Job Strain Exposures vs.

Stress-Related Workers' Compensation Claims in Victoria, Australia: Developing a Public Health Response to Job Stress Author(s): Tessa Keegel, Aleck Ostry and Anthony D. LaMontagne Source: Journal of Public Health Policy, Vol. 30, No. 1 (Apr., 2009), pp. 17-39 Published by: Palgrave Macmillan Journals Stable URL: http://www.jstor.org/stable/40207220 . Accessed: 10/09/2013 05:08
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Journal of Public Health Policy.

http://www.jstor.org

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Original Article

Job strain exposures vs. stress-related workers' compensation claims in Victoria, Australia: Developing a public health response to job stress
Tessa Keegel3, Aleck Ostryb and Anthony D. LaMontagnea'*
aSchool of Population Health, McCaughey Centre: Vichealth Centre for the Promotion of Mental Health and Community Wellbeing, University of Melbourne, Parkville, VIC 3010, Australia. E-mail: alamonta@unimelb.edu.au bDepartment of Geography, University of Victoria, Victoria, BC, Canada. * author. Corresponding

We presenta comparativeanalysis of patternsof exposure to job Abstract stressorsand stress-relatedworkers' compensation (WC) claims to provide an evaluation of the adequacy of claims-drivenpolicy and practice. We assessed job strainprevalencein a 2003 population-basedsurveyof Victorian[Australia] workers and compared these results with stress-relatedWC statistics for the same year. Job strain prevalence was higher among females than males, and elevatedamong lower vs. higheroccupationalskill levels. In comparison,claims were higher among females than males, but primarilyamong higher skill-level workers. There was some congruence between exposure and WC claims patterns. Highly exposed groups in lower socio-economic positions were underrepresented in claims statistics, suggesting that the WC insurance perspective substantially underestimates the job stress problems for these groups. Thus to provide a sufficient evidence base for equitable policy and practice responses to this growing public health problem, exposure or health outcome data are needed as an essential complement to claims statistics. Journal of Public Health Policy (2009) 30, 17-39. doi:io.iO57/jphp.2Oo8.4i Keywords:job strain; job stress; policy; workers' compensation

Background
Population-levelOccupationalHealth and Safety (OH&S) policy and practice priorities can be determined from two main data
2.009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 www.palgrave-journals.com/jphp/

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

^j

Keegelet al

sources: exposure or health outcome patterns (OH&S data), and workers' compensation (WC) claims patterns (insurance data). The WC insurance (illness-focused) should operate as a backup to primary (exposure-focused) and secondary (worker-focused)preventive efforts, compensating and rehabilitating those whom the OH&S regulatory system has failed to protect.1 If the WC system effectively fulfils this function, the occupational groups experiencing higher exposures should also show higher claims rates. Lack of congruence between exposure/illness and claims data would suggest problems in the OH&S system deserving further investigation or correction. Most comparative studies have focused on injuries and injury claims. Historically, studies have shown widely varying degrees of coverage, from as low as roughly 25 per cent of occupational injuries registered in hospital emergency departmentsfor a poor, predominantlyAfricanAmerican, inner-city,Philadelphia population2 to as high as of 90 per cent coverage for serious injuries requiring hospitalization in a cohort of highly unionized Canadian sawmill workers.3 Studies of most work-related chronic disease outcomes, however, are complicated by long latency periods, lack of recognition of relatedness to work, multiple contributing causes (both work and non-work-related)and other factors. Where exposure-disease relationships are established, such as for work-related contributions to common chronic diseases like depression and cardiovascular disease (CVD), population-level exposure patterns offer a way to guide and evaluate policy and practice.4 Internationally, job stress is a substantial and growing concern for working people, their advocates, employers, occupational health and safety regulators, and WC insurance schemes. Job stress has been linked to a range of adverse physical and mental health outcomes, including CVD,5"8 musculoskeletal disorders,9 depression and 710~17 anxiety. Figure 1 presents an overview of the job stress process. Poor working conditions, or job stressors, may lead to worker distress, which in time may result in enduring adverse health outcomes. In theory, workers experiencing job stress-related health outcomes should be compensated, rehabilitated, and returned to work by the WC system. Primary intervention aims to eliminate or reduce job stress by targeting working conditions. Secondary intervention attempts to alter the ways that individuals perceive or respond to job stressors. Tertiary intervention treats, compensates, and rehabilitates workers who have developed job stress-related
18 2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

/f?

Intervention Primary To eliminate or reducejob stressors

Secondary Intervention To alterthe ways thatindividuals perceiveor respondto job stressors

Tar*,~n, in*An.Ar.i^n To treatcSnsSHnd rehabilitate J^' ^Sb^^Stl^m^

I
Distress USIBM -

Working a Conditions

Modifying Variables : Individual or Situational Social -Non-work related stressors -Socioeconomic status Biophysical -Age -Sex -Health status Psychological -Personality -Coping abilities

t1-

n1-

Short Term Responses

T1

Enduringhealth outcomes
-

Characteristics Genetic -Inheritedpredisposition to mental illness, heart disease

Behavioural -Exercise -Recreational activities -Nutrition

Figure i: The job stress process, modifying variables and intervention points. Source: From LaMontagne et al.

In additionto being a commonlyreportedcause of occuillnesses.18 illness, job stress has been linked to other unfavourable pational such as lost workdays, low productivity, outcomes organizational and high turnoverrates.19For some leading chronic diseases, the proportionsattributableto job stress are substantial.An estimated to job strain- the 7-16 per cent of CVD among men is attributable combinationof high job demandsand low job control.7In a recent attributable risk Australian study,we reportedjob strain-population (PAR)for depressionat 13 per cent for males and 17 per cent for females.20 Depression, anxiety, and other common mental disorders are The most recently widely prevalentamong working Australians.21 available Australian National Survey of Health and Wellbeing (1997) found the prevalenceof 12-month depressionat 3.89 per cent for employed men and 8.41 per cent for employed women. Depression was defined using DSM-IV criteria and a modified DiagnosticInterview(CIDI)method.20An CompositeInternational set data of the by Andrewsand co-authorsfoundonly 3 5 per analysis cent of peoplewith any mentaldisorderreporteda consultationfor a mentalhealth problemduringthe past year.22 Although occupational stress is a significant public health problem, in most jurisdictions population-level information is lacking regardingthe patterns of exposure and associated health outcomes.In Australia,and many other industrialized democracies,
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 19

