Despite the incidence of work-related musculoskeletal injuries in paramedics (Rodgers, 1998), little has been published on risk factors for these injuries. This contrasts with the greater numbers of studies undertaken in nurses, including two systematic reviews (Lagerström et al, 1998; Sherehiy et al, 2004), which has resulted in an improved understanding of the important risk factors contributing to musculoskeletal injuries in nursing work.
The work performed and the workplace both contribute to multifactorial risk factors associated with musculoskeletal injuries (Armstrong et al, 1993). Factors related to the performance of work include the physical and psychological requirements of the job, while factors associated with the workplace extend to the manner in which work is organised and the organisational and social context in which work is performed (Armstrong et al, 1993). Furthermore, the relative contribution of risks associated with injuries is likely to vary between industries according to the specific work undertaken and the organisational context (Broniecki et al, 2010).
The performance of work and the organisational and social context in which work occurs presents modifiable risk factors for musculoskeletal injuries. These modifiable risk factors impact on the personal tolerances for mechanical strain so that when personal tolerances are exceeded, mechanical strain occurs (Macdonald and Evans, 2006).
An important risk factor for musculoskeletal symptoms that lead to varying degrees of pain, reduced function and subsequent disability is the cumulative effect of high and low intensity loads over time (Podniece and Taylor, 2008). Specific physical job requirements can increase the risk of work-related back injuries. These job requirements include the rapid pace of work and repetitive motion patterns, heavy lifting and forceful manual exertions, awkward static and dynamic positioning, and frequent bending and twisting (Bevan et al, 2007). Furthermore, these specific risk factors can be compounded by insufficient recovery time (Bevan et al, 2007).
Specific risk factors have also been found to be important in neck pain. Dynamic risk factors include neck flexion, arm force, arm posture, duration of sitting, twisting or bending of the trunk, hand-arm vibration and workplace design, while static risk factors for the neck include sedentary posture and twisting or bending of the trunk (Ariëns et al, 2000).
Psychosocial risk factors for musculoskeletal problems are generally assessed through the perceptions of the worker and carry an emotional value (Devereux et al, 2004). Perceived high work demands, for example, has been found to be a psychosocial risk factor in two systematic reviews that demonstrated an association with musculoskeletal symptoms of the neck and upper limb (Ariëns et al, 2001; Bongers et al, 2006). The psychosocial risk factors of perceived low social support, poor relationships at work, and low job satisfaction have been found to be associated with musculoskeletal symptoms of the back in two systematic reviews (Hoogendoorn et al, 2001; Linton, 2001), while one of these reviews also found that high work demands and monotonous work were also important risk factors (Linton, 2001).
Psychosocial risk factors were found to be important in the development of work-related musculoskeletal symptoms in Swedish ambulance personnel. These risk factors included the psychological demands of the job, lack of social support and worry about work conditions (Aasa et al, 2005). In addition, the psychosocial factors of higher perceived effort of the job, and low job satisfaction were important in the development of back musculoskeletal injuries in a study of Hong Kong emergency medical technicians (Tam and Yeung, 2006).
The objective of this study was to establish whether working conditions, physical and psychological factors over the previous 12 months were associated with aches, pain or discomfort, sustaining injuries or submitting claims, including income maintenance claims.
Methods
Study design
This cross-sectional study was conducted using an on-line survey of paramedics at an Australian ambulance service conducted between April and June 2011. Operational paramedics, that is, those paramedics undertaking direct patient care activities, were invited by email to participate in the study. The email included information about the study, the requirements of participants and a hyperlink to a secure online survey tool (‘Lime Survey’, Germany). The initial email was then followed by two reminder emails at 3-week intervals to maximise the response rate.
