campusvej 55, Department of Psychology, University of Southern Denmark, 5230 Odense M, Denmark. Tel.: þ45 6550 7591. E-mail addresses: [email protected] (L.N. Andersen), bjuul-kristensen@ health.sdu.dk (B. Juul-Kristensen), [email protected] (K.K. Roessler), lghe@ sonderborg.dk (L.G. Herborg), [email protected] (T.L. Sørensen), [email protected] (K. Søgaard).
http://dx.doi.org/10.1016/j.math.2015.04.017 1356-689X/© 2015 Elsevier Ltd. All rights reserved.
with the highest risk of long-term sick leave and early retirement (Danish Working Environment Authority, 2010). Health care workers are at high risk of long-term sick leave, as they are subject to considerable physical work demands that involve patient handling and manual tasks. Their work is also characterized by having prolonged periods of standing and walking and by requiring awkward postures that are potentially harmful for n et al., 1995; Waters et al., their musculoskeletal system (Torge 2006). Moreover, the demographic increase in age in the population as a whole can lead to an increased pressure on the nursing sector and may result in a parallel development of increased musculoskeletal stress among health care workers (Danish Working Environment Authority, 2010). Musculoskeletal disorders are often recurrent: they can result in chronic pain and can affect employees' functional capacity. In a
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
physically demanding job it is particularly important to maintain good musculoskeletal health, not least to ensure the workers' continuing ability to perform their task (Pohjonen, 2001). Alongside depression, pain-related work interruption and work-related pain had the largest total effect on the duration of absences from work (Franche et al., 2011). This means that focussing on pain to prevent or reduce work absence is important in any intervention aimed at preserving or enhancing work ability (Andersen et al., 2012a). Physical activity interventions involving exercises to strengthen muscles have been tested among various occupational groups and have proved to be effective in reducing pain and improving muscular strength (Andersen et al., 2008; Zebis et al., 2011). However, there is a lack of evidence on preventive activities among employees whose work demands heavy physical work, and this study will contribute knowledge on preventive activities for a specific employment group with high physical work demands. The strategy we pursue is to improve the physical capacity of the health care worker in order to prevent musculoskeletal disorders and n et al., 1995, 1999). preserve or enhance work ability (Torge The aim of this study was to evaluate “Tailored Physical Activity” (TPA) against a reference group (REF) with regard to its efficacy in reducing the number of self-reported days with sick leave and improving secondary outcomes related to pain, physical capacity, work ability and kinesiophobia. The intervention in the study arm was carefully chosen based on the results in previous evidencebased studies that have shown themselves to be effective in corresponding occupational groups (Andersen et al., 2008, 2010; Jorgensen et al., 2011; Zebis et al., 2011; Andersen et al., 2012b) and takes the participants' pain history into consideration (Horneij et al., 2004; Andersen et al., 2013). 2. Methods and design
667
The inclusion criteria were that they should: (1) be health care workers performing manual work and (2) have a history of selfreported work-related musculoskeletal pain in back or upper body. Participants who were excluded or eligible participants who did not wish to participate were registered as is recommended by the CONSORT statement (Schulz et al., 2011). 2.3. Procedure for recruitment, randomisation and allocation Workers from the Social and Health Affairs Department of Sønderborg Municipality were informed about the project through internal correspondence. Eligible participants who volunteered received written information including an invitation to an information meeting. After the meeting, written informed consent for participation in the study was provided by those participants who fulfilled inclusion criteria. The participants who were recruited were randomised in permuted blocks of 2 and 4 according to computer-generated random numbers, to participate in either TPA or REF. The study utilised an allocation concealment procedure ensuring that group allocation was not known until the participant was enrolled in the study. Concealment of the assigned intervention was ensured by having a secretary in the administration of Social and Health Affairs in Sonderborg Municipality obtain the opaque, sealed envelope containing the participant's assigned intervention in the period after the participants received health guidance and just before the intervention was initiated. Neither the investigator nor health personnel at the Health Care Centre had any other role in the sequence generation or subsequent allocation concealment. Health care professionals and participants were aware of the allocation arm but blinded to the results of baseline assessment. Health care professionals who were outcome assessors were blinded to participants' allocation.