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Keegel et al

job stress policy and practice responses are driven primarily by stress-related WC claims statistics. In Australia, claims for stressrelated effects on mental health are identified by a 'mechanism of injury or disease classification' of 'mental stress.' According to the National Occupational Health and Safety Commission (NOHSC, now the Australian Safety and Compensation Council), the incidence rate for such WC cases was 0.9 per 1000 (7480 cases) in 2003. 23 For the state of Victoria, the number of successful WC stress claims increased from 2417 cases in 2001-2002 to 2922 cases in 2003-2004, despite an overall downward trend for WC claims of all types.24 Stress-related claims are more costly and complex to manage than other injury claims. The Victorian Workcover Authority has estimated that double the compensation is paid to workers suffering from stress compared to physical injuries - in 2004/2005, $AUS 133.9 million.24 As overall WC claims numbers are decreasing over time, claims for mental stress increased by 83 per cent from 19961997 to 2003-2004. The Compendium of Workers' Compensation Statistics Australia 2004-2005 examined the most common subcategories of 'mental stress' claims. Work pressure accounted for 41 per cent of all mental stress claims, followed by harassment (22 per cent), workplace or occupational violence (16 per cent), and other mental stress factors (16 per cent).25 WC claims result when workers seek compensation for conditions identified by medical practitioners as having an occupational cause. When a worker presents to a medical practitioner for a job stressrelated condition - whether the worker suspects stress-relatedness or not - the practitioner may or may not identify an underlying occupational cause or contribution. Even if job stress is medically recognized as a contributory factor, Australian general practitioners have been found to be reluctant to initiate WC claims for patients presenting with job stress-related conditions.26'27These barriers or filters operating in the WC claims process, combined with the under-diagnosis of common mental disorders from all causes - workrelated and other - suggest that only the most severe and persistent cases of job stress-related illness result in compensated claims. We believe other methods of assessing the magnitude of job stress-related mental illness are needed. Work contexts demonstrating elevated stress-related claims education and health-care sectors, for example - clearly warrant
2.0 2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

*^*

preventive intervention. Yet elevated exposure prevalence and associated adverse health effects may not always be manifest in elevatedclaimsrates.The true numberof individualsaffectedby job stress in Australiaand the associatedeconomic and personalcosts may be far higherthan claims statisticssuggest. Many workersare for example not coveredby WC systems- self-employed individuals, - and some groupsof workers,such as those precariously employed, tend not to file claims for work-related illness.29 Occupational disease, particularlyfor multi-factorialdisease outcomes such as those associatedwith job stress, is not recognizedor reportedfully. for occupationaldiseasesin Leighand Robbinslooked at WC claims 30 the United Statesfor the year 1999. Among other disease claims, they reporteda total of 2272 claimsfor 'mentalstress'and concluded that in general WC statistics substantiallyunderestimateoccupational disease. They did not publish the denominators.In a recent study,we comparedthe numberof cases of job strain-attributable depression(21437) to the numberof 'mental stress'compensation claims (696), in Victoria, and found that claims statistics underestimatejob strain-attributable depressionby at least 30-fold.20 This paperpresentsa descriptivecomparativeanalysisof population-level patterns of exposure to job stressorswith stress-related WC claims. Building on our previous finding that job stressmentalillnessis far more prevalentthan claimsstatistics attributable would suggest, in this analysis we set out to identify exposed over occupationalgroupsin need of policy and practiceintervention and above those identifiedby elevatedclaims rates. Our goal is to evaluationof the adequacyand equity provide(1) a population-level of claims-driven policy and practice,and (2) informationfor policy and practice action that complementsand extends a claims-based perspective.

Methods
Data sources The Victorian Job StressSurvey(VJSS) The VJSS was conducted using a cross-sectional representative sample of 1101 workers (526 men and 575 women) in the southeasternAustraliastate of Victoria. Furtherdetails on the sampling
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 21

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

^j

Keegelet al

frame and data collection procedures are described elsewhere.31"33 This study was reviewedand approvedby the Universityof Melbourne's Human Research Ethics Committee (HREC protocol #030398). Job stress measures We selected Karasek and Theorell's demand and control model, in particular their measure of job strain, as the job stressor, or exposure measure of choice because it is the most widely studied measure, integrates job control and job demands, and demonstrates strong associations linking it predictively to adverse effects on mental and physical health.7'34'35 The model focuses on task-level job characteristics, postulating that psychological strain results from the interaction of job demands and job control, with the combination of low control and high demands producing 'job strain'. We used standard methods, as described previously, to compute these measures, with job control and psychological demand dimensions meeting international norms of reliability (Cronbach a's of 0.80 and 0.66, respectively).32'33 Covariates Covariate data were collected for a range of sociodemographics. Workers were asked if they were a member of a union, and whether they worked for a government, private sector, or not-for profit, religious, or community organization and their average weekly working hours. Occupations were collapsed into five Australian Bureau of Statistics (ABS) skill levels - level one lowest to level five highest. Occupational skill level was dummy coded with the highest skill level serving as the reference category. Industrial sector information was collected according to 17 ABS categories and then collapsed into manufacturing or service. Hostility was assessed using the sum of three Likert-scaled items36 with higher scores indicating greater hostility. Victorian WC data The Australian National Occupational Health and Safety Commission (NOHSC) compiles a publicly accessible national WC statistics database.23 Numbers of cases are derived from compensation claims received from insurance companies, self-insurers and government departments at commonwealth, state, and territory level. The denominators which are used by NOHSC were calculated by the
22 2.009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

a publichealthresponse to job stress Developing

ABS using Labour Force Survey and the Survey of Employee Earnings and Hours.23The WC database was queried for incidence rates of job stress claims in Victoria for the same year as the VJSS (2003), as identified by mechanism of injury or disease classification of 'mental stress'. Claims incidence rates were filtered by age, gender, ABS classifications for occupational levels, and ABS categories for industry. Statistical analyses Job strain exposure data from the VJSS were stratified by industry, age, and occupation with proportions calculated by group. We analysed males and females separately. Bivariate analyses compared categorical variables using a * test, or a Fisher's exact test when appropriate. We performed four sets of multivariate logistic regression analyses to identify determinants of job strain, with risk expressed by odds ratios (OR) and 95 per cent confidence intervals. We assessed model fit using Hosmer-Lemeshow tests; all models presented had acceptable test statistics (>o.2o). For the WC data from Victoria, incidence rates and numbers of cases with 'mental stress' claims were stratified by age, and occupation. For both the VJSS data set and the WC claims data set, we noted where proportions or rates were higher or lower than the overall WC incidence rates (number of occupational disease cases/number of employees x iooo).23 We analysed data using STATA 8 (Stata Corporation, College Station, TX).