Survey tool
We used a composite survey tool designed specifically for the purposes of this study. Most questions were sourced from well-validated questionnaires. Questions on the psychosocial demands of the job were sourced from the UK Health and Safety Executive Indicator Tool (Cousins et al, 2004; Mackay et al, 2004) (35 questions) and the Copenhagen psychosocial questionnaire (Kristensen et al, 2005) (seven questions). Seven questions on the perceived physical demands of the job were sourced from the Australian Safety and Compensation Council survey for nurses (Driscoll, 2008), while three questions on worry about work conditions were sourced from a previous study on Swedish ambulance personnel (Aasa et al, 2005). The researcher developed 13 questions about the ambulance service work context, while seven demographic questions and one question each on barriers to claim submission and increased alcohol consumption were also included. Two questions requiring narrative responses were included to determine factors paramedics felt were causal to sustaining back, neck and shoulder musculoskeletal injuries, as well as possible solutions to mitigate the risk of these injuries.
The questions were developed in consultation with an ambulance service medical director and four paramedics of another ambulance service in which the survey was initially run. This helped ensure the questions and response options, scope of questions, and structure of the survey were appropriate. Consultation also occurred with paramedics from the ambulance service involved in the study reported here to ensure terminology was appropriate to that service.
Outcome variables were based on the Standardised Nordic questionnaire (Kuorinka et al, 1987), and included questions on back, neck and shoulder aches, pain or discomfort, and injuries and claims in the previous year.
A ‘musculoskeletal injury’ was defined in the questionnaire as an impairment of the bodily structures, such as muscles, joints, tendons, ligaments, nerves or the localised blood circulation system, which are caused or aggravated primarily by the performance of work and by the effects of the immediate environment in which work is carried out (Podniece and Taylor, 2008).
A ‘claim’ represented more severe injuries for which income maintenance payments were made when the claimant required time off work due to work injury and/or the payment of work injury-associated medical expenses. Therefore, a claim was defined as the submission of a form for workers compensation by the injured ambulance officer.
Statistical considerations
A power calculation was undertaken based on the Poisson regression mode. Assuming a baseline injury rate of 20 %, and a single normally-distributed predictor variable (Likert scale) with a mean of 3 and a standard deviation of 1, then a sample of 176 achieved 80 % power at the 0.05 significance level to detect a response rate ratio of at least 1.5 due to a one-unit change in the predictor variable.
Data analysis
Log binomial generalised linear model (GLM) regression was used for the dichotomous outcome measures, with the Poisson robust model used when the GLM did not converge. Analysis of count outcome variables was undertaken using a Poisson regression and a thematic content analysis was undertaken for the narrative responses.
The association between injuries and claims, and the frequency of exercise and manual handling training undertaken in the previous year were determined using the Fisher's exact test.
Ethical considerations and approval
Ethics approval for this study was obtained on 9 July 2010 from the ethics review committee of an Australian Health Service. Information regarding confidentiality and privacy was provided to potential respondents in the initial recruitment email, and return of a questionnaire was considered to imply consent. No identifying information was collected, and as work computers were assumed to be used by most respondents, internet provider addresses or log-in locations were not recorded.
Results
A total of 832 survey responses were received. Responses that were 80 % or more complete were included in the analysis. This provided 723 responses available for analysis, a response rate of 21.4 %.
The demographic and work characteristics of the respondents are shown in Table 1.
| Variable | N | % | |
|---|---|---|---|
| Age (years) | ≤30 | 157 | 22 |
| 31–40 | 236 | 33 | |
| 41–50 | 223 | 31 | |
| >50 | 108 | 14 | |
| Total | 723 | 100 | |
| Gender | Male | 465 | 64 |
| Female | 258 | 36 | |
| Total | 100 | ||
| Marital status | Married/de facto | 563 | 78 |
| Never married | 104 | 14 | |
| Other | 56 | 8 | |
| Total | 723 | 100 | |
| Years in operational role | 0–5 | 218 | 30 |
| 6–10 | 180 | 25 | |
| 11–20 | 174 | 24 | |
| >20 | 151 | 21 | |
| Total | 723 | 100 | |
| Full time | Yes | 681 | 94 |
| No | 42 | 6 | |
| Total | 723 | 100 | |
| Location | Metro | 303 | 56 |
| Regional/rural/remote | 320 | 44 | |
| Total | 723 | 100 | |
| Do you mainly work as | Crew member | 652 | 90 |
| First responder | 71 | 10 | |
| Total | 723 | 100 | |
The percentage of survey responses for each age category was broadly similar to the population of paramedics employed.
