Does medical certification of workers with injuries influence patterns of health service use?

Abstract

BACKGROUND: Among workers with injuries who seek compensation, a general practitioner (GP) usually plays an important role in a person’s return to work (RTW) by advising if the worker is unfit for work (UFW), is able to work on alternate (ALT) duties or is fit for work and also providing referrals to other health service providers.

OBJECTIVE: To examine patterns of health service utilization (HSU) in workers with injuries by condition and type of certificate issued by GP.

METHODS: Zero-inflated negative binomial and logistic regressions were conducted for major healthcare services accessed over the 12-month period post-initial medical examination. Services included GP consultations, pharmacy, physiotherapy, occupational rehabilitation and psychology.

RESULTS: The average number of physiotherapy services was greater in workers with musculoskeletal disorders, back pain and fractures. In contrast, the median number of psychological services was greater in mental health conditions (MHC). Workers with ALT certificates were more likely to use GPs, pharmacy and physiotherapy services.

CONCLUSION: HSU in the 12 months post-initial medical certification varied substantially according to the worker’s condition, certificate type, age, gender and residential location. Understanding these factors can facilitate more appropriate resource allocation; strategic thinking on optimal use of particular health services and enables better targeting of particular provider groups for more education on the health benefits of RTW.

Keywords

Injury workers’ compensation return-to-work mental health

1 Introduction

Workers’ compensation claims data have been an important source of information to describe the incidence and effects of work-related injury and disease within and across jurisdictions [1 –3]. Among those workers with injuries who seek compensation, a general practitioner (GP) usually is the first point of call [4]. GPs not only treat injuries and diseases but also play an important role in a person’s return to work (RTW) [5 –9].

Being unable to work may negatively influence not only worker’s with injury recovery but also their family, employer and society [10 –16]. In order to lodge a worker’s compensation claim in some jurisdictions, workers need to obtain a medical certificate, usually issued by a GP [6 , 17–21]. For example, in Victoria Australia the vast majority (74%) of workers with injuries receive unfit for work (UFW) medical certificates the first time they visit a GP, with only 23% of certificates recommending alternate duties (ALT) and 3% certifying workers with injuries as fit for work [6, 22]. Being certified as UFW infers that a person is unable to perform his/her regular duties at work, is physically and/or mentally unwell and needs medical help to treat his/her disability.

In order to treat their injuries, workers will not only seek GPs help, but also will be referred to other health professionals depending on the nature and the type of the injuries. The type and volume of health services accessed by a person with injury will be determined by various factors such as the nature of the medical condition and demographic characteristics of the worker [23 –26]. Previous studies demonstrated substantial differences in the type of HSU between workers with different medical conditions. The most common differences were observed between workers with musculoskeletal disorders (MSD) [27] and mental health conditions (MHC); the former predominantly using physiotherapy, the latter - psychology services [28 –30]. More granular analyses exploring injury type and HSU reveal more subtle differences: For example, the findings of a study conducted in the US on workers suffering from low back pain [31], demonstrated that physicians were most often preferred providers by both employers and workers, but some workers were also likely to only visit chiropractors. Similarly, a Canadian study by Fleury et al. [32] revealed that while the majority of patients with MHC visited psychiatrists and case managers, only half of the study population consulted GPs. Wong et al. [33] also suggests that older age, severity of symptoms, type of previous treatment, and physician referral are all associated with increased utilization of mental healthcare services in patients with MHC. In terms of demographic factors, Field et al. [34] demonstrated that younger (below 16 years of age) and older (65+ years of age) people, especially females report higher rates of HSU, as do non-manual workers and those who are unemployed. In many studies, female sex remains a predictor of higher HSU over time [35, 36].

However, it is not yet known what happens when a worker with injury is recommended to RTW on alternate or modified duties. Such workers would potentially have less time and less need to see health care practitioners; the type and volume of HSU would also potentially be different from those who are completely unfit. Therefore, in this study we aimed to examine patterns of HSU post medical certification in workers with different conditions who have received different types of medical certificate at the initial GP consultation. Our hypothesis was that workers with injuries receiving ALT certificates would have fewer instances of HSU over the initial 12 months post-medical certification. Findings of previous studies suggest there is a dissonance between whom the initial sickness certifier might be and the provider ultimately responsible for worker’s with injury rehabilitation and RTW. Understanding these pathways to recovery through detailing worker’s with injury use of health services can allow for more targeted interventions to particular health providers about the benefit of work and return to safe work. We also believe that the findings of this study will help us to better understand patterns of HSU, and the resources required and associated with the type of certification and may lead to policy changes associated with type of certificate written.

2 Methods

2.1 Settings

As of June 2011 the state of Victoria, Australia had a total population of 5.5 million and working population of ∼2.8 million [37]. Employers in the state are required to maintain workers’ compensation insurance through the Victorian WorkCover Authority (VWA) unless they are able to self-insure or obtain insurance through other schemes. The VWA system provides coverage for ∼85% of the Victorian labour market for healthcare expenses and wage replacement costs for injuries and illnesses occurring during the course of employment. All work-related injuries and illnesses that exceed the pecuniary threshold for healthcare expenses ($642 in the 2013/14 financial year) or resulted in >10 days absence from work, are required to be lodged with the VWA via one of six private insurers (known as a claims agents) within 10 days of receipt of the employee’s injury claim form. Case-level data is collected by private insurers and provided to the regulator on an on-going basis for the purpose of managing and analysing the performance of the workers’ compensation scheme.

2.2 Data source

For the present study we accessed the Compensation Research Database (CRD) established at the Institute for Safety Compensation and Recovery Research (ISCRR) at Monash University. The CRD contains de-identified case-level administrative data received from the VWA between years 1986–2012 [38, 39]. More detailed information on the dataset used in this study is provided elsewhere [6].

2.3 Participants

All data of accepted compensation claims lodged by working age adults (15–65 years) with a date of injury between 1/01/2003 and 31/12/2010 were extracted from the database (n = 207,949). Nature of injury or disease was recorded using the VWA coding system, which was mapped to the Type of Occurrence Classification System (TOOCS, 3rd edition, http://www.safeworkaustralia.gov.au) for workplace injury and disease recording, and is reported using this classification system. Following several consultations within the research team, which included GPs, four categories of the most frequent reasons for issuing medical certificates were included into the analysis: (1) fractures, (2) MSD, (3) back pain and strains and (4) MHC [6]. Claims were excluded if:

They were accepted prior to 2003, as there were no adequate data on sickness certificates available;

The claim was for healthcare expenses only (i.e., the claim did not meet the 10 day work absence threshold, therefore no medical certificate was issued) (n = 78,086, 37.6%);

The initial medical certificate was written by a health practitioner other than a GP (n = 5,439, 2.6%);

The information on duration of certificates contained logical errors, such as certification date prior to injury date (n = 82, 0.04%);

Claims were for other traumatic injuries and other diseases (n = 27,870, 13.3%);

Claims were associated with fit for work certificates (n = 3,915, 1.9%).

Claims were grouped according to the type of the initial medical certificates: UFW (n = 71,561) and ALT (n = 20,996). The records of the included claims records were then linked to the VWA services database, containing the data of all the services paid for by VWA on behalf of workers with injuries from 2003 to 2011 (n = 33,646,888). The services’ data was then restricted to medical, rehabilitation and allied health services utilised within 12 months post-initial medical examination date. These services were then grouped into the eight most frequently accessed categories: (1) GPs, (2) hospital, (3) pharmacy (prescription pharmaceuticals), (4) physiotherapy, (5) occupational therapy (OT), (6) occupational rehabilitation and (7) psychology (services provided by psychologists only). ‘Hospital services’ refers to inpatient and outpatient services in public and private hospitals; the remaining services are primary and community care-based. The combined dataset resulted in 92,557claims and 3,912,027 healthcare services.

2.4 Statistical analysis

Collectively, a total number of claims and healthcare services were considered in this paper. Inferential statistics described the type of services used across health conditions and medical certificate types.

Due to a large amount of excess zero in the count data (i.e. no services utilized) observed within most healthcare services zero-inflated negative binomial regressions (ZINB) were conducted. ZINB models are described by two sets of predictors: 1) the logit model, employed to predict the probability of being a non-user of HSU, and (2) the negative binomial (NB) regression model estimating the number of HSU [40]. In addition, Vuong tests [41] were conducted to assess the appropriateness of a ZINB model against the standard negative binomial model.

Both models were adjusted for workers’ with injuries gender, age, medical condition, medical certificate type and residential location. As the ZINB models for ‘hospital’ and OT services did not converge, the outcome was dichotomised to ‘any vs. none’ services and logistic regression models were conducted instead. The results were summarized as beta coefficients, but are also shown as incident rate ratios (IRRs) and their 95% confidence intervals [95% CI]. The level of significance in this study was set at p < 0.