Results
Socio-demographic and employment characteristics for the VJSS are summarized in Table 1. Working males were older and had a lower educational level than working women. More males were selfemployed, and more females were employed in their main job for ^35 hours/week. Most respondents were employed by private companies or not-for-profit agencies and less than a third of workers belonged to a union. Job strain The prevalence of job strain was higher in females than in males (25.4 per cent vs. 18.6 per cent, P<o.O5). Younger males had the
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 2.3

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

^z

Keegelet al

Table i: Victorian Job Stress Survey: Socio-demographic and employment characteristics Males n (%) Whole sample Age ^51 years 41-50 years 30-40 years <3O years Educational level Post-graduate Undergraduate Vocational Completed high school Completed primary or some high school Occupation Level five (highest skill level) Level four Level three Level two Level one (lowest skill level) Union membership Industrial sector Manufacturing Service Location Urban Rural/regional Employed by Government Private/not-for-profit agency Self-employed Size of workplace ^20 <2O Average weekly hours (ABS) ^35 hours 36-49 hours ^50 hours ^=526 Females n (%) n=575 Total N (%) N=iioi

122(23.2) 122 (23.2) 161 (30.6) 121 (23.0)

117(20.3) 162 (28.2) 159 (27.7) 137 (23.8)

239(21.7) 284 (25.8) 320 (29.1) 258 (23.4)

47 (8.9) 132 (25.1) 128 (24.3) 90(17.1) 124(23.5)

56 (9.7) 217 (37.7) 76 (13.2) 111(19.3) 112(19.5)

103 (9.4) 349 (31.7) 204 (18.5) 201(18.2) 236(21.4)

115 41 130 119 121

(21.9) (7.8) (24.7) (22.6) (23.0)

164 47 80 142 142

(28.5) (8.2) (13.9) (24.7) (24.7)

279 88 210 261 263

(25.3) (8.0) (19. 1) (23.7) (23.9)

148(28.1)

165(28.7)

313(28.4)

339(64.5) 187(35-5)

235(40.9) 339 (59-o)

574(52.1) 526(47.8)

377 (71.7) 149 (28.3) 61 (11.6) 462 (87.8) 121 (23.0)

417 (72.5) 158 (27.5)

794 (72.1) 307 (27.9)

175 (30.4) 392 (68.2) 66 (11. 5)

236 (21.4) 854 (77.6) 187 (17.0)

273(49.2) 253(46.0) 106 (20.1) 250 (47.5) 160(30.4)

306(53.2) 269(46.8)

579(52.6) 522(47.4)

296 (51.5) 198 (34.4) 65(11.3)

402 (36.5) 448 (40.7) 225(20.4)

24

Macmillan 2009 Palgrave 0197-5897 Journalof PublicHealthPolicy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

sfc

highest prevalence of high job strain. They were also most likely to be in passive jobs (Table 2, Males). Older males had the lowest prevalence of job strain. We found significant differences according to occupational skill level, with the prevalence of high job strain and passive jobs increasing stepwise with decreasing skill level. Being self-employed was highly protective against job strain. Male unionized workers had a demand-control profile similar to non-union members. For females (Table 2, Females), job strain prevalence was highest among middle-aged women vs. those aged ^51. Similar to the pattern for males, the prevalence of high job strain and passive jobs was highest in the lowest skill group, but with less of a clear gradient. Self-employment among females was also highly protective against job strain. Among female union members vs. non-members, however, there was a higher prevalence of job strain in combination with a markedly higher prevalence of active jobs and a lower prevalence of passive jobs. As observed in males, females with higher skill-level jobs generally had lower job strain prevalence and those with lower skill-level jobs had higher prevalence of passive jobs. Logistic regression modelling of job strain In bivariate analyses for males (first column, Table 3, Males), before adjustment for age, being an employee (vs. self-employed), and working longer hours, were associated with higher odds of experiencing job strain. Multivariate modelling was then conducted to assess the relative contributions of potential determinants of job strain. Several covariates were not significant and dropped from Models A and B (Table 3, Males). Because negative personality - hostility may represent both a predisposition to and a consequence of job strain, to be conservative, we present models with and without adjustment for hostility (Models A and B, Males). Hostility is significantly associated with job strain, but with a very small magnitude in comparison to other identified determinants. The final models for males (Models C and D, Males) show that the risk of job strain is elevated among young males, males in lower skill-level jobs and males working longer hours. Effect size estimates for these job strain determinants remained fairly stable with varying combinations of covariates modelled, and were little affected by adjustment for
Journal of Public Health Policy Vol. 30, 1, 17-39

2009 Palgrave Macmillan 0197-5897

25

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

*^>* Keegelet al

Table 2: Victorian Job Stress Survey: Four-way demand control measures Males (n-^01) Passive Low job Active High job ? -value strain strain jobs jobs n (row %) n (row %) n (row %) n (row %) 15(13.6) 9 (23.7) 28 (22.2) 49(44.9) 55 (46.6) 47(39-5) 40 (25.8) 30 (26.6) 39(34.2) 21 (19.1) 135 (34.5) 109 (30.3) 46 (32.9) 13(11.8) 5 (13.2) 22 (17.5) 24(22.0) 29 (25.6) 28(23.5) 27 (17.4) 24 (21.2) 14(12.2) 11 (10.0) 82 (21.0) 65 (18. 1) 28 (20.0) 0.000 0.044 0.000 0.840 0.099 0.364 0.000