The prevalence of aches, pain or discomfort, injuries and claims for the 723 survey participants is shown in Table 2. For aches, pain or discomfort, injuries and claims, the back was the most affected area. The main reason for not submitting a claim for those who were injured was that they were not concerned about it.
| Body part | Aches, pain, discomfort | Injury | Claim | ||||||
|---|---|---|---|---|---|---|---|---|---|
| N | % | 95 % Cl | N | % | 95 % Cl | N | % | 95 % Cl | |
| Back | 589 | 81 | 79–84 | 307 | 42 | 39–46 | 93 | 13 | 11–15 |
| Neck | 379 | 52 | 49–56 | 99 | 14 | 11–16 | 18 | 2 | 2–4 |
| Shoulder | 273 | 38 | 34–41 | 178 | 25 | 22–28 | 41 | 6 | 4–8 |
The statistically significant predictors for aches, pain or discomfort, injuries and claims are shown in Tables 3–5. Individuals were more likely to experience back, neck or shoulder musculoskeletal aches, pain or discomfort if required to move or lift heavy loads, were not consulted about workplace changes, felt worn out or increased their alcohol consumption in the previous 12 months.
| Independent variable | Univariate model | Multivariate model | ||||
|---|---|---|---|---|---|---|
| RR | P | 95% CI | RR | P | 95% CI | |
| Move or lift heavy loads | 1.06 | 0.001 | 1.02–1.10 | 1.04 | 0.026 | 1.00–1.08 |
| Staff are not consulted re change | 1.03 | <0.001 | 1.03–1.09 | 1.04 | 0.003 | 1.02–1.08 |
| Feeling worn out | 1.05 | <0.001 | 1.03–1.08 | 1.05 | 0.027 | 1.03–1.08 |
| Increase in alcohol consumption | 1.07 | <0.001 | 1.03–1.12 | 1.04 | 0.003 | 1.00–1.08 |
| Independent variable | Univariate model | Multivariate model | ||||
|---|---|---|---|---|---|---|
| RR | P | 95% CI | RR | P | 95% CI | |
| Move or lift heavy loads | 1.29 | <0.001 | 1.18–1.41 | 1.19 | <0.001 | 1.08–1.30 |
| Unable to take sufficient breaks | 1.21 | <0.001 | 1.13–1.29 | 1.08 | 0.025 | 1.01–1.16 |
| Unrealistic time pressures | 1.20 | <0.001 | 1.13–1.29 | 1.08 | 0.042 | 1.00–1.16 |
| Staff are not consulted re change | 1.19 | <0.001 | 1.00–1.62 | 1.08 | 0.042 | 1.01–1.18 |
| Unable to talk to line manager when upset | 1.08 | 0.004 | 1.03–1.15 | 1.08 | 0.017 | 1.00–1.15 |
| Line manager is not encouraging | 1.34 | <0.001 | 1.02–1.19 | 1.11 | 0.005 | 1.04–1.19 |
| Feeling worn out | 1.19 | <0.001 | 1.12–1.27 | 1.07 | 0.033 | 1.02–1.17 |
| Independent variable | Univariate model | Multivariate model | ||||
|---|---|---|---|---|---|---|
| RR | P | 95% CI | RR | P | 95% CI | |
| Older age | 1.15 | 0.001 | 1.06–1.24 | 1.14 | <0.001 | 1.06–1.24 |
| Gender (female) | 1.38 | 0.043 | 1.01–1.89 | 1.38 | 0.015 | 1.01–1.89 |
| Having a busy last shift rotation | 1.59 | 0.043 | 1.23–2.05 | 1.59 | 0.030 | 1.23–2.05 |
| Repeated and strenuous pushing, pulling | 1.66 | <0.001 | 1.35–2.04 | 1.66 | 0.008 | 1.34–2.04 |
| Worry about patient violence | 1.54 | <0.001 | 1.31–1.79 | 1.54 | 0.000 | 1.31–1.79 |
| Increased alcohol consumption | 1.76 | 0.001 | 1.27–2.44 | 1.76 | 0.001 | 1.20–2.44 |
Paramedics were more likely to sustain an injury if they reported that they were required to move or lift heavy loads, were unable to take sufficient breaks, experienced unrealistic time pressures, were not consulted about change, were unsupported by their line manager or if they felt worn out.
