Integrated health programme: a workplace randomized controlled trial
Abstract
Title. Integrated health programme: a workplace randomized controlled trial.
Aim. This paper is the report of a pilot study to assess if an Integrated Health Programme would reduce sick leave and subjective health complaints, and increase coping in a population of nursing personnel.
Background. The work group in Norway with most sick leave is healthcare workers. More than 50% of the sick leave is because of subjective health complaints. Work place physical exercise interventions have a documented positive effect on sick leave.
Method. After baseline screening, employees who had agreed to participate (n = 40) were randomized to an intervention or control group. The intervention group participated in an Integrated Health Programme twice weekly during working hours. The programme consisted of physical exercise, stress management training, health information and an examination of the participants’ workplace. The control group was offered the same intervention after the project was finished. This study was carried out from 2001 to 2002.
Findings. There were no statistically significant effects on sick leave or health-related quality of life. The intervention group reported fewer neck complaints compared to the control group, but otherwise there were no effects on subjective health complaints. However, the subjective effects were large and highly statistically significant, the intervention group reporting improvement in health, physical fitness, muscle pain, stress management, maintenance of health and work situation.
Conclusion. The Integrated Health Programme was not effective in reducing sick leave and subjective health complaints, but may be of use to employers wanting to increase employee job satisfaction and well-being.
What is already known on this topic
- •
Workplace physical exercise interventions have a documented positive effect on musculoskeletal complaints and sick leave.
- •
Nursing personnel report high levels of musculoskeletal complaints.
- •
Nursing personnel is a work group with high levels of sick leave.
What this paper adds
- •
Physical exercise, stress management training, and health information had no effect on sick leave and subjective health complaints.
- •
Physical exercise, stress management training, and health information had no effect on health related quality of life.
- •
The intervention increased employee satisfaction and well-being.
Implications for practice and/or policy
- •
Employers could use the Integrated Health Programme to increase employee job satisfaction and well-being.
Introduction
In western countries the prevalence of subjective health complaints (SHC) is high (Agreus 1998, Eriksen et al. 1998, Kind et al. 1998, Bassols et al. 1999, Makela et al. 1999, Ihlebaek et al. 2002, Picavet & Hazes 2003). Healthcare workers report the highest level of SHCs compared with other work groups (Ihlebæk & Eriksen 2003). Physical exercise is found to have positive effect on both complaints and sick leave from musculoskeletal complaints (Linton & Van Tulder 2001, Waddell & Burton 2001, Tveito et al. 2004a, Busch et al. 2007).
A high level of sick leave (National Insurance Administration, 2006b) and retirement with disability pension (National Insurance Administration, 2006c) is becoming an economic burden to the Norwegian society (National Insurance Administration, 2006a). The work group with the highest level of sick leave is nursing personnel (Statistics Norway, 2006) and this also is the work group with the highest number of disability pensioners (Norges Offentlige Utredninger 2004).
Subjective health complaints, including musculoskeletal complaints, are the cause of more than half of the sick leave in Norway (National Insurance Administration, 2005). SHCs are conditions without objective pathological findings, or where the complaints are stronger than expected based on the objective findings (Eriksen et al. 1999). Musculoskeletal complaints is one group of SHCs, and pseudoneurology (i.e. tiredness, sleep problems, mood changes) and gastrointestinal complaints are others (Eriksen et al. 1999) . Musculoskeletal complaints alone are the cause of nearly half of the sick leave in Norway (National Insurance Administration, 2005).
Nursing personnel are considered a high risk group for musculoskeletal complaints (Kaplan & Deyo 1988, Ljungberg et al. 1989, Eriksen 2003, Vieira et al. 2006) and reports of musculoskeletal complaints are associated with reports of low health-related quality of life (Tveito et al. 2004b). Studies of the effect of physical exercise and musculoskeletal symptoms report contradictory findings, with some demonstrating reduced symptoms (Alexandre et al. 2001) and reduced risk of sick leave (Gundewall et al. 1993, Eriksen 2003) in nursing personnel and other working populations (Tveito et al. 2004a), whilst others have not found this (Horneij et al. 2001, Oldervoll et al. 2001, Eriksen et al. 2002, Pernold et al. 2005). Occupational interventions in the nursing profession not including physical exercise or where physical exercise was a minor part of the intervention; did not have any positive effect on sick leave or musculoskeletal complaints (Bos et al. 2006).
