You are here

Article Text

Article menu

Download PDFPDF

Workplace
Original article
Time trends in the incidence of work-related mental ill-health and musculoskeletal disorders in the UK
  1. Melanie Carder1,
  2. Roseanne McNamee2,
  3. Susan Turner1,
  4. John Timothy Hodgson3,
  5. Fiona Holland2,
  6. Raymond M Agius1
  1. 1Faculty of Medical and Human Sciences, Centre for Occupational and Environmental Health, Institute of Population Health, The University of Manchester, Manchester, UK
  2. 2Health Methodology Research Group, University of Manchester, Manchester, UK
  3. 3Statistics Branch, The Health and Safety Executive, Bootle, UK
  1. Correspondence to Dr Melanie Carder, Faculty of Medical and Human Sciences, Centre for Occupational and Environmental Health, Institute of Population Health, The University of Manchester, Room C4.2, Ellen Wilkinson Building, Oxford Road, Manchester M13 9PL, UK; melanie.carder{at}manchester.ac.uk

Abstract

Objectives To determine UK trends (from 1996 to 2009) in incidence of work-related mental ill-health and musculoskeletal disorders, for all industry as well as for health and social care employees. Second, to investigate whether there may have been a recent shift from a physical to psychological perspective in how patients present their illness by comparing reporting trends for back pain and ‘other work stress’.

Methods Multilevel models were used to investigate changes in incidence of work-related illness, as diagnosed by specialist physicians. The dependent variable comprised case reports to The Health and Occupation Research network. Comparisons were made between medical specialties, industry (health and social care vs all other employees), gender and diagnosis.

Results Trends for Occupational Physicians’ (OP) reporting mental ill-health (average annual increase +3.7% (95% CI +2.2% to +5.2%)) differed significantly (p<0.001) from psychiatrists’ reporting over the same time period (−5.9% (95% CI −7.6% to −4.2%)). For OPs’ reporting, the rate of increase was greater for females and for health and social care employees. A fall in incidence of musculoskeletal disorders for OPs of −5.8% (95% CI −7.3% to −4.3%) and rheumatologists’ reporting −6.6% (95% CI −8.3% to −4.8%) was found, with little variation by gender or industry. Within health and social care, an increase in incidence of ‘other work stress’ was accompanied by a similar decrease in ‘spine/back pain’.

Conclusions The evidence presented is consistent with a shift in the presentation of ill-health from a physical to psychological perspective, although changes in hazards, prevention measures and physician awareness should also be considered as explanations.

https://doi.org/10.1136/oemed-2012-100904

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What this paper adds

    • Musculoskeletal and mental ill-health diagnoses are the two most frequently reported causes of work-related illness in Great Britain.

    • There is evidence that the reported incidence of musculoskeletal disorders may be decreasing and that the incidence of mental ill-health diagnoses may be increasing, with psychological illness frequently cited as the reason for time off work.

    • The incidence of work-related mental ill-health as reported by occupational physicians from 1996 to 2009 increased, while the incidence of work-related musculoskeletal disease decreased over the same time period.

    • Within health and social care employees, the results were consistent with the hypothesis of a shift in how work-related illnesses might be presented/labelled, with an increase in incidence of ‘other work stress’ being accompanied by a decrease in incidence of ‘spine/back pain’.

    Introduction

    Musculoskeletal and mental ill-health diagnoses are the two leading causes of work-related illness in Great Britain, with an estimated 92% of working days lost attributed to these two diagnostic categories.1 There is evidence that the incidence of musculoskeletal disorders (MSDs) such as back pain may be decreasing2 and that the incidence of mental ill-health diagnoses may be increasing, with psychological problems frequently cited as the reason for time off work.3 The Department of Work and Pensions in 2010 estimated that 183 000 people were receiving employment and support allowances for mental ill-health diagnoses, compared with 71 000 receiving support for MSDs.4 One hypothesis is that the changing patterns in incidence of these two diagnostic groups may reflect a change in how patients present their symptoms. The combination of mental ill-health being perceived as less of a stigma and the policy of active treatment for MSDs5 may mean that some cases previously presenting as MSDs might now be more likely to present as psychological. If true, an increase in the incidence of psychological problems might be accompanied by a corresponding, but opposing decrease in the incidence of MSDs, if other factors remain stable.