2.1. Study design 2.4. Interventions The study is a parallel, randomised, single-blind controlled trial designed to evaluate the efficacy of TPA, including general aerobic training and specific strength training, against REF in relation to the participants' self-reported number of days of sick-leave after three months. Secondary outcomes related to pain, physical capacity, work ability, productivity and kinesiophobia. The study was conducted in Sonderborg Municipality, Denmark, between January 2012 and April 2014. In addition to the interventions, pre- and post-intervention tests and assessments were performed at the Health Care Centre in Sonderborg meaning that participants may have up to 30 min transport time from their different work places. Outcome evaluations will be performed three months after baseline and 12 months after baseline. In the present paper three months results are reported. The protocol was approved by The Regional Scientific Ethics Committee for Southern Denmark (project-ID S-20110040), The Danish Data Protection Agency, registered in the ClinicalTrials.gov, number NCT01356784 and was also published (Andersen et al., 2013). 2.2. Study population The study population consisted of health care workers e.g. nurses and nurses' aids with varying levels of health care education. They were all employed by the Social and Health Affair Department of the Municipality of Sonderborg and included participants working in municipal health care, e.g. nursing homes or home care (n ¼ 560).
All randomised participants received health guidance for 1.5 h from a trained supervisor. While the REF group received the health guidance only, the intervention group was offered TPA as an add-on health promotion activity. TPA started within one week of baseline measurements, health guidance and randomisation and was additional to any existing life style activity. Health guidance was a 1.5 h dialogue between the participant and health supervisor, based on the participant's lifestyle, motivation, resources and power to act. During the conversation participants had the opportunity to prepare a goal-oriented health plan identifying the means at their disposal to achieve the changes that they wanted and needed. The health supervisor provided input and encouraged participants to take an active part in optimising their own lives by focussing on well-being in everyday life, physical activity and weight loss. 2.5. Tailored physical activity-group (TPA) TPA sessions were performed in teams of up to 10 participants, were supervised by physiotherapists and included a standardised combination of aerobic fitness and strength training for 50 min three times per week over ten weeks during work hours. The participants were referred to one of three training programmes based on their primary region of musculoskeletal problems (neck and shoulder pain, arm and/or hand pain, lower back pain). The three standardised training programmes all started with a 5 min warm-up, during which the participants gradually increased their heart rate (HR) followed by aerobic fitness training for 20 min,
668
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
at intensities ranging from 50% with a progression up to 80% heart rate reserve. During the following weeks, progression of training was tailored to the participant's current training status and pain problems (Hofmann and Tschakert, 2011). For the warm-up and the aerobic fitness training, the participants could choose between ergometer cycling, rowing, stepping or cross training. The choice was taken after consultation with the physiotherapist and taking into consideration the participant's current musculoskeletal problems and general health. The relative workload was estimated on the basis of the known relationship between HR and oxygen uptake. HR at rest was set at 70 beats per minute and maximum HR was estimated (Tanaka et al., 2001). HR was monitored during each training session to ensure training intensity. Participants who had pain relating to the upper body and the neck were referred to high-intensity strength training programmes modified from earlier studies (Andersen et al., 2008, 2010; Zebis et al., 2011; Andersen et al., 2012b). The programme for neck and shoulder pain contained five different dumbbell exercises; one-arm row, shoulder abduction, shoulder elevation, reverse flies and upright row. The programme for pain primarily located in the arm and/or hand region contained five different dumbbell exercises: front raise, shoulder abduction, reverse flies, shoulder elevation and wrist extension. During the intervention period, the training load progressively increased from 15 repetitions maximum (~70% of maximal intensity) at the beginning of the training period to 8e12 repetitions maximum (~75e85% of maximal intensity) during the later phase. The strengthening exercises were performed in a conventional manner using consecutive concentric and eccentric muscle contractions. Three of the five different exercises with three sets per exercise were performed during each training session in an alternating manner, with shoulder elevation being the only exercise that was performed during each session (Andersen et al., 2008). Participants who had symptoms in the lower back were referred to specific strength training, corresponding to the exercises for the upper body. In addition, they performed a standardised coordination exercise programme for the lower spine composed of stabilising exercises activating the rectus abdominis, erector spinae and oblique externus muscles at intensities of more than 60% of their maximal voluntary contraction strength (Jorgensen et al., 2010, 2011). The rate of progression of all the exercises was dependent upon strength gains. To ensure standardised guidance only physiotherapists trained in accordance with the manuals for the training concepts took part in the project. The physiotherapists were encouraged to use their professional judgement to tailor each participant's programme, so that the physical demands of the programme matched the musculoskeletal condition. Training activity was recorded in a diary.
collected from a questionnaire. In addition, secondary outcome measurement were performed at baseline and repeated after three months (Andersen et al., 2013). Aerobic capacity was estimated using the Aastrand-Rhyming Test, and the result was adjusted for age and gender and normalised to body weight (Aastrand et al., 2003). Aerobic capacity testing was performed by trained physiotherapists. Hand-grip strength was measured using a digital hand-held dynamometer (American Society of Hand Therapist, 1992). The rationale for measuring hand grip strength is that it is among others correlated with mortality and development of disability (Bohannon, 2008). Additionally, responses to questions on general pain, work ability, productivity and kinesiophobia are reported. Questionnaires were completed twice, at baseline and at the end of the three months intervention. The questionnaire was constructed for use in the present project but consisted mainly of standardised and validated scales that had previously been used (Andersen et al., 2013).