These conflicting findings point to a need for more research to assess the effect of physical exercise on the health of nursing personnel. In one of the studies, where physical exercise did not have a positive effect on musculoskeletal pain (Horneij et al. 2001), the authors suggest that as the aetiology of musculoskeletal complaints is multifactorial, and that a combination of the two interventions offered in the study, physical exercise and stress management training, might be effective and should be further studied. In another study (Eriksen et al. 2002) the intervention, which combined physical exercise with cognitive factors and health information, seemed the most efficient, even if the effects were not statistically significant. Based on these results, an intervention consisting of physical exercise, stress management training and health information seems to be the most efficient approach to motivate employees to life-style changes.
The study
Aims
The aims of the study were to assess if an Integrated Health Programme would reduce sick leave and SHCs, and to increase coping in a population of nursing personnel.
Our hypotheses were: (1) the intervention group will have a reduction in musculoskeletal complaints and sick leave compared with the control group after having participated in the Integrated Health Programme and (2) the intervention group will increase their coping skills after having participated in the Integrated Health Programme.
Design
A randomized controlled pilot study was performed and the data were collected in 2001.
Participants
The 62 employees in a nursing home for older people in Norway, all working more than 49% of full time, were invited to participate in the project ‘An integrated health programme for employees in nursing homes for older people’. A convenience sample of 45 employees agreed to participate. Most of the employees were nursing auxiliaries and nurses, but the sample also included some assistants without formal education and other helping staff.
Data collection
Baseline data were collected in September 2001 and the study was completed in 2002. There were only five males in the study and they were excluded from the analyses as the number was too small to allow separate analyses. The groups were also too small to allow separate analyses based on professional background. The participants were randomized to intervention (n = 19) or control (n = 21) group (see Figure 1). The intervention group was granted leave from work to participate in the intervention twice a week. The post-test was performed in June 2002 when the intervention was finished.
Flow chart.
Intervention
The intervention group
The ‘Integrated Health Programme’ (Eriksen et al. 1994, 2002) consisted of three main parts: physical exercise, health information/stress management training and a practical examination of the workplace. The programme was delivered by an aerobics instructor with healthcare education (nurse). The physical exercise was based on a standardized aerobic dancing programme (Bø & Kamhaug 1989, Eriksen et al. 1994) and was given for one hour three times a week for 9 months. The programme consisted of body awareness (5 minutes), warm-up/aerobics/ergonomics (25 minutes), cool-down exercises (5 minutes), strength/stabilizing (10 minutes), stretching (10 minutes) and relaxation (5 minutes). The general aim of the programme was to improve physical capacity, muscle strength and flexibility.
During the nine months of the intervention the participants were given 15 hours of information on stress, coping, health and lifestyle and a practical examination of the workplace was carried out. The information sessions were held once a week from September to December 2001 and lasted 1 hour. Health and lifestyle information focused on exercise, nutrition, sleep, smoking, activity and musculoskeletal problems. The stress management training focused on both the positive and negative consequences of stress and how to cope with stress. Discussions were encouraged. In the practical examination of the work place, the participants contributed with their own experiences of how to organize and cope with the job.
The control group
The control group participated in the pre- and post-tests. They were offered the intervention programme after the intervention was finished and were also granted leave from work to participate.
Randomization
The randomization was concealed and performed in blocks of ten. The participants were stratified at their department level to ensure that the departments were left with sufficient staff when the intervention group participated in the intervention programme. The randomization procedure was performed with a table of random numbers and a list of the employees who had agreed to participate in the project. The administrative staff of the research group performed the randomization. The research group had no knowledge of the participants. The participants were randomized after the initial screening and testing.