    One factor that might change over time is type of employment, and the type of workplace hazards encountered may vary across different industries. Therefore, it would be informative (in addition to measuring overall change in incidence of work-related illness) to investigate change within specific sectors. The largest employment sector in the UK is health and social care6 and as such was considered an important workforce for analysis. This sector can also be considered fairly homogenous regarding the type of employer (predominantly public sector) and its underlying hazards. The Government White Paper ‘Choosing Health’ acknowledges that ‘mental illness’ is a problem for National Health Service employees and that ‘stress-related’ conditions and MSDs are currently the most frequently reported causes of work-related sickness absence.7 It is also important to investigate whether the rate of change in incidence of work-related illness is influenced by gender; because while it is generally recognised that psychological disorders such as depression and anxiety are more prevalent in females compared with males,8 less is known about how incidence rates may be changing, and whether there are differences in one gender compared with the other.

    There are two main sources of routinely collected information relating to work-related illness in the UK which can be used to monitor trends in incidence: the Self-reported Work-related Illness (SWI) survey9 and The Health and Occupation Research (THOR) network.10 Reports to the SWI are worker (patient) reported, while reports to THOR are medically diagnosed.

    The aim of this study was to determine the trend in incidence of work-related mental ill-health and MSDs within the UK during the period from 1996 to 2009, for industry overall and specifically for health and social care employees, and to examine whether observed trends varied by gender. It also aimed to investigate whether any observed change in the incidence of psychological illness was accompanied by opposing changes in the incidence of MSDs.

    Methods

    This study used data from three THOR schemes: Musculoskeletal Occupational Surveillance Scheme (MOSS), Surveillance of Occupational Stress and Mental Ill-health (SOSMI) and Occupational Physicians Reporting Activity (OPRA).11–13 MOSS and SOSMI were established in 1999 to enable consultant rheumatologists and psychiatrists throughout the UK to report cases of work-related MSDs and mental ill-health, with data collection ceasing at the end of 2009. OPRA commenced in 1996 (with data collection ongoing) to enable occupational physicians (OPs) to report all diagnoses of work-related illness.

    Participation in THOR is voluntary with physicians asked to report cases of work-related illness seen within their reporting month/s. The decision about whether a case is work-related (caused or aggravated by work) is left to the physician (although guidance is provided).10 If no cases are seen, physicians should return report cards saying ‘I have no new cases to report’. Reports include diagnosis, age, gender, occupation, industry (subsequently coded using Standard Industrial Classification4) and suspected agent. Psychiatrists and rheumatologists are asked to assign their case to one (or more) predefined diagnostic groups.10 OPs provide a diagnosis which, if mental ill-health or an MSD, is then assigned to a category used in the psychiatrists’ or rheumatologists’ reporting scheme.10

    Between 1999 and 2001, all physicians participating in MOSS and SOSMI reported for one randomly allocated month each year (as ‘sample’ reporters). In 2002 a number of these sample reporters commenced reporting every month (as ‘core’ reporters). All OPRA reporters were ‘sample’ until 2004 when a group of 30 were asked to report as ‘core’ (returning to sample in 2005), with another group of 30 ‘sample’ reporters reporting as ‘core’ in 2005.15 At the end of 2005, OPRA ‘core’ reporters from 2004/2005 were invited to continue ‘core’ reporting, and 38 did so. Thus, there was a maximum of data from 12 months per year for each core reporter and from 1 month per year for each sample reporter.

    The methods used to analyse trends in THOR surveillance data have been described previously.16 Briefly, change in incidence of specialist-diagnosed work-related illness among the UK workforce was estimated, with the STATA software command xtnbreg being used to fit longitudinal, negative binomial (ie, overdispersed) models with a β distributed random intercept term. These models are also known as multilevel models with a hierarchical structure, where the repeated reporting occasions over time are nested within individual reporters. The dependent variable was the number of cases, including zeros, per reporter per month. This method is able to correctly allow for variation in the number of reports between reporters (eg, core vs sample and due to non-response). The main predictor of interest, calendar time, was represented either as a categorical variable with 2002 as the base year for comparison, or as a continuous variable with a scale of years. OP analyses were based on data from 1996 to 2009 (full data set) and from 1999 to 2009 (to enable a direct comparison with the available MOSS and SOSMI data). In the categorical analysis, the width of the 95% CIs is affected by the numbers of cases in the comparison and base years. The year with the maximum number of cases (and thus, ideally the base year) varied between groups, therefore the logical compromise was to choose the midyear (2002) as the base year.