3. Outcomes
4. Results
The primary aim was to make a comparison af the participants' self-reported sickness absence due to musculoskeletal troubles after three months (Andersen et al., 2013). It was evaluated using a modified question from the Nordic Musculoskeletal Questionnaire “How many days in total have you been on sick leave because of musculoskeletal trouble (such as ache, pain, discomfort) during the last 3 months?” (0 days, 1e7 days, 8e30 days, >30 days) (Kuorinka et al., 1987). At baseline, the question refers to the three months previous to the trial, while the primary endpoint measurement reflects the three months intervention period. Height and body weight were measured without shoes and wearing light clothes. Characteristics of the participants were
In the Sonderborg Municipality, Denmark, a total of 54 health care workers were included in this randomised controlled trial. The flow of participants' progress of the study is shown in Fig. 1. Baseline characteristics of the participants for TPA and REF are given in Table 1. No differences between the groups were found at baseline.
3.1. Statistical analysis A power calculation was made based on the assumption that 15% or less of the health care workers in the TPA group and 50% of the participants in the REF group reported that they have had sickness absence within the last three months, a sample size of 22 individuals in each group was required to achieve greater than 80% statistical power (one-sided, alpha ¼ 0.05). Differences in baseline characteristics between TPA and REF were tested with Pearson's c2 for categorical variables, while continuous variables were tested with an independent samples ttest. The primary analysis compared the proportion of self-reported sickness absence from work due to musculoskeletal complaints at three months follow-up between TPA and REF using a Pearson's c2. The participants' self-reported sickness absence was analysed as a dichotomous outcome indicating whether the participant had no sickness absence (0 days) because of musculoskeletal troubles, or had sickness absence during the last three months (1e7 days, 8e30 days, >30 days). When comparing continuous secondary outcomes, a linear mixed model was used for analysis of group differences, and a paired samples t-test was used for comparison of baseline and follow-up within groups. All analyses were performed in accordance to the intention-totreat-principle, i.e. all randomised participants were included in the analyses, missing values being substituted by carried forward or backward measured values leaving a delta-value of null. All results are given as mean (95% CI) or numbers (%). Similarly, mean and change values are reported with 95% CI, with p 0.05 considered statistically significant. IBM SPSS Statistics for Windows Version 21.0. (Armonk, NY:IBM Corp) was used for data analysis.
4.1. Sickness absence At follow-up, more participants from TPA than REF reported that they had no sickness absence because of musculoskeletal troubles within the last three months (Table 2), however, the difference was not significant (p ¼ .40).
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
669
Table 2 Sickness absence days. TPA, n ¼ 27
0 days (no (%)) 1e7 days (no (%)) 8e30 days (no (%)) >30 days (no (%))
REF, n ¼ 27
Baseline
Follow-up
Baseline
Follow-up
13 10 3 1
18 6 3 0
13 12 1 1
15 10 0 2
(48.1) (37.0) (11.1) (3.7)
(66.7) (22.2) (11.1) (.0)
(48.1) (44.4) (3.7) (3.7)
(55.6) (37.0) (.0) (7.4)
P ¼ .4
(Table 3). Participants in TPA group maintained hand-grip strength, while those in the REF group decreased significantly in REF from baseline to follow-up. This resulted in a significant difference between groups for left-hand but not the right-hand grip strength. No significant difference from baseline until follow-up was shown between TPA and REF for BMI. 5. Discussion
Fig. 1. Flow diagram of the progress through the study.
4.2. Pain intensity, work ability, productivity and kinesiophobia There was a significant reduction in pain intensity in TPA compared to REF from baseline to follow-up (p .01) (Table 3). Also a significant improvement in work ability (p ¼ .04) and kinesiophobia (p .01) were seen in TPA versus REF from baseline to follow-up, while no significant between-group difference was found in productivity. 4.3. Aerobic capacity, grip strength and BMI Aerobic capacity showed a strong trend toward a significant improvement in TPA compared to REF from baseline until follow-up
Table 1 Baseline characteristics.