Lost to follow-up
Eleven people did not attend for the post-test – seven in the intervention group and four in the control group – a drop-out rate of 28%. In the intervention group, two of the drop-out individuals were on leave (family reasons) and one was on leave receiving disability compensation. One individual never participated in the intervention. For the last three, the reason for drop-out was unknown. Of the four drop-out individuals in the control group, three were on leave (family reasons) and one was off sick.
Ethical considerations
The project received the appropriate ethics committee approval. Before the start of the study, the employees were informed about the project by their managers and employee organizations. An information pamphlet describing the project was distributed. All participants signed an informed consent form and were informed about their rights according to the Helsinki declaration. They were assigned an identification number and treated anonymously in all analyses.
Outcome measures
The main outcome measures were sick leave and SHCs. Secondary outcome measures were coping, job stress, effort reward imbalance, demands, control, health related quality of life and subjective effects of the intervention.
All participants in the intervention and control groups had a brief health check and filled in a questionnaire at pre- and post-test. Data on sick leave from the year before the start of the intervention, the intervention year, and the year after the intervention were supplied by the nursing home.
All data were measured with Norwegian versions of questionnaires covering a broad range of variables including demographic variables, lifestyle, health-related quality of life, SHCs, coping, social support and work variables.
Subjective health complaints were measured using the SHCs inventory (Eriksen et al. 1999). This lists 29 items on subjective somatic and psychological complaints in the previous 30 days. Severity was scored on a 4-point scale, from 0 – no complaints, to 3 – severe complaints. Five subscales were computed, allergy (five items, α = 0·43), flu (two items, α = 0·84), musculoskeletal complaints (headache, neck pain, upper back pain, low back pain, arm pain, shoulder pain, migraine and leg pain, eight items, α = 0·78), ‘pseudoneurology’ (American Psychiatric Association, 1994) (palpitations, heat flushes, sleep problems, tiredness, dizziness, anxiety, and sadness, seven items, α = 0·62), and gastrointestinal complaints (heartburn, epigastric discomfort, ulcer/non-ulcer dyspepsia, stomach pain, gas discomfort, diarrhoea, and constipation, seven items, α = 0·36).
The generic health status inventory SF-36 for health situations during the last 4 weeks was used to measure health-related quality of life (Ware et al., 1994, 2000, Loge et al. 1998). The 36 items are grouped into eight factors: physical functioning (10 items, α = 0·89), role limitations caused by physical problems (four items, α = 0·84), role limitations caused by emotional problems (three items, α = 0·7), bodily pain (two items, α = 0·94), social functioning (two items, α = 0·81), mental health (five items, α = 0·65), vitality (four items, α = 0·88) and general health perceptions (five items, α = 0·89). In addition, health transition over the past year was measured. The scoring of the items varied from dichotomous scales (yes/no) to 6-point ordinal scales.
Adjusted SF-36 scores were calculated by using each individual’s score on the eight SF-36 factors, subtracting the age and gender specific normative score, dividing by the standard deviation in the general population, multiplying by 10 and adding 50 (Loge & Kaasa 1998, Ware et al. 2000). The mean is 50, and a deviation of 10 points from the mean represents one standard deviation. A low score is one below 50 and implies low quality of life, while a high score is one above 50 and implies high quality of life.
Coping was measured by the Instrumental Mastery Oriented Coping Factor from the CODE (Eriksen et al. 1997), based on the Utrecht Coping List (UCL) (Schreurs et al. 1988, 1993). Instrumental Mastery Oriented Coping (active problem-solving, avoidance and passive expectancy, and depressive reaction pattern, 22 items, α = 0·77) implies an instrumental, active, goal-oriented coping style, with strategies such as direct intervention, considering different solutions to the problem, and considering the problem a challenge (Eriksen et al. 1997). For a high coping score, the score on active problem-solving must be high and scores on avoidance and passive expectancy and on depressive reaction pattern must be low.