    Variables representing ‘season’, ‘reporter type’ (‘core’ or ‘sample’), ‘first month as a new reporter’ and ‘first month as a new ‘core’ reporter’ were also included.16 To account for changes in population base an offset variable representing the UK working population (obtained from the Labour Force Survey) for each year, industrial group (all industry, health and social care or other than health and social care) and (where appropriate) gender was included in the model. Finally, to assess whether there were differences in trends between two groups, a statistical interaction term was added to the model.

    Analyses were undertaken comparing the trend in incidence of total MSDs reported by rheumatologists with total MSDs reported by OPs. Further MSD subgroups were investigated for OPs only, including upper limb disorders (hand, wrist, arm, elbow, shoulder) and spine/back disorders (neck, thoracic spine, lumbar spine, trunk). Similarly, trends in incidence of total mental ill-health as reported by psychiatrists and OPs were compared, with additional analyses of the predefined categories of anxiety and/or depression, and ‘other work stress’ undertaken for OPs only.

    To identify those OPRA reporters who covered health and social care (which it was not possible to determine a priori), the assumption was made that an OP covered this sector if the OP had ever reported one or more case (of any work-related illness) within this sector (ie, cases assigned a SIC92 code of 85.11 to 85.32, inclusive14). The dependent variable was the number of (musculoskeletal or mental ill-health) cases reported in health and social care, per reporter, per month. The comparison group was all other industry (ie, total industry excluding health and social care), with OPs included in this group if they had reported one or more case in all other industry. Based on these criteria, OPs could be selected as reporting cases in either or both of these groups. Analyses were also undertaken to investigate whether the trend in incidence of total MSDs or total mental ill-health varied by gender, for health and social care and for all other industry.

    To investigate whether an increase in incidence of psychological problems has been accompanied by a similar decrease in incidence of MSDs, a comparison of OP reported trends for two frequently reported diagnoses that is, ‘other work stress’ and ‘spine/back pain’ was undertaken. To determine whether any observed changes were of a similar magnitude in terms of number of cases, estimates of absolute incidence were calculated arbitrarily for the year 2002. This comparison was carried out for health and social care only, therefore minimising any bias arising from changes in the types of industry covered by OPs over time.

    Multicentre Research Ethics Committee approval has been granted to THOR (Reference number MREC 02/8/72).

    Results

    Reporters and reporting behaviour

    Between 1996 and 2009, 969 OPs were enrolled in OPRA with 96% responding at least once (ie, returning at least one report card with cases or declaring ‘I have nothing to report this month’). On average, 57 reporters were selected to report to OPRA each month with an overall response rate (number of report cards returned/number sent out) of 83%. MSDs and mental ill-health diagnoses comprised 40% and 41%, respectively, of the total (17 760) cases reported by OPs (table 1). Upper limb disorders and anxiety and/or depression were the most frequently reported musculoskeletal and mental ill-health diagnoses.

    View this table:
    Table 1

    Case reports per month from occupational physicians (OPRA, from 1996 to 2009), rheumatologists (MOSS, from 1999 to 2009) and psychiatrists (SOSMI, from 1999 to 2009)

    In MOSS, 417 rheumatologists were enrolled between 1999 and 2009 with 93% responding at least once. The mean number of reporters approached to report each month was 31, with (on average) 75% of these responding each month. Rheumatologists reported 3128 cases with upper limb disorders and spine/back disorders most frequently reported. OPs and rheumatologists reported approximately one case of work-related MSD per active reporter (ie, excluding physicians who did not respond) per month.

    For SOSMI, 91% of the 1055 psychiatrists enrolled responded at least once (from 1999 to 2009). On average, 58 psychiatrists were asked to report each month with an overall response rate of 72%. Most of the 3639 case reports from psychiatrists were anxiety and/or depression with a smaller proportion of ‘other work stress’. On average, psychiatrists and OPs reported 0.7 and 0.9 mental ill-health cases per active reporter per month, respectively.

    In OPRA, 303 OPs reported at least one case of work-related illness in health and social care between 1996 and 2009 (ie, 32% of OPs who responded during that period). Of these OPs, 86% also reported at least one case from all other industry with only 14% reporting exclusively from health and social care. A further 382 OPs (41% of total OPs responding during this period) reported at least one case in all other industry but did not report a case in health and social care.