Paramedics were more likely to submit a claim if they were older, female; if their last shift rotation was reported as busy; they perceived that they undertook repeated and strenuous pushing, pulling and awkward positioning; or they felt they were exposed to higher levels of patient violence. There was also a statistical association between an increase in their stated alcohol consumption in the previous 12 months and the likelihood of submitting a claim.
In relation to whether manual handling training in the last 12 months was protective of injuries, there was no statistical association between those with and those without recent manual handling training for either injury rate (54 % vs 56 % P=0.785) or claim rate (18 % vs 17 %, P=0.813).
Paramedics were asked to provide a narrative response to two open-ended questions at the end of the survey. The first question requested information on what paramedics thought were the causal factors for injuries; the second question requested information on interventions paramedics thought would mitigate the risk of injuries.
Manual handling of kit and equipment and vehicle storage issues were the most frequent perceived causal factors for injuries followed by patient extrication difficulties and manual handling involving patients. Other comments were made about the general heavy manual handling requirements of the job including repetitive and awkward movements.
The most frequent suggestion for mitigating the risk of injuries related to kit and lifting equipment. Suggestions included ergonomic interventions such as reducing the weight of kit and equipment and improving access to kit in vehicles, and the use technical interventions such as the use of hydraulics wherever possible, particularly for stretchers.
Discussion
This online survey of over 700 Australian paramedics has identified information that may be useful in formulating organisational strategies aimed at decreasing the incidence of musculoskeletal injuries among operational paramedics.
We found statistically significant associations related to two perceived aspects of physical job demands, moving and lifting heavy loads (objects) aches, pain and discomfort and injuries; and repeated strenuous pushing, pulling and awkward positioning for claims. These findings are consistent with four systematic reviews that demonstrated an association between physical job demands and pain of the neck (Ariëns et al, 2001), combined neck and upper limb (Bongers et al, 2006) and low back (Hoogendoorn et al, 2001).
Furthermore, high levels of, and repeated and strenuous manual handling associated with equipment and patients, were important perceived risks for injuries identified through the narrative responses.
‘The most frequent suggestion for mitigating the risk of injuries related to kit and lifting equipment’
High perceived physical demands of the job was also found to be associated with activity limitation affecting the neck-shoulder area in females and the back in males and females, in a study of Swedish ambulance personnel (Aasa et al, 2005), and low back musculoskeletal injuries in Hong Kong emergency medical technicians (Tam and Yeung, 2006). Furthermore, the high physical demands of the job would appear to contribute to feeling worn out, which was significantly associated with for aches, pains and discomfort, and injuries.
The organisational factor of paramedics feeling that they were not consulted about changes in the work they do was significantly associated with aches, pains and discomfort, and injuries. Changes in the health care system that impact on patient service delivery and ultimately on the work undertaken has been found to be associated with musculoskeletal problems of the back, neck and shoulder in nurses (Lipscomb et al, 2004).
The significant association between injuries and the work organisation risk factors of feeling unable to take sufficient breaks and unrealistic time pressures, and the association between claims and the business of the last shift rotation, may reflect the limited control paramedics have over the allocation of cases, which is directed by the communications centre. Low control over the allocation of work would appear to be compounded by the high work demands. The combined effect of low job control and high work demands have been found to be related to neck and upper limb symptoms in a systematic review (Bongers et al, 2006).
The association between injuries and the reported psychosocial risk factors of being unable to talk to the line manager when upset and the lack of encouragement from the line manager is consistent with two systematic reviews on psychosocial risk factors (Hoogendoorn et al, 2001; Linton, 2001). These reviews demonstrated that the quality of work relationships and low social support and relationships at work, were important psychosocial risk factors for back pain. Similar findings were demonstrated in a study of Swedish paramedics (Aasa et al, 2005) that showed a lack of social support was statistically associated with neck-shoulder complaints.
Worry about patient violence, which was statistically associated with claims, has also been found to be a risk factor amoung Swedish paramedics. Concern about working conditions was found to be associated with low back complaints amoung female paramedics and neck-shoulder and low back complaints in male paramedics (Aasa et al, 2005). However, concern about work conditions in the Swedish study extended beyond patient violence and included concerns about maintaining clinical competence and the risk of contracting an infectious disease. No statistically significant association was found in our study for the explanatory variable of maintaining clinical competence.
Two demographic factors associated with claims were increasing age and being female. Increasing age was found to be a risk factor for low back musculoskeletal injuries in Hong Kong paramedics (Tam and Yeung, 2006) while a systematic review found consistent associations between age and back musculoskeletal disorders (Frank et al, 1996).
The significant association between being female and claims is consistent with a study examining the gender differences for upper limb musculoskeletal disorders taking into account differences in work exposures between men and women (Treaster and Burr, 2004). In both general and working populations women had a higher incidence of various upper limb musculoskeletal disorders after adjusting for confounders such as age and physical work factors. These gender differences were explained by job-related, cultural, and biological factors. The work undertaken by women is generally considered to be less physically demanding with less forceful exertion, although the work is often characterised by high static load of the neck and shoulder region that contributes to musculoskeletal problems. Another factor that may contribute to increased musculoskeletal disorder risks in women may be related to the responsibility of childcare and housework still falling predominantly on women resulting in greater overall exposure as well as reduced time for recovery after the working day (Eriksen et al, 2004).
‘The organisational factor of paramedics feeling that they were not consulted about changes in the work they do was significantly associated with aches, pains and discomfort, and injuries’
Furthermore, differences in strength, muscle mass, tendon and muscle composition, aerobic capacity and hormonal fluctuations may contribute to the increased risk in musculoskeletal disorders in women. Women have been found to have more Type I muscle fibres in the trapezius muscle than men, where myofascial pain is thought to originate (Lindman et al, 1990, 1991). This provides a plausible explanation for the higher level of neck and shoulder musculoskeletal pain experienced by women. In addition, women have a smaller crosssection of trapezius muscle fibres that may further impact on the propensity to develop neck and shoulder musculoskeletal disorders (Treaster and Burr, 2004).
The finding that increased alcohol consumption over the past 12 months was associated with aches, pain or discomfort, and claims, could be explained by paramedics self medicating the symptoms of depression (Grant et al, 2009; Young-Wolff et al, 2009) or drinking to avoid or alleviate negative emotional states such as anxiety (Cooper, 1994). A negative affective state is related to self reported stress and poor coping, health complaints and frequency of unpleasant events (Cooper et al, 1995). Increased alcohol consumption due to the negative affective states of anxiety or depression has been found to be related to a higher frequency of drinking and heavy drinking (Cooper, 1994). In this way, alcohol is used as a way of strategically coping with negative emotions to escape, avoid or otherwise regulate negative emotions (Cooper et al, 1995). Furthermore, negative affective states have in turn been related to increased propensity to injury (Iverson and Erwin, 1997) while individuals with positive affectivity have been shown to demonstrate a more stable disposition with a greater tendency to actively control their environment than those individuals with negative affectivity (Iverson and Erwin, 1997; Ghosh and Bhattacherjee, 2007).
Overwhelmingly, paramedics felt that reducing the weight of equipment (kit, stretchers and carry chairs), reducing the requirement to carry three kits to every case and improving the storage of kit in the ambulance would assist in reducing musculoskeletal injuries. The Australian National Code of Practice for Manual Handling (NOHSC, 2005) states that there is some evidence that the risk of back injury increases significantly with objects above 16–20 kg when lifting, lowering or carrying loads, and that it is advisable to limit manual handling loads to 15 kg. In the service in which the survey was conducted, the approximate weight of kit being carried to every emergency case is likely to be between 29.0–30.9 kg (P Simpson, personal communication, 4 August 2011). This is in excess of the weight recommended in the National Manual Handling Code. Given the repetitive and awkward nature of manual handling activities in ambulance work, the reduction in weight of equipment, better ergonomic design, and reduced frequency of lifts is likely to have a positive impact on reducing injuries and fatigue.
Paramedics who responded to the survey appeared to believe that in order to reduce injuries, the weight of kit should be reduced and technical interventions introduced, for example through the use of hydraulics, particularly for stretchers. They also expressed the need for greater consultation on workplace changes that affect them. Consultation approaches through participative ergonomics (consultation with end users on ergonomic factors) which include workers in workplace change decisions have been found to improve the chances of technical interventions succeeding (Amick III et al, 2006; Podniece and Taylor, 2008; Haslam et al, 2007).
The strong views by some paramedics that fitness will assist in reducing the risk of injuries was not supported through the statistical analysis of the frequency of physical exercise and injury rate or claim rate for the survey sample. However there is evidence from a systematic review that vigorous exercise three times a week is protective of injuries (Podniece and Taylor, 2008). The reasons for not detecting an association between fitness and injury risk might be explained in two ways. First, ambulance work may in itself have a work hardening effect, which may then provide a buffer against injury risk. Alternatively, it could be that those individuals who cannot cope with the heavy physical job demands of ambulance work self select out of the service in the longer term. Self selection out of work by individuals who are unable to meet required physical job demands has been described as the healthy worker effect in a review of the relationship between psychosocial work characteristics and low back pain (Punnett, 1996).
While a proportion of paramedics felt that manual handling training would help mitigate the risk of injuries, an examination of the quantitative data showed no statistically significant association between injury and claim rates for paramedics with recent manual handling training compared with those without recent training.
This finding is consistent with five systematic reviews concluding that technique-based manual handling training as the only intervention strategy may have a limited impact on reducing injuries (Hignett, 2003; Amick III et al, 2006; Haslam et al, 2007; Podniece and Taylor, 2008; Martimo et al, 2009). One of these, a Cochrane review (Martimo et al, 2009), concluded that moderate evidence exists demonstrating that manual handling advice and training is no more effective at preventing back pain or back pain-related disability than no intervention or minor advice. The implications of this would appear to suggest that technique-based manual handling training may be one element of a range of interventions to mitigate the risk of manual handling injuries, and that greater emphasis should be placed on eliminating or reducing the causal factors for musculoskeletal injuries.
Limitations
The limitations of this study relate mainly to the response rate, the generalisability of the study findings and the cross-sectional study design.
While the 21 % response rate of this study of is relatively low, it is similar to a recent survey in another Australian ambulance service. It is possible that a selection bias occurred and that those who responded may have responded differently compared to those who did not respond.
In relation to the generalisability of the findings, the recruitment of paramedics, organisational structures and culture may differ between Australian ambulance services, and may therefore not be applicable to all services.
Finally, this was a cross-sectional study, and associations don't necessarily imply causality. Respondents were asked about pain, injury and claims in the previous 12 months, and it could be that positive responses to risk factors could have occurred after the outcome of interest, with the potential for reverse causality.
‘An examination of the quantitative data showed no statistically significant association between injury and claim rates for paramedics with recent manual handling training compared with those without’
Conclusions
The main implications of this study is that consideration should be given to reducing the physical demands of the job and consequent fatigue of ambulance work, including the weight of the lifting equipment and kit, general manual handling requirements and work organisation factors. Also strategically focused equipment and vehicle research to determine appropriate technical interventions that will reduce the hazard burden on paramedics in the future may be of benefit.
Greater awareness in the organisation on the importance of physical and psychological wellbeing, particularly from line managers may mitigate the risks of injuries. Greater general awareness on the importance of how managers and staff recognise and deal with paramedics who are showing signs of distress, anxiety and depression may also be of assistance in reducing injuries.
Finally, consideration of further research that will objectively measure the activity level (through energy expenditure) of paramedics, and the degree to which ambulance work is aerobic or strength based. This will also help provide information on the type of exercise that should be undertaken and inform dietary requirements.