The psychological demands factor was measured by five questions from the short Swedish version (Theorell et al., 1991) of the Psychological Demands dimension from the Demand/Control Model (Karasek & Theorell 1990). The items were scored on a 4-point scale from 1 (yes, often) to 4 (no, nearly never), yielding a sum score for psychological demands (α = 0·79). High demands are related to working hard and fast, excessive work, insufficient time for work tasks or conflicting demands.
Control (decision latitude) was measures by six items from the short Swedish version (Theorell et al., 1991) of the Decision Latitude dimension from the Demand/Control Model (Karasek & Theorell 1990). Four items refer to skill discretion and two items to decision authority. The items were scored on a 4-point scale from 1 (yes, often) to 4 (no, nearly never), yielding a sum score for control (α = 0·56). High score indicates high level of control.
Subjective effects of the intervention were measured by seven items developed for an earlier intervention study to capture the perception of improvement attributed to the intervention by the employees (Eriksen et al. 2002). The seven items are used descriptively as single items. Participants were asked what the intervention had meant to them by scoring the following statements on a 5-point scale from 1 (much better) to 5 (much worse):
- •
My health has become…
- •
The work environment has become…
- •
My physical fitness has become…
- •
My work situation has become…
- •
My muscle pain has become…
- •
My management of stress has become…
- •
My ability to maintain my health has become…
Data analysis
All statistical analyses were performed with spss 13.0 for Windows. Differences between groups were assessed by one-way anova and chi-squared tests. Risk ratios and 95% confidence intervals (CI) were calculated for the subjective effects of the intervention.
Results
Baseline characteristics
There were no statistically significant differences between the participants in the intervention and control groups at baseline (see Table 1).
| Intervention group (n = 19) | Control group (n = 21) | ||
|---|---|---|---|
| Sick leave | |||
| Baseline | 19·7 (5·7–33·7)* | 20·6 (6·5–34·7) | F(1,38) = 0·008, P = 0·931 |
| Post-test | 36·0 (5·2–66·8) | 35·2 (14·1–56·2) | F(1,37) = 0·002, P = 0·963 |
| 1 year follow-up | 52·6 (12·8–92·3) | 54·4 (16·8–91·9) | F(1,31) = 0·005, P = 0·945 |
| Musculoskeletal complaints | |||
| Baseline | 5·6 (3·4–7·7) | 5·7 (3·6–7·7) | F(1,38) = 0·003, P = 0·955 |
| Post-test | 3·5 (0·9–6·2) | 5·5 (3·4–7·7) | F(1,27) = 1·580, P = 0·220 |
| Pseudoneurological complaints | |||
| Baseline | 4·2 (2·8–5·6) | 2·7 (1·8–3·5) | F(1,38) = 4·036, P = 0·052 |
| Post-test | 2·8 (1·1–4·5) | 2·5 (1·2–3·8) | F(1,27) = 0·094, P = 0·761 |
| Gastrointestinal complaints | |||
| Baseline | 1·5 (0·7–2·2) | 1·6 (0·8–2·4) | F(1,38) = 0·077, P = 0·783 |
| Post-test | 0·8 (0·1–1·6) | 1·7 (0·8–2·6) | F(1,27) = 2·069, P = 0·162 |
| Coping | |||
| Baseline | 3·1 (3·0–3·2) | 3·0 (2·9–3·1) | F(1,38) = 2·126, P = 0·153 |
| Post-test | 3·0 (2·9–3·2) | 3·0 (2·9–3·1) | F(1,27) = 0·026, P = 0·872 |
| Jobstress | |||