    Within health and social care, OPs reported 1779 cases of MSDs (55% of whom had a ‘spine/back’ diagnosis) and 2965 (50% of whom had a diagnosis of ‘other work stress’) mental ill-health cases.

    Time trends by disease category—rheumatologists/psychiatrists versus OPs

    Rheumatologists and OPs reported an overall downward trend in the incidence of total work-related MSDs, although this fall appears to be largely confined to the latter years (figure 1A). The average change in annual incidence (from 1999 to 2009) was estimated at −6.6% (95% CI −8.3% to −4.8%) for rheumatologists and −5.8% (95% CI −7.3% to −4.3%) for OPs. A test for significance showed that the trends were not statistically different (p=0.48). The observed fall in incidence (from 1996 to 2009) of OP reported upper limb disorders was not statistically significantly different (p=0.2) from that seen for spine/back disorders (figure 1B).

    Figure 1
    Figure 1

    Relative rates by year (2002=1), with 95% CIs, of (A) total musculoskeletal disorders as reported by rheumatologists to Musculoskeletal Occupational Surveillance Scheme (MOSS) and Occupational Physicians to Occupational Physicians Reporting Activity (OPRA) and (B) upper limb and spine/back disorders as reported by Occupational Physicians to OPRA.

    OPs’ reports to OPRA suggested an increase in the incidence in work-related mental ill-health, mainly occurring between 1996 and 2004 (figure 2A). Overall (1996 to 2009), the estimated average change in annual incidence was +6.2% (95% CI +4.8% to +7.5%). Restricting the OP data to the same time period as that for psychiatrists (1999 to 2009) saw the estimated change in annual incidence fall from +6.2% to +3.7% (95% CI +2.2% to +5.2%). This was in the opposing direction and significantly different (p<0.001) from the trend observed for psychiatrists over the same time period: average change in annual incidence estimated at −5.9% (95% CI −7.6% to −4.2%).

    Figure 2
    Figure 2

    Relative rates by year (2002=1), with 95% CIs, of (A) total mental ill-health as reported by psychiatrists to Surveillance of Occupational Stress and Mental Ill-health (SOSMI) and Occupational Physicians to Occupational Physicians Reporting Activity (OPRA) and (B) anxiety/depression and other work stress as reported by Occupational Physicians to OPRA.

    The trend in incidence of OPs’ reports of anxiety and/or depression was similar to that for mental ill-health overall (figure 2B). However, the plot for ‘other work stress’ suggests that the reported incidence continues to rise (estimated average annual change +10.1% (95% CI +8.1% to +12.2%)) and that this trend was statistically different (p=0.03) from the trend for anxiety and/or depression.

    Time trends by industrial sector—health and social care versus all other industry as reported by OPs in OPRA

    The trend in reported incidence of MSDs in health and social care (figure 3A) was similar to that observed for all other industry, with a relatively flat trend in the earlier years (up to 2004/2005) followed by a fall in incidence. Overall (1996 to 2009), the average change in annual incidence of MSDs for health and social care (−4.2% (95% CI −6.4% to −2.0%)) was not statistically different (p=0.68) from the average change in annual incidence for all other industry (−4.0% (95% CI −5.4% to −2.6%)).

    Figure 3
    Figure 3

    Relative rates by year (2002=1), with 95% CIs, of (A) total musculoskeletal disorders as reported by Occupational Physicians to Occupational Physicians Reporting Activity (OPRA) in the health and social care sector versus all other industry and (B) total mental ill-health in the health and social care sector versus all other industry.

    The trend in reported incidence for total mental ill-health in health and social care was also similar to all other industry although the ‘peak’ appeared to occur slightly later compared with all other industry (figure 3B). A test for significance suggested that the average increase in annual incidence (1996 to 2009) for health and social care at +6.0% (95% CI +4.0% to +8.1%) was significantly different (p<0.001) to that observed for all other industry at +3.9% (95% CI +2.3% to +5.5%).

    Time trends by gender—health and social care versus all other industry as reported by OPs in OPRA

    Observed trends (1996 to 2009) for MSDs in health and social care did not vary significantly (p=0.32) by case gender although the average decrease in annual incidence for females of −4.8% (95% CI −7.1% to −2.4%) was greater than for males: −3.1% (95% CI −7.1% to +1.1%). For all other industry, a greater average annual decrease in incidence was also observed for females (−5.8% (95% CI −7.9% to −3.7%)) compared with males (−3.7% (95% CI −5.1% to −2.3%)) (p=0.06).