Occupation Level five (highest skill level) Level four Level three Level two Level one (lowest skill level) Age <3O years 30-40 years 41-50 years ^51 years Self employed or employee Self-employed Employee Union membership Non-union member Unionized Industrial Sector Manufacturing Service Employed by Government Private/not-for-profit Average weekly hours (ABS) ^35 hours 36-49 hours ^50 hours Females (n=j$o) Occupation Level five (highest skill level) Level four Level three Level two Level one (lowest skill level) Age <3O years 30-40 years

41(37.3) 13 (34.2) 45 (35.7) 19(17.4) 22 (18.6) 22(18.5) 52 (33.6) 30 (26.6) 36(31.6) 49 (44.6) 91 (23.3) 103 (28.6) 37 (26.4)

41(37.3) 11 (28.9) 31 (24.6) 17(15.6) 12 (10.2) 22(18.5) 36 (23.2) 29 (25.7) 25(21.9) 29 (26.4) 83 (21.2) 83 (23.1) 29 (20.7)

101 (31.0) 39 (22.4)

74 (22.6) 38 (21.8)

91 (27.8) 65 (37.4)

61 (18.7) 32 (18.4)

12(20.0) 126 (28.8)

18(30.0) 94 (21.5)

19(31.7) 136 (31.1)

11(18.3) 82 (18.7)

26(25.2) 67(27.9) 44(29.7)

10(9.7) 39(16.2) 61(41.2)

52.(50-5) 78(32.-5) 22(14.9)

i5(i4-6) 56(23.3) 21(14.2)

36 (22.6) 9 (20.0) 15(20.0) 33 (24.4) 10 (7.4) 22(16.5) 29 (18.7)

75 (47.2) 6 (13.3) 15(20.0) 19 (14.1) 8 (5.9) 18(13.5) 35 (22.6)

19 (11.9) 16 (35.6) 25(33.3) 52 (38.5) 72 (52.9) 57(42.9) 46 (29.7)

29 (18.2) 14 (31. 1) 20(26.7) 31 (23.0) 46 (33.8) 36(27.1) 45 (29.0)

0.000 -

2.6

Macmillan of PublicHealthPolicy Vol. 30, 1, 17-39 2009 Palgrave 0197-5897 Journal

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

^-

Table 2: Continued

Females(n-^o)

Low job Active Passive High job ? -value strain strain jobs jobs n (row %) n (row %) n (row %) n (row %) 27(17.8) 25 (22.7) 21 (35.6) 82 (16.7) 83(21.3) 20 (12.5) 52(22.8) 51 (16.0) 35 (20.6) 67 (17.9) 56(19.6) 33 (17.7) 10(16.1) 38(25.0) 32 (29.1) 46(30.3) 35 (31.8) 41(27.0) 18 (16.4) 0.035

41-50 years ^51 years Self employedor employee Self-employed Employee Unionmembership Non-unionmember Unionized Industrial sector Manufacturing Service by Employed Government Private/not-for-profit
Average weekly hours (ABS):

15 (25.4) 18 (30.5) 5 (8.5) 108 (22.0) 166 (33.8) 135 (27.5) 0.000 69(17.7) 54 (33.8) 59 (*5-9) 64 (19.9) 52 (30.6) 68 (18.2) 47(16.4) 42 (22.6) 29(46.8) 154(39.5) 84(21.5) 30 (18.8) 56 (35.0) 7* (31-^) 45 (*9-7) in (34.6) 95 (29.6) 39 (22.9) 44 (25.9) 143 (38.2) 96 (25.7) 109(38.1) 74(25.9) 63 (33.9) 48 (25.8) 9(14.5) 14(22.6) 0.000 0.012 0.001 0.000

^35 hours 36-49 hours ^50 hours

hostility.Workplacesize, public vs. privateorganization,urbanvs. were not regionallocation, and beingan employeevs. self-employed associatedwith job strainin men. Bivariateanalyses for females (first column, Table 3, Females) lowest occupashowed a wider range of job strain determinants: tional skill, working in the servicecomparedto the manufacturing all age groupscompared sector,beingan employeevs. self-employed, to the oldest, and for union memberscomparedto non-members. Basedon the resultsof Models E and F (Table3, Females),a similar set of nonsignificantcovariates was dropped. Hostility was not associatedwith job strain in women. The final models (G and H) show that the risk of job strain is elevated among middle-aged women, among women in low and middle skill-leveljobs, and - in contrast to males - among employees vs. self-employedworkers, among union members vs. non-members, and among women
2.009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 27

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

etal ^j Keegel

o
C

VJ

pt

*-^ .- . ^-~. ^^
Tj- H rA> O

- ^ .- . .- .
IAi ON H

^-^ - *
H H

,- ~
H

% 0
* g CJ

|S o
O Q

l^->qi dd-HH
H OC cA U-> MMHrl

|ol^i odd
I\Tj"O\ H * M

i vt
m o
*"! "T HHM

4 Sh 11
*"!

ON

-^

"c

00~ V0~ ^ 00"