| Baseline | 23·7 (15·6–31·9) | 26·9 (19·6–34·2) | F(1,38) = 0·361, P = 0·552 |
| Post-test | 22·1 (14·7–29·6) | 28·5 (18·6–38·4) | F(1,27) = 1·119, P = 0·299 |
| Effort reward imbalance† | |||
| Baseline (n = 35) | 27% | 15% | χ2 = 0·729, P = 0·393 |
| Post-test (n = 24) | 18% | 15% | χ2 = 0·034, P = 0·855 |
| Demands | |||
| Baseline | 14·6 (13·4–15·8) | 14·2 (12·9–15·5) | F(1,38) = 0·199, P = 0·658 |
| Post-test | 14·5 (12·6–16·5) | 14·3 (12·8–15·8) | F(1,27) = 0·041, P = 0·842 |
| Control | |||
| Baseline | 17·1 (16·3–17·9) | 17·8 (16·7–18·9) | F(1,37) = 1·047, P = 0·313 |
| Post-test | 17·4 (16·2–18·5) | 17·6 (16·3–19·0) | F(1,27) = 0·100, P = 0·754 |
| Physical functioning | |||
| Baseline | 48·1 (44·7–51·5) | 49·5 (45·9–53·1) | F(1,36) = 0·353, P = 0·556 |
| Post-test | 49·9 (44·1–55·7) | 49·1 (43·2–55·0) | F(1,27) = 0·039, P = 0·845 |
| Role physical | |||
| Baseline | 49·8 (45·2–54·5) | 47·8 (42·4–53·1) | F(1,37) = 0·355, P = 0·555 |
| Post-test | 50·0 (44·2–55·9) | 48·9 (43·6–54·3) | F(1,27) = 0·085, P = 0·773 |
| General health | |||
| Baseline | 42·3 (37·8–46·8) | 45·7 (41·7–49·7) | F(1,38) = 1·374, P = 0·248 |
| Post-test | 49·4 (43·5–55·3) | 44·7 (38·1–51·2) | F(1,27) = 1·264, P = 0·271 |
| Bodily pain | |||
| Baseline | 46·9 (41·5–52·3) | 45·6 (41·3–49·9) | F(1,38) = 0·163, P = 0·689 |
| Post-test | 49·9 (43·3–56·4) | 45·3 (39·4–51·3) | F(1,27) = 1·201, P = 0·283 |
| Vitality | |||
| Baseline | 44·9 (40·1–49·7) | 47·3 (42·2–52·4) | F(1,38) = 0·517, P = 0·477 |
| Post-test | 48·6 (43·5–53·7) | 48·5 (42·5–54·6) | F(1,26) = 0·000, P = 0·991 |
| Social functioning | |||
| Baseline | 45·2 (40·1–50·3) | 49·3 (45·5–53·1) | F(1,38) = 1·873, P = 0·179 |
| Post-test | 50·4 (46·0–54·7) | 48·5 (42·9–54·1) | F(1,27) = 0·294, P = 0·592 |
| Role emotional | |||
| Baseline | 51·0 (45·8–56·3) | 50·9 (46·4–55·3) | F(1,37) = 0·003, P = 0·954 |
| Post-test | 51·3 (47·1–55·6) | 49·4 (43·3–55·5) | F(1,27) = 0·287, P = 0·596 |
| Mental health | |||
| Baseline | 47·3 (42·7–51·9) | 49·8 (45·9–53·7) | F(1,38) = 0·762, P = 0·388 |
| Post-test | 52·9 (48·4–57·3) | 52·9 (49·4–56·4) | F(1,26) = 0·001, P = 0·976 |
- *Mean (95% CI).
- †Differences tested with χ2 tests. Percentage of participants with effort reward imbalance.
- All other differences were tested with one-way anova.
Participants lost to follow-up
The participants lost to follow-up were younger [mean 37 (95% CI 28·7–45·6)] than those who completed the study [mean 45 (95% CI 42·0–48·9), F(1,38) = 5·4, P = 0·025]. There were no statistically significant differences on sick leave at baseline (September 2000–2001, year 1), but participants who completed the study had statistically significantly fewer days of sick leave in the intervention year (September 2001–2002, year 2) [mean 25·6 (95% CI 8·4–42·8); F(1,37) = 4·3, P = 0·045] and the year after the intervention (September 2002–2003, year 3) [mean 38·4 (95% CI 14·9–61·9); F(1,31) = 6·0, P = 0·020] than those lost to follow-up. Otherwise there were no statistically significant differences on any demographic or outcome variables at baseline between participants who completed the study and those who were lost to follow-up.