    In contrast, the reported incidence of mental ill-health appeared to be increasing at a significantly greater rate for females compared with males. For health and social care, the estimated average annual change in incidence (1996 to 2009) was +6.5% (95% CI +4.3% to +8.9%) for females compared with +3.2% (95% CI +0.2% to +6.3%) for males (p=0.007) while for all other industry it was +5.5% (95% CI +3.4% to +7.6%) for females compared with +3.3% (95% CI +1.4% to +5.3%) for males (p=0.03).

    Time trends by disease category—‘other work stress’ versus spine/back pain as reported by OPs in OPRA

    The increase in (OP reported) incidence of ‘other work stress’ within health and social care—average annual change of +9.8% (95% CI +6.8% to +12.9%)—has been accompanied by a decrease in the incidence of back pain—average annual change of −7.1% (95% CI −9.8% to −4.3%)—over the same time period (1996 to 2009); these trends were statistically different (p<0.001) (figure 4). The crude estimates of absolute incidence for the year 2002 were similar for spine/back and ‘other work stress’ (47 and 48 per 100 000 persons employed in health and social care, respectively). These two sets of data were combined to estimate the absolute changes in incidence per 100 000 persons, from 2002 to 2007. For ‘other work stress’ this was from 47 to 75 (=47×1.0985) while for back pain it was from 48 to 33 (=48×(1–0.071)5). Thus, on an absolute scale, the fall in back pain incidence is half the increase in ‘other work stress’.

    Figure 4
    Figure 4

    Relative rates by year (2002=1), with 95% CIs, of back pain versus other work related stress as reported by Occupational Physicians to Occupational Physicians Reporting Activity (OPRA) in the health and social care sector.

    Discussion

    This study provides evidence that the incidence of work-related mental ill-health as reported by OPs to OPRA (1996 to 2009) has increased and although the reported incidence of anxiety and/or depression may have peaked, the incidence of ‘other work stress’ might still be increasing. However, data from other reporters of mental ill-health (psychiatrists) suggested a fall in incidence of work-related mental ill-health (1999 to 2009). OPs’ data for a specific sector, health and social care, showed that the incidence of work-related mental ill-health appeared to be increasing at a greater rate compared with all other industry, with a statistically significantly higher rate of change for females compared with males. In contrast, data from rheumatologists and OPs suggested a fall in the incidence of work related MSDs over the study period, with no significant variation (for OP reported trends) by industry (health and social care vs all other industry) or gender. Of interest, within health and social care, the results were consistent with the hypothesis of a shift in how certain work-related illnesses might be labelled, with an increase in incidence of ‘other work stress’ being accompanied by a decrease in incidence of spine/back pain.

    To date, the only other UK-wide investigation of trends in these work-related illnesses is an analysis of data reported to the SWI.17 SWI data for the period from 2001 to 2008 suggest a fairly stable (perhaps slightly downward) trend in the estimated incidence rates of ‘stress, depression and anxiety’ and also in work-related MSDs.

    The observed disparity between OP and psychiatrist reported trends is most likely explained by changes in management of these conditions during the study period, and in corresponding referral patterns. Increasingly, psychological cases are managed within primary care, with hospital referrals to psychiatrists limited to cases where serious pathology is suspected.18 Thus, cases of ‘milder’ psychological illness such as anxiety/depression/stress (which OPs would have continued to see throughout the study period) would have been increasingly less likely to be referred to a psychiatrist; onward referral data from THOR-General Practitioner (GP) indicates that only 1% of the mental ill-health case reports in 2010 were referred to a psychiatrist.19 It is unlikely that the different trends reflect different rates of change in diagnostic labelling or work attribution between OPs and psychiatrists; previous work by this group has shown that when presented with the same information, there was no systematic difference in diagnostic labelling between the two groups of reporters and psychiatrists were in fact more likely to classify a case as work-related compared with OPs.20

    The observed downward trend for psychiatrists’ reporting could also reflect a decline in psychiatrist motivation. Participation in SOSMI was lower compared with other THOR schemes (probably reflecting the comparative difficulty (compared with lung or skin) in attributing a case to work) but this would only impact on the trend estimate if motivation changed over time. Analyses of SOSMI data found some evidence of an increase in non-response and the return of zero cases (‘I have nothing to declare this month’) over time; both of which could be indicative of reporter fatigue. Adjusting for possible fatigue is a major methodological challenge for the investigation of trends in incidence. Various approaches have been adopted previously, but none has been entirely satisfactory.16 ,21 ,22 In view of this, the trend estimates presented here have not been formally adjusted for the potential effect of ‘fatigue’.