1\ '5*"^

'hT 'Pi

P5 ^i >2(^ O

^S

o^ 5 ? O

l^iV^i

ddoH ' >-' "~ ^-^ q vc h tjMMriri

I ot iU i
MMri

ooh -^-^^ ^ 1* *1

I I I I I I

| i vl
ho - '
Hh

| <5H

H. *?

2
"O ^ 'rT"^ ^ 7T ^ OO' ^ ^ ^ 'T 'T ^ ^ ^ "^ "^T

1
<u
>^

3S

o^"
<3

44 i UA m" h
d d
MMHri

vi^J.^^V loLv^ol dddddd odd

HHHHHOOdd

i Asvi 4 Sdo hH
hhh

to

S
"C O

5^

o^" O Q

ooom "-" ^ ~~^ ^ O N m ir> mmHH

-odd - >tv-^-^v hhh

6 6 6 6 6 6 ^^-^^^^ i- O\O0 OO00 O hOOOOh

ho ^,^_ ^^? hh

il s

j> "* .S (3 j
"^

^U ^^ 1 .^"^
^

^ ^ ^ ^ r^r^-^-TJ,1111 dodo
HHMH

't *9" 't A rn 4 ,111 ooh

HHH

'noo ^o '^^'q tJ-hhhhH I I I I I I Hodddd

HHHOOH

^^ roH I I ho
i-^ M

^T H I h
H

1
s 1 vi

f
a

15 , s g|g a
^>* ^2i--^_

fills
^oo^

5li!ia
a>g^,Jc>

1 J
^^on-^

1 1 1
z8

|i!J!j

kjmfji|ii

rij

2009 Palgrave Macmillan 0197-5897

Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

h\^o

o h pj m

H ^ *>

* 9 ^

^^
Q

|iH<k<k
HHHrr,

IcAiiL
m H H

U vt i I I I
rA r4 M

III

ck
H

I Ss
Q

linivvi
HHHfT,

| ol i ik
MHH

Uh
mHH

| | |

III

i S

^5
Q

i i i U
HHHfn

i U/,
HHM

^
fAHH

i i ^
HM

i JLvl cki
HH M

^ ^ ^_^ ^_^ ^_^

^_^

^ ^^

^^

^ ^_^ ^_^ ^__^ ^_^

^_^ ^_^

^S; >W O Q

|ii.U -ii >ii>2^i ^ ^ *^'I HHhA

i <kA ^ J^vi i L ^ ^ >2>ii3- ^-!^^,3,33, ^hJTfu?v^r^'^riri hHh roritHHHH

| JL>fc i S. r 3,3, ^^ ^ mm

I
^2 "sP ^ fT\ H f^

Is

~ ^^ ^

1^^

3,3,3,-H^, 9^*?^
HhhH

i 4 i 4
hHh

tf\

(f|

fT%

HNHHHM

MM

J. Jkivi JLJ,
-!!, , 3, 3, 3, H. o noo h q
t fi h (5 h h

i vi 4
3,3, qoo
m 6

i
i q
m

-H ,, ^ H. >

"^ 73
^

It!

As5,

i !

*i.8.sii

^v^l

linn

fiiii!

im jjw

ijifu

niif?

!HJ
ri

t 1

2009 Palgrave Macmillan 0197-5897

Journal of Public Health Policy Vol. 30, 1, 17-39

29

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Keegel et al

working in the service vs. manufacturing sector. The association between job strain and female union members may be related to highly unionized industries such as health and community services also having a high percentage of female workers (n = 35 females and n = 6 males in VJSS as well as an increased risk of job strain). Effect size estimates (adjusted ORs) for these job strain determinants remained stable with varying combinations of covariates modelled. Job stress-related WC claims vs. job strain exposure Table 4 presents 'mental stress' WC claims patterns by occupation and age. Similar to the VJSS, the incidence of claims was higher among females than males (0.9 per 1000 vs. 0.7 per 1000). There were also many differences between the patterns emerging from the two sources. Claims data show the highest rates among workers employed in higher skill-level jobs, and for the 45-59 age range for both males and females. This contrasts with job strain patterns in the VJSS,where the highest prevalence of job strain was among lower skill levels and the youngest age group in males and 30-40-year olds in females. Finally we compared job stress WC claims in Victoria and VJSS job strain prevalence stratified by the 17 ABS industrial sector categories. Both claims rates and job strain prevalence were elevated for males and females in the health and community services sector (Figures 2 and 3). For males, the education and transport and storage sectors had high claims as well as high job strain prevalence (Figure2), as was the case for females in personal and other services, and finance and insurance sectors (Figure 3). However, there were a number of industries where the elevated prevalence of job strain was not reflected in claims patterns. These included manufacturing, construction, and wholesale trade for men (Figure 2), and retail for women (Figure3). Notably, job strain prevalence was particularly high in the accommodation, cafes, and restaurant sector for both males and females, but claims were not (Figures 2 and 3).

Discussion
Population-level exposure patterns showed that women were more likely to be exposed to job strain than men and that job strain prevalence was higher in younger workers and workers in lower
2.009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

30

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

a publichealthresponse to job stress Developing

Table 4: Victorian Workers' Compensation Data for 2003 - Case numbers and incidence rates for mental health stress claims (per 1000 workers) Male cases (IR)* Occupation (9 categories) Managers and administrators (skill level five) Professionals (skill level five) Associate professionals (skill level four) Tradespersons and related workers (skill level three) Advanced clerical and service workers (skill level three) Intermediate clerical, sales and service workers (skill level two) Intermediate production and transport workers (skill level two) Elementary clerical, sales and service workers (skill level one) Labourers and related workers (skill level one) Age (11 categories) < 20 years 20-24 years 25-29 years 30-34 years 35-39 years 40-44 years 45-49 years 50-54 years 55-59 years 60-64 years 65+ years Total 60 (0.6) 128 (0.6) 208 (1.5) 61 (0.3) 14 (1.2) 74 (0.8) 146 (0.8) 44 (0.5) Female cases (IRf 46 (1.7) 266 (1.2) 133 (1.3) 25(1.1) 62 (0.8) 236 (0.8) 22 (0.8) 93 (0.5) Total cases (IRf 106 (0.9) 394 (0.9) 341 (1.4) 86 (0.4) 76 (0.9) 310 (0.8) 168 (0.8) 137 (0.5)

55 (0.5)

51 (0.7)

106 (0.6)

n.p.b (0.1) 26 (0.2) 44 (o-3> 84 (0.6) 120 (0.9) 148 (1.1) 138 (1.1) 109(1.0) 85 (1.1) 31 (o-8) n.p.b (0.3) 792. (0.7)

8(0.1) 48 (0.4) 94 (0.7) 123 (1.0) 121 (1.1) 148 (1.2) 165 (1.4) 157(1.6) 55 (0.9) 14 (o-7) o (0.0) 933 (-9)

13(0.1) 74 (0.3) 138 (0.5) 207 (0.8) 241 (1.0) 2.96 (1.1) 303 (1.3) 266(1.3) 140 (1.0) 45 (-8) n.p.b (0.2) 17*5 (0.8)

aIR= Incidence rate data where the mechanism of the injury of disease was 'mental stress'. Data available from NOSI exclude self-employed individuals from the denominator. Vp. = data not published by NOHSC due to confidentiality restrictions.

skill-leveljobs. There were some areas of congruencebetween job strain exposureprevalenceand WC claims patterns- for example, higherprevalenceof job strain and higher claims incidenceamong women compared to men. We also observed some important
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 31