Effects of the intervention
There were no statistically significant differences between the intervention and control groups regarding sick leave (see Table 1). Number of days of sick leave increased by a factor of 2·6 in the study population from the year before the intervention [mean 20 (95% CI 11–30)] to 1 year post-intervention [mean 53 (95% CI 28–79)].
There were no statistically significant differences between the intervention and control groups on SHCs sum scales, coping, job stress, effort reward imbalance, demands, control or any of the SF-36 scales (see Table 1). The intervention group reported fewer neck complaints than the control group [8% vs. 48% reported substantial complaints (score above 1), χ2 = 5·2, P = 0·023]. There were no differences in reports of complaints in the upper back (15% vs. 13% reported substantial complaints, χ2 = 0·024, P = 0·877) or the low back (8% vs. 13% reported substantial complaints, χ2 = 0·179, P = 0·672).
There were statistically significant differences between the intervention and control groups in subjective effects (see Table 2). The intervention group reported that their health, physical fitness and work situation had become better, they had less muscle pain, were better at stress management and better at maintaining their health than the control group. In reports of how the project had influenced their work environment, there were no differences between the groups. We recorded no negative effects of the intervention.
| RR (95% CI) n = 27 |
|
|---|---|
| Better health | |
| Control group | 1 |
| Intervention group | 3·3 (1·3–8·0) |
| Better work environment | |
| Control group | 1 |
| Intervention group | 1·5 (0·7–3·2) n.s. |
| Better physical fitness | |
| Control group | 1 |
| Intervention group | 11·1 (1·7–74·1) |
| Better work situation | |
| Control group | 1 |
| Intervention group | 2·2 (1·1–4·2) |
| Less muscle pain | |
| Control group | 1 |
| Intervention group | 3·2 (1·5–7·0) |
| Better stress management | |
| Control group | 1 |
| Intervention group | 2·5 (1·3–4·8) |
| Better health maintenance | |
| Control group | 1 |
| Intervention group | 2·8 (1·1–6·9) |
Discussion
There were no statistically significant effects of the intervention on sick leave, SHCs summed scales, health-related quality of life, stress or coping. Complaints in the neck were statistically significantly better after the intervention in the intervention group compared with the control group, but there were no differences between the groups regarding complaints in the upper and low back. The subjective effects of the health and well-being of the participants were statistically significantly better in the intervention group than in the control group.
Mean number of days of sick leave per year in the study population more than doubled during the 3 years of measurement. There was an increase in sick leave for all Norwegian employees in the health sector in the same period, but this was much smaller: level of sick leave increased from 9·8% to 10·5% (Statistics Norway, 2006). Why the study nursing home had such a large increase in sick leave is difficult to explain. It may have been a coincidence or some of the explanation may lie in organizational factors. The intervention may have contributed by increasing the workload of the control group, but then the increase in sick leave should have been in the control group only. This explanation is not supported in the process evaluation of the project (Sivertsen & Saksvik 2002). Informants reported that the majority of employees, regardless of being in the intervention or the control group, were positive about the project and made an effort to make the work run as smoothly as possible. If the intervention itself was harmful to participants, one would not expect to see an increase in sick leave in the control group.
The intervention group reported statistically significantly fewer neck complaints after the intervention compared to the control group. This is in accordance with the results from a Finnish workplace exercise intervention (Sjogren et al. 2005), and also with the conclusions of previous reviews on physical exercise and neck complaints (Linton & Van Tulder 2001, Kay et al. 2005). Otherwise there was no statistically significant difference between the intervention and control groups regarding SHCs. However, when asked if the project had had any positive or negative effects, there were large statistically significant differences between the intervention and control groups, the intervention group reporting improvement in health, physical fitness, muscle pain, stress management, maintenance of health and work situation. This is in accordance with the findings in an earlier study of employees in the Norwegian postal service (Eriksen et al. 2002). One interpretation of these seemingly conflicting results may be that the participants in the intervention group have learned effective ways of coping with the complaints and accordingly think the problems have improved. Successful conservative treatment of low back pain is aimed at reducing pain-related fear of movement and conveying the message that activity may be painful, but not harmful (Indahl et al. 1995, Hagen et al. 2000). Participating in physical exercise may have helped the intervention group to experience that even if it is painful, activity does not make the complaints worse in the long run, and accordingly they paid less attention to the complaints and were more satisfied because the complaints did not hamper them in their daily activities. This is in accordance with the fear avoidance model (Vlaeyen & Linton 2000).