Female/male sex e no (%) Age e mean, yr Civil status e no (%) Married/cohabitant Other Education e no (%) No and low education Short education Medium education Smoking e no (%) Yes No, but previous smoker No, never smoked Use of pain killers the last three months Daily basis One or more times a week One or more times a month Infrequent or never Physical activity status e no (%) Almost completely passive or light physical activity less than 2 h/week Light physical activity 2e4 h/week Light physical activity more than 4 h/week or more strenuous activity 2e4 h/week Strenous activity more than 4 h/week a
n ¼ 26.
TPA, n ¼ 27
REF, n ¼ 27
25/2 (92.6/7.4) 45.1 (41.4e48.8)
26/1 (96.3/3.7) 47.1 (43.6e50.6)
20 (74.1) 7 (25.9)
22 (81.5) 5 (18.5)
12 (44.4) 6 (22.2) 9 (33.3)
12 (44.4)a 4 (14.8) 10 (37.0)
6 (22.2) 9 (33.3) 12 (44.4)
5 (18.5)a 9 (33.3) 12 (44.4)
5 10 8 4
(18.5) (37.0) (29.6) (14.8)
5 6 9 6
(18.5) (22.2) (33.3) (22.2)
4 (14.8)
4 (14.8)
12 (44.4) 10 (37.0)
12 (44.4) 8 (29.6)
1 (3.7)
3 (11.1)
The present RCT on a standardised tailored physical activity intervention did not show any statistical significant effect on selfreported sickness absence days. This is in line with the systematic review from Proper et al. (2002), which overall also showed limited evidence for reduction in sickness absence. However, this is due primarily to the lack of high-quality studies. Important findings in the present study were that it showed significant improvements for TPA participants compared to REF in pain intensity, work ability and kinesiophobia and additionally a strong trend towards improved aerobic capacity. However, in spite of these positive effects TPA was not more effective than REF in reducing sickness absence days. This could probably be explained by the fact that the outcome measure was not sensitive enough to measure the changes or by the fact that sickness absence as a multifaceted outcome is difficult to improve. Another explanation of the lack of change could be recall bias, as the participants were expected to remember the number of sickness absence days during the last three months. On the other hand, strategies in the workplace aiming at enhancing physical capacity and the ability to cope with musculoskeletal pain have been successful (Andersen et al., 2010, 2012b), but mainly among sedentary workers. In the present study, where participants are required to perform more heavy work, the physical activity intervention aimed at enhancing participants' physical activity, and therefore aerobic capacity was used as an indicator of progression. Our results on aerobic capacity for health care workers are in line with previous results, and the improvement in aerobic capacity is an important indicator of the ability to prevent decline in work ability and also of the prevention of age-related deterioration of physical capacity (Pohjonen, 2001). The result on work ability strengthens the results on the health care workers' physical capacity as a highly significant improvement for TPA compared to REF. This indicates that the improvements of aerobic capacity positively affected the work ability of the health care workers as was also shown in a previous study (Pohjonen and Ranta, 2001). Furthermore, productivity improved for TPA, although not at a significant level compared to REF. The improvements in work ability are generally important for maintaining participants' attachment to the labour market. Although improvements in work ability in the present study may seem relatively small, even small changes may be of importance as indicated by a study of Sell et al. (Sell et al., 2009) showing that a one point decrease in work ability is associated with 15% increased risk of long-term sickness absence. Health care workers have previously been shown to have a high prevalence of overweight and obesity (Christensen et al., 2012) and
670
L.N. Andersen et al. / Manual Therapy 20 (2015) 666e671
Table 3 Secondary outcomes. TPA, n ¼ 27 Baseline Pain VAS 0e100 (mm (CI)) Work ability, 0e10 (score (CI)) Productivity, 0e10 (score (CI)) Kinesiophobia, TSK-17, 17e68 (score(CI) BMI (kg/m2 (CI)) Aerobic capacity (ml/min/kg (CI)) Grip strength, right (kg (CI)) Grip strength, left (kg (CI))
REF, n ¼ 27 Follow-up
p
Baseline
Follow-up
p
p-Value for change
47.9 7.3 7.5 26.7
(37.9e57.8) (6.9e7.8) (6.9e8.1) (23.7e29.7)
21.8 8.1 8.1 22.5
(10.5e33.1) (7.4e8.9) (7.4e8.9) (20.5e24.5)