    The observed fall in incidence of rheumatologist reported MSDs is also probably due to changes in management and referral patterns with only the more severe cases currently referred to rheumatologists (6% of MSDs reported to THOR-GP in 2010 were referred on to a rheumatologist).19 ,23 Data from OPs also suggested a fall in incidence in MSDs. The similar trends observed for upper limb and spine/back disorders could suggest that the decline reflects reporter fatigue (which would be expected to have a similar influence on both groups) rather than prevention success (which may have different effects on different groups). However, although some measures introduced in the UK to address MSDs have been targeted at specific MSDs, much of the guidance is generic and could therefore feasibly result in a reduction in incidence across several different MSDs and sectors.24 There may have also been a degree of diagnostic inaccuracy arising from difficulties in reliably reporting the site of the musculoskeletal pain (which may be affecting a number of different body regions). MSD subgroup analyses were not carried out for rheumatologists because of the aforementioned concerns about differential referral.

    The higher annual rate of change in incidence of work-related mental ill-health observed for health and social care compared with all other industry may be a result of increased awareness of work-related ill-health by employees within health and social care employment, perhaps also being associated with implementation of guidance from the Health and Safety Executive in a sector where access to occupational health services is relatively good.25 However, this was not reflected in the findings for MSDs, where the results for health and social care did not differ significantly to those observed for all other industry. A limitation of the present study is that the approach adopted to identify OPs covering health and social care (or all other sectors) would not have captured changes in the population served by the OPs over time (which could influence trends). Continued efforts have been made to identify the population covered by OPRA reporters through surveys but difficulties in doing so persist perpetuated by a move away from ‘inhouse’ OPs towards ‘large service providers’ (who may cover a wide range of companies and involve rapid changes in the populations for whom they provide Occupational Health (OH) services). As these issues are unresolved, selecting OPs based on the industries they reported cases from was deemed the best method available.

    This study also observed that, for health and social care and for all other industry, the incidence of mental ill-health (as reported to OPRA) appears to be increasing at a significantly greater rate for females compared with males. If true, this is important, since females comprise approximately 80% of the total health and social care workforce.6 It is well documented that the incidence of mental ill-health is greater for females compared with males,8 however it is less clear why the incidence of work-related mental ill-health is increasing at a greater rate for females. One possible explanation could be the increasing pressure on females to balance work and home responsibilities.26 In comparison, a greater decrease in annual incidence of MSDs was observed for females compared with males in health and social care and all other industry, but these differences were not statistically significant.

    Within health and social care, measures of relative increase in incidence of ‘other work stress’ (as reported by OPs) was accompanied by a similar (in terms of magnitude) relative decrease in incidence of back pain, but our calculations suggest that the absolute increase in the former is likely much greater than the absolute fall in the latter. Nevertheless, these findings are consistent with the hypothesis that there may have been a shift over recent years in how some patients present their symptoms (ie, from a physical to a psychological perspective). However, the observed shift in case reporting could reflect other factors such as a shift in the types of hazards faced by workers over the study period. A recent review of 11 studies investigating interventions to reduce back pain found no evidence that measures such as advice, training and lifting equipment were effective, suggesting that the observed shift in case reporting is unlikely to be because prevention programmes focusing on MSDs have been more effective than those focusing on mental ill-health.27 Changes in physicians’ and patients’ awareness of these two major diagnostic groups may also have played a role; an increased focus on mental ill-health in recent years may have led to an increased likelihood of patients seeking help for these conditions and an increase in the awareness of physicians to accurately pursue possible physiological symptoms (and hence make psychological diagnoses).

    For MOSS and SOSMI, including data from the Labour Force Survey (LFS) in the multilevel models to represent the UK working population was considered appropriate, as according to the principles underpinning the National Health Service, all of the UK workforce has access to a clinical specialist. In turn, since all eligible rheumatologists and psychiatrists had (in theory) an equal chance to participate in THOR (and the proportion actively doing so has remained relatively constant throughout the study period), there was no reason to believe that cases reported to these two schemes are unduly biased. It therefore seems reasonable to assume that the trends observed here are representative of UK rheumatologists and psychiatrists in general.