However, there was no difference in coping, measured as instrumental mastery oriented coping (Eriksen et al. 1997), between the groups. If better coping is one way of explaining incongruence in the lack of effect seen on reports of SHCs and the large subjective effects, it is an aspect of coping not caught by the coping scale we used (Eriksen et al. 1997).
In summary, the intervention did have a positive effect on perceptions of health and well-being of the participants, but not on the actual level of sick leave, complaints or stress.
Study limitations
The groups were small and the number of drop-outs high, implying that large effects are needed to show statistically significant results. There may have been a prophylactic effect, but follow-up time was not long enough to capture this. Post hoc power calculations showed that unrealistically large sample sizes were needed to show an effect, i.e. the effect was so small that it was not clinically interesting. The exceptions were musculoskeletal complaints, gastro-intestinal complaints, job stress, general health, and bodily pain, where the sample size needed to show effect varied from 71 to 151.
Recent research has shown job satisfaction to be an important predictor of health of employees (Faragher et al. 2005). This variable was not measured in our study, but employees in the intervention group reported a statistically significantly better work situation after the intervention period compared with their colleagues in the control group. This may be interpreted as implying an improvement in job satisfaction and thereby supporting the suggestion of a prophylactic effect.
The number of employees who dropped out of the study is a major concern. Drop-out may be selective and in our study the drop-outs were younger and had more sick leave than those who completed the study. It may be that the employees who most need an intervention like this choose not to participate in the study or drop out; the higher level of sick leave in the drop-out group points in this direction. Seasonal effects may be a problem in studies with an intervention period of 9 months. However, this being a randomized controlled study, any seasonal effects should be evenly distributed in the two groups and thus should not have influenced our results.
Given the limitations of the study, there was no support for our hypothesis that the intervention group would have less musculoskeletal complaints and lower level of sick leave than the control group. Our second hypothesis of an increase in coping skills in the intervention group was based on the negative association between SHCs and coping (Eriksen & Ursin 1999) and was not supported, as would be expected because of the lack of reduction of SHCs.
Conclusion
The Integrated Health Programme may be of use to employers who want to increase job satisfaction and well-being among their employees based on the positive subjective effects reported by the intervention group. In view of the large increase in sick leave in both groups during the intervention, this recommendation may be questioned, but we have no cause to believe that the increase in sick leave was connected to the intervention. Nevertheless, considering the positive subjective effects found in this study and the strong data on the association between employee health and job satisfaction, the Integrated Health Programme may have a positive effect on employee health in the longer run. This needs to be assessed in future studies of long enough duration to show possible prophylactic effects of this type of intervention, and cost-benefit analyses should also be included.
Acknowledgements
We would like to thank the employees and managers of the study nursing home for their collaboration. We are grateful to Stein Atle Lie for statistical counselling. We are also grateful to Anette Harris for her good work and enthusiasm, and also would like to thank Aud Skogen for administration of the data collection. We thank Randi Espelid, Nina Konglevoll, Eli Nordeide and Linda Sandal for valuable technical assistance.
Funding
This study was funded by grants from the University of Bergen, Norwegian Research Council, and Ministry of Health and Social Affairs through the Research Unit of the Norwegian Back Pain Network.
Author contributions
THT & HRE were responsible for the study conception and design, performed the data collection and analysis, obtained funding, drafted the manuscript and made critical revisions to the paper for important intellectual content. HRE supervised the study.
References
The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the advancement of evidence-based nursing, midwifery and health care by disseminating high quality research and scholarship of contemporary relevance and with potential to advance knowledge for practice, education, management or policy. JAN publishes research reviews, original research reports and methodological and theoretical papers.
For further information, please visit the journal web-site: http://www.journalofadvancednursing.com