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Keegel et al

Figure 2: Victorian stress-related workers' prevalence, by industrial sector: Males

compensation

claims

incidence

vs. )ob strain

Figure 3: Victorian stress-related workers' compensation claims incidence vs. job strain prevalence by industrial sector: Females

The industrialsector with the highest prevalenceof discrepancies. strain for both males and females - accommodation, cafes, job and restaurants- had low stress claims rates. Job strain was most prevalentamong younger workers in lower skill-level/status
32 2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

jHfc'

workersin occupations,but claimswere highestamong middle-aged factors skill-level/status occupations.Many may explainthese higher findings: particular groups of workers have received insufficient OHS of educationand are unawareof the potential stress-relatedness their illnesses; workersfearlosingtheirjobs if they seek compensation, especially if precariously employed37; WC medical practitionersare reluctantto initiate stress-related claims26'27; submittedclaims are denied, or other reasons. Our study based on the same population and survey data mental illness greatly estimatedthat the cases of job strain-related this exceed the number of 'mental stress' claims.21Unfortunately, remainsa commonproblemwith occupationaldisease.We have also documented the under-recognitionand under-compensationof The analysis occupationalskin disease in working Australians.38'39 presented in our study suggests that a considerable amount of mental illness is likely to be occurring stress-related uncompensated workers. This interpretation is low skill-level among younger, and with consistent supported by a recent longitudinal study in New Zealand,in which it was estimatedthat 45 per cent of incident depressionand anxiety disordersamong a previouslyhealthy birth cohort of 32-year-oldworkerswas job stress-related.40 Claims-drivenprimary prevention intervention efforts, as in educational and health services sectors in Victoria, should be continued and expanded to integrate primary, secondary, and in a systemsapproach(Figurei).1 Our findings tertiaryinterventions data are an inadequate evidence base for that claims suggest developing comprehensivepolicy responses to job stress. Current efforts in Victoria in the education and health service sectors, for example, are limited to the public sector, yet our study found no differencesin job strainexposurelevels between public and private sectors for these industries.Further,our study findings show that the shortcomingsof claims-drivenpolicy and practice responses affect groups that are socially and economically disproportionately disadvantaged.
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 33

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

~t

Keegelet al

We acknowledgecertain limitations of the study. Although the of the working population, VJSSwas designedto be representative the study sample was taken from publicly available telephone exclude those workerswho listings, which may disproportionately are in less secure employment and in lower status groups. Shift workers and those working longer hours may also be underas participants were contactedon their home telephone represented numbers.These considerations suggestthat the disparitiesobserved are likely to be underestimates. There are also limitationscaused by comparingpatternsof job strainexposureprevalencefrom the VJSSwith claims patterns.WC statistics are based on accepted claims; informationregardingthe numbersof claims submittedis unavailable.Given the adversarial natureof the WC system, it is likely that many workerswith stressrelatedillnesseshave their claims rejected,or may be deterredfrom filing a claim. Anotherlimitationresultsfrom the classificationterm 'mentalstress'for stress-related claims. This narrowdefinitionmay cause an underreporting of stress-related illness. Possibly a wider definition might have includedcauses of other conditions, such as stress-relatedCVD.41 But job stress-relatedclaims for CVD are extremelyrare and thus would not appreciablyaffect our results. Thus, althoughthe narrowdefinitionused in WC and the likely low claims acceptancerates qualify our interpretation, these limitations also highlight the inadequacies of WC data as a proxy for occupationalexposureor disease surveillance. Some of the observedvariancebetween job strain exposure and claims patternsmight be explained by disease latency.The latency period between job strain exposure and the manifestationof job stress-related diseaseis not fully understood.Currentbest estimates indicatethat exposureto poor psychosocialworkingconditionscan be linked to adversemental health outcomes with a i-year latency was period.42Given that mean job tenureamong VJSSrespondents 7.5 years with 90 per cent of respondentsin their currentjob for 6 months or longer,33 it is reasonableto compareclaims rates and job strain prevalence from the same year. The age groups with the highest successfulclaims rates were between 40 and 59 years old, comparedto the age groups shown to have the highestexposureto job strain- 30-40-year-oldfemalesand < 30-year-oldmales. As the difference in age groups that we observed, between job strain
34 2.009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

*7*v

exposure and claims rates, ranges from 10 to 20 years, disease latencyis unlikelyto explain the observedvariation.
Implications for policy and practice

Development of a public health response to job stress requires informationregardingwhere the problem is at its worst, and thus where policy and practiceinterventionis most urgentlyneeded. In Australia and many other jurisdictions, government and other responses to job stress are primarily driven by WC insurance concerns.Our findingssuggestthat those most likely to be exposed to and adversely affected by job stress are the least likely to be compensated through WC; this situation exacerbates the gap between those who are successful in accessing compensation and those who do not, therebycontributingto health inequalities. Thus WC statisticsare an inadequateevidence base for developing equitable public health policy and practice responses to the job stressproblem.Indeed,a historicalrelianceon successfulWC claims as the sole indicator of job stress epidemiology has resulted in efforts for preventionof job stress being concentratedin specific industries and among particular groups of people when there may be others who are also highly exposed and are at risk. Public and evidence are urgently needed to health-basedunderstandings views, and to ensure that policy and complementinsurance-based practice responsesinclude those groups that are most affected by job stress. exposuredata on job stressare relatively Populationsurvey-based easy and inexpensive to obtain, and can be combined with risk estimatesfrom the best availableepidemiologicstudies for various associated health outcomes - available in some instances from published meta-analyses- to estimate PARs for common multifactorialchronic diseases (for example, depression,anxiety, CVD). The exposure or hazard data provide information for action identificationof prioritiesoccupationalgroups or work settings and the attributablerisk estimates provide the justification for both an essentialcomplementto claims-based action.This represents priority setting, and an efficient means of generatingthe public healthevidencebase neededto developequitablepolicy and practice responsesto this growingproblem,as well as to other occupational
2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39 35

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

yfc

Keegelet al

disease problems where exposure-diseaserelationshipshave been established. Claims-driven intervention efforts should be continued and and tertiaryinterventions expandedto integrateprimary, secondary, in a systems approach.This effort, however, needs to be complemented by similar comprehensiveinterventionefforts for younger and lower status workers,particularly for females in those groups, wherestress-related effectson healthcould be preventedby reducing job stressors and mitigated by effective compensation for stressrelatedillness. Acknowledgements AD LaMontagnewas supportedby a VictorianHealth Promotion Foundation Senior Research Fellowship (#2001-1088). Project National Heart fundingwas providedby grantsfrom the Australian Foundation (#G 01M 0345) and the Victorian Health Promotion Foundationto AD LaMontagne.T. Keegel was supportedin part from the Australian National Health and througha PhD scholarship Medical Research Council (#359306). A. Ostry was supported awardfrom the CanadianInstitutesfor througha New Investigator Health Research and a Scholar Award from the Michael Smith Foundationfor Health Researchin BritishColumbia. Supportfor collaborationbetween the Universityof BritishColumbia and the Universityof Melbourne was provided by an internationalcollaborations small grant from the Canadian Institute for Health Research(Grant#2oR 91434).