    All UK specialist OPs are equally eligible to report to OPRA with estimates suggesting that at least 50% of OPs were enrolled in OPRA during the study period. However, OPs differ from the other clinical specialists in that access to an OP among the UK workforce is biased towards the public sector and larger employers; earlier work (based on 503 OPs who reported to OPRA from 1998 to 2000) suggested that approximately 12% of the UK working population had access to an OP.25 However, this study involved measuring relative incidence (not absolute incidence) so it was deemed reasonable to include data from the LFS to represent population changes for OPs. This approach makes the assumption that the relationship between the size of the population with access to an OPRA reporter and the size of the total working population (LFS) has remained fairly stable over the study period. This assumption is probably more valid for industry as a whole, and for a large public sector such as health and social care, compared with other, smaller sectors. To enable a direct comparison of the magnitude of change in the incidence of ‘other work stress’ and ‘spine/back’ pain (in health and social care) crude estimates of absolute incidence for the year 2002 were made. The denominator used to calculate these rates was obtained from an OPRA denominator survey. Although there are a number of unresolved issues with these denominator data (and therefore actual incidence rates may be higher or lower than reported here), it was the relative magnitude (or not) of the incidence rates rather than the absolute figures that was of interest in this study.

    It has been shown that different THOR data sources (rheumatologists/psychiatrists vs OPs) provide sometimes complementary and sometimes different perspectives on the trend in incidence of work-related illness in the UK. Differences between the data which might at first appear to suggest inconsistency, can however be explained plausibly by superimposed factors (such as differential access to physicians). The differences do however serve to underline the value of multiple data sources. Data from GP sources (not yet published in this analytical format) are likely to further strengthen the validity and value of this output.28 The longitudinal analysis of temporal trends that we have undertaken are important in terms of identifying those diseases for which the incidence is increasing (or decreasing) and in helping to identify groups who may be at an increased risk (therefore aiding efficient targeting of resources to reduce incidence). Furthermore, work has now also begun to investigate the impact of government interventions and initiatives on the incidence of work-related illness.29

    Acknowledgments

    We are grateful to the physicians who report to THOR for their continuing support. Physicians who wish to join THOR and participate in reporting can find further details at: http://www.medicine.manchester.ac.uk/oeh/research/thor/