About the Authors

Tessa Keegel is a research fellow and PhD candidate at the Melbourne School of Population Health, The University of Melbourne. Her research interests are in work and health, and healthinequalities. Shehas a particular interestin exploringthe ways that policy operatesin the context of the workplace. Dr Aleck Ostry is currentlyan associateprofessorin the Facultyof Social Sciences at the Universityof Victoria. He holds a Canada
36 2009 Palgrave Macmillan 0197-5897 Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

y!

ResearchChair in the Social Determinantsof CommunityHealth and is also a SeniorScholarwith the Michael SmithFoundationfor Health Researchin British Columbia. He has an MSc in Health ServicePlanning,an MA in history (specializingin the history of He conductsan extensive publichealth)and a PhD in epidemiology. social determinants of health with a focus on food on the programme nutrition and policy. security Associate Professor Anthony D. LaMontagne's interest is in of work as a social developingscientific and public understanding in policy of health, and contributingto improvements determinant and practiceaimed at protectingpeople from the harmfuleffects of work as well as optimizingthe health-promoting aspects.He draws but integratesit with on his strongoccupationalhealth background, healthpromotion,sociological,historical,labourrelationsand other He collaborateswidely and across multipledisciplines perspectives. of the relationshipsbetween work and to advance understanding health,and to translatesuch knowledgeinto workplacehealthpolicy and practice.

References
i. LaMontagne, A.D., Keegel, T. and Vallance, D.A. (2007) Protecting &cpromoting mental health in the workplace: Developing a systems approach to job stress. Health Promotion
Journal of Australia 18(3): 221-228.

2. Frumkin, H., Williamson, M., Magid, D., Holmes, J.H. and Grisso, J.A. (1995) Occupational injuries in a poor inner-city population. Journal of Occupational and Environmental Medicine 37(12): 1374-1382. 3. Alamgir, H., Koehoorn, M., Ostry, A., Tompa, E. and Demers, P. (2006) An evaluation of hospital discharge records as a tool for serious work related injury surveillance. Occupational and Environmental Medicine 63(4): 290-296. 4. LaMontagne, A.D., Oakes, J.M. and Lopez-Turley, R. (2004) Long-term ethylene oxide exposure trends in US hospitals: Relationship with OSHA regulatory and enforcement actions. American Journal of Public Health 94(9): 1620-1626. 5. Schnall, P.L., Belkic, K., Landsbergis, P. and Baker, D. (2000) The workplace and
cardiovascular disease. State of the Art Reviews: Occupational Medicine 15(1): 1-334-

6. Peter, R. and Siegrist, J. (2000) Psychosocial work environment and the risk of coronary heart disease. International Archives of Occupational and Environmental Health 73(Suppl): S41-S45. 7. Belkic, K., Landsbergis, P., Schnall, P. and Baker, D. (2004) Is job strain a major source of cardiovascular disease risk? ScandinavianJournal of Work Environment and Health 30(2):
85-128.

2009 Palgrave Macmillan 0197-5897

Journal of Public Health Policy Vol. 30, 1, 17-39

37

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Keegelet al

8. Kivimaki, M., Leino-Arjas, P., Luukkonen, R., Riihimaki, H., Vahtera, J. and Kirjonen, J. (2002) Work stress and risk of cardiovascular mortality: Prospective cohort study of industrial employees. BMJ 325: 857-861. 9. Grosch, J. and Sauter, S. (2005) Psychologic Stressors and Work Organization. In: L. Rosenstock, M. Cullen, C. Brodkin and C. Redlich (eds.) Textbook of Clinical Occupational & Environmental Medicine, 2nd edn. Philadelphia: Elsevier, pp. 931-942. 10. Stansfeld, S., Bosma, H., Hemmingway, H. and Marmot, M. (1998) Psychosocial work characteristics and social support as predictors of SF-36 functioning: The Whitehall II study. Psychosomatic Medicine 60: 247-255. 11. Stansfeld, S.A., Fuhrer, R., Shipley, M.J. and Marmot, M.G. (1999) Work characteristics predict psychiatric disorder: Prospective results from the Whitehall II study. Occupational and Environmental Medicine 56: 302-307. 12. Niedhammer, I., Goldberg, M., Leclerc, A., Bugel, I. and David, S. (1998) Psychosocial factors at work and subsequent depressive symptoms in the Gazel cohort. Scandinavian Journal of Work Environment and Health 24(3): 197-205. 13. Bourbonnais, R., Comeau, M., Vezina, M. and Dion, G. (1998) Job strain, psychological distress, and burnout in nurses. American Journal of Industrial Medicine 34(1):
20-28.

14. Kawakami, N., Araki, S., Kawashima, M., Masumoto, T. and Hayashi, T. (1997) Effects of work-related stress reduction on depressive symptoms among Japanese blue-collar workers. Scandinavian Journal of Work Environmental and Health 23(1): 54-59. 15. Bosma, H., Marmot, M.G., Hemingway, H., Nicholson, A.C., Brunner,E. and Stansfeld,S.A. (1997) Low job control and risk of coronary heart disease in Whitehall II (prospectivecohort) study. BMJ 314(7080): 558-565. 16. Mino, Y., Shigemi, J., Tsuda, T., Yasuda, N. and Bebbington, P. (1999) Perceived job stress and mental health in precision machine workers of Japan: A 2 year cohort study. Occupational and Environmental Medicine 56: 41-45. 17. Cheng, Y, Guo, Y.L. and Yeh, W.Y (2001) A national survey of psychosocial job stressors and their implications for health among working people in Taiwan. International Archives of Occupational and Environmental Health 7Ah): 495-504. 18. LaMontagne, A.D., Keegel, T, Louie, A.M., Ostry, A. and Landsbergis, P.A. (2007) A systematic review of the job stress intervention evaluation literature:
1990-2005. 268-280. International Journal of Occupational and Environmental Health 13(3):

19. Fletcher, B. (1988) The Epidemiology of Occupational Stress. In: C. Cooper and R. Payne (eds.) Causes, Coping and Consequences of Stress at Work. Chichester: John Wiley 6c Sons, pp. 3-50. 20. LaMontagne, A.D., Keegel, T, Vallance, D.A., Ostry, A. and Wolfe, R. (2008) Job strain Attributable depression in a sample of working Australians: Assessing the contribution to health inequalities. BMC Public Health 8(181), http://www.biomedcentral.com/content/ pdf/1471-2458-8-1 81.pdf. 21. Sanderson, K. and Andrews, G. (2006) Common mental disorders in the workforce: Recent findings from descriptive and social epidemiology. Canadian Journal of Psychiatry 51(2): 63-75. 22. Andrews, G., Henderson, S. and Hall, W. (2001) Prevalence, comorbidity, disability and service utilisation. Overview of the Australian National Mental Health Survey. British Journal of Psychiatry 178: 145-153. 23. NOHSC. (2005) NOHSC On-Line Statistics Interactive 2 (NOSI 2). Canberra: National Occupational Health & Safety Commission (NOHSC), www.nohsc.gov.au/ OHSInformation/NOSI/default.asp, accessed and data downloaded 17 May 2005.

38

(Q 2009 Palgrave Macmillan 0197-5897

Journal of Public Health Policy Vol. 30, 1, 17-39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions

Developing a public health response to job stress

/rC*

24. Victorian Workcover Authority. (2006) Stressing the Point: A Study into What Causes Stress in a Budget Sector Workplace,and How it can be Controlled. Melbourne: Victorian Workcover Authority, April. 25. ASCC. (2007) Part E: The Mechanism Mental Stress. In: Compendium of Workers' Compensation Statistics Australia 2004-05. Canberra:Australian Government: Australian Safety and Compensation Council (ASCC), pp. 71-78. 26. Russell, G.M. and Roach, S.M. (2002) Occupational stress: A survey of management in general practice. Medical Journal of Australia 176(8): 367-370. 27. Steven, I.D. and Shanahan, E.M. (2002) Work-related stress: Care and compensation. Medical Journal of Australia 176(8): 363-364. 28. Azaroff, L.K., Levenstein, C. and Wegman, D.H. (2002) Occupational injury and illness surveillance: Conceptual filters explain underreporting.American Journal of Public Health 92(9): 1421-1429. 29. Macaskill, P. and Driscoll, T.R. (1998) National Occupational Injury Statistics: What Can the Data Tell Us? In: A.-M. Feyer and A. Williamson (eds.) Occupational Injury: Risk, Prevention and Intervention. London: Taylor & Francis, pp. 5-14. 30. Leigh, J.P. and Robbins, J.A. (2004) Occupational disease and workers' compensation: Coverage, costs, and consequences. Milbank Quarterly 82(4): 689-721. 31. Louie, A.M., Ostry, A.S., Quinlan, M., Keegel, T.G., Shoveller, J. and LaMontagne, A.D. (2006) Empirical study of employment arrangements and precariousness in Australia. Relations Industrielles/Industrial Relations 61(3): 465-489. 32. Ostry, A., Radi, S., Louie, A.M. and LaMontagne, A.D. (2006) Psychosocial and other working conditions in relation to Body Mass Index in a representative sample of Australian workers. BMC Public Health 6(1): 53. 33. Radi, S., Ostry, A. and LaMontagne, A.D. (2007) Job stress and other working conditions: Relationships with smoking behaviors in a representative sample of working Australians. American Journal of Industrial Medicine 50(8): 584-596. 34. Kivimaki, M., Virtanen, M., Elovainio, M., Kouvonen, A., Vaananen, A. and Vahtera, J. (2006) Work stress in the etiology of coronary heart disease - A meta-analysis. Scandinavian Journal of Work Environment and Health 32(6): 431-442. 35. Stansfeld, S. and Candy, B. (2006) Psychosocial work environment and mental health A and Health Work Environment Scandinavian review. 32(6): 443-462. Journalof meta-analytic 36. Koskenvuo, M. et al (1988) Hostility as a risk factor for mortality and ischemic heart disease in men. Psychosomatic Medicine 50: 330-340. 37. Quinlan, M. (2004) Workers' compensation and the challenges posed by changing patterns of work: Evidence from Australia. Policy & Practice in Health & Safety 2(1): 25-52. 38. Keegel, T, Erbas, B., Cahill, J., Noonan, A., Dharmage, S. and Nixon, R. (2007) Occupational contact dermatitis in Australia: Diagnostic and management practices, and severity of worker impairment. Contact Dermatitis 56(6): 318-324. 39. Keegel, T. et al (2005) Incidence and prevalence rates for occupational contact dermatitis in an Australian suburban area. Contact Dermatitis 52(5): 254-259. 40. Melchoir, M., Caspi, A., Milne, B.J., Danese, A., Poulton, R. and Mottitt, T.E. (2007) Work stress precipitates depression and anxiety in young, working women and men. Psychological Medicine 37: 1 1 19-1 129. 41. Driscoll, T. (2004) Overview Report on Work-Related Cardiovascular Disease. Canberra: National Occupational Health & Safety Commission. Report No.: NOHSC Paper Number 3/8. 42. de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D. and Bongers, P.M. (2004) The relationships between work characteristics and mental health: Examining normal, reversed and reciprocal relationships in a 4-wave study. Work Stress 18(2): 149-166.

Z009 Palgrave Macmillan 0197-5897

Journal of Public Health Policy Vol. 30, 1, 17-39

39

This content downloaded from 121.54.54.33 on Tue, 10 Sep 2013 05:08:14 AM All use subject to JSTOR Terms and Conditions