    References

    1. ↵
      Health and Safety Executive. Table THORGP01. Top level data from THOR-GP for 2007 to 2009—Number of diagnoses, calculated incidence rates and associated sickness absence by broad diagnostic category of work-related ill-health. http://www.hse.gov.uk/statistics/tables/index.htm#thor (accessed Apr 2012).
    2. ↵
      1. Waddell G,
      2. Aylward M,
      3. Sawney P
      . Back pain, incapacity for work and social security benefits: an international literature review and analysis. London: Royal Society of Medicine Press, 2002.
    3. ↵
      1. Hussey L,
      2. Turner S,
      3. Thorley K,
      4. et al
      . Work-related sickness absence as reported by UK general practitioners. Occup Med (Lond) 2012;62:105–11.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      Department of Work and Pensions. Employment and Support Allowance—Caseload by medical condition. Caseload: February 2010. http://research.dwp.gov.uk/asd/index.php?page=esa (accessed Apr 2012).
    5. ↵
      British Occupational Health Research Foundation (BOHRF). Back pain at work: a guide for people at work and their employers. 2000. http://www.bohrf.org.uk/downloads/ (accessed Apr 2012).
    6. ↵
      Office for National Statistics. Labour Force Survey 1996–2009: The Stationery Office, 2008.
    7. ↵
      Department of Health. Choosing health: making healthier choices easier. London: Department of Health, 2004. http://webarchive.nationalarchives.gov.uk/20100407034821/http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4094550 (accessed Apr 2012).
    8. ↵
      World Health Organisation. Gender and women's mental health: gender disparities and mental health: the facts. http://www.who.int/mental_health/prevention/genderwomen/en/ (accessed Apr 2012).
    9. ↵
      Health and Safety Executive. Self-reported work-related illness (SWI) and workplace injuries (LFS). http://www.hse.gov.uk/statistics/publications/swi.htm (accessed Apr 2012).
    10. ↵
      THOR—The Health and Occupation Research network. http://www.medicine.manchester.ac.uk/oeh/research/thor/ (accessed Apr 2012).
    11. ↵
      1. Cherry NM,
      2. Meyer JD,
      3. Chen Y,
      4. et al
      . The reported incidence of work-related musculoskeletal disease in the UK: MOSS 1997–1999. Occup Med 2001;51:450–5.
      OpenUrlAbstract
    12. ↵
      1. Cherry NM,
      2. Chen Y,
      3. McDonald JC
      . Reported incidence and precipitating factors of work-related stress and mental ill-health in the United Kingdom (1996–2001). Occup Med 2006;56:414–21.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Cherry NM,
      2. McDonald JC
      . The incidence of work-related disease reported by occupational physicians, 1996–2001. Occup Med 2002;52:407–11.
      OpenUrlAbstract
    14. ↵
      Central Statistical Office. Indexes to the Standard Industrial Classification of Economic Activities 1992. London: HMSO, 1993:15.
    15. ↵
      1. McNamee R,
      2. Chen Y,
      3. Hussey L,
      4. et al
      . Randomised Controlled Trial comparing time-sampled versus continuous time reporting for measuring incidence. Epidemiology 2010;21:376–8.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. McNamee R,
      2. Carder M,
      3. Chen Y,
      4. et al
      . Measurement of trends in incidence of work-related skin and respiratory diseases, UK 1996–2005. Occup Environ Med 2008;65:808–14.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      Health and Safety Executive. Analysis of the trends in self-reported work-related ill-health with special reference to the trend reversal in 2006/07. HSE Statistics Branch. 2008. http://www.hse.gov.uk/statistics/pdf/lfsissue1.pdf (accessed Apr 2012).
    18. ↵
      NICE Guidelines CG123: Common mental disorders: Identification and pathways to care. http://publications.nice.org.uk/common-mental-health-disorders-cg123 (accessed Apr 2012).
    19. ↵
      The Health & Safety Executive. Table THOR-GP02. Ill-health: number of diagnoses, incidence rates, sickness absence and secondary referral by diagnostic category, latest year only http://www.hse.gov.uk/statistics/tables/index.htm#thor (accessed Apr 2012).
    20. ↵
      1. O'Neill E,
      2. Turner S,
      3. McNamee R,
      4. et al
      . Do psychiatrists and occupational physicians have different diagnostic preferences for cases of work-related psychological ill-health? Occup Environ Med 2008;65:726–31.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. McNamee R
      . Fatigue discussion. Report to HSE, 12th March 2007.
    22. ↵
      1. McNamee R
      . Surveillance and reporter fatigue. Report to HSE: 8th May 2007.
    23. ↵
      National collaborating centre for primary care. Low back pain: early management of persistent non-specific low back pain. Full guideline, May 2009. http://guidance.nice.org.uk/CG88/Guidance/pdf/English (accessed Apr 2012).
    24. ↵
      Health and Safety Executive. Legislation covering MSDs http://www.hse.gov.uk/msd/legislation.htm (accessed Apr 2012).
    25. ↵
      1. McDonald JC
      . The estimated workforce served by occupational physicians in the UK. Occup Med 2002;52:401–6.
      OpenUrlAbstract
    26. ↵
      1. Lundberg U,
      2. Frankenhaeuser M
      . Stress and workload of men and women in high ranking position. J Occup Health Psychol 1999;4:142–51.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Martimo KP,
      2. Verbeek J,
      3. Karppinen J,
      4. et al
      . Effect of training and lifting equipment for preventing back pain in lifting and handling: systematic review. BMJ 2008;336:429–31.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Hussey L,
      2. Turner S,
      3. Thorley K,
      4. et al
      . Work-related ill health in general practice, as reported to a UK-wide surveillance scheme. Br J Gen Pract 2008;58:637–40.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Stocks SJ,
      2. McNamee R,
      3. Turner S,
      4. et al
      . Has European Union legislation to reduce exposure to chromate in cement been effective in reducing the incidence of allergic contact dermatitis attributed to chromate in the UK? Occup Environ Med 2012;69:150–2.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Contributors RM, MC and RA designed the study. MC and RM analysed the data. All authors contributed to interpretation of data. MC drafted the article and all authors critically reviewed and provided further input to the article. All authors have approved the version submitted for publication.

    • Funding THOR is partially funded by the Health and Safety Executive and has also received funding from the Department of Health, and from charitable sources. This paper expresses the views of the authors, and not necessarily of the funders.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    Linked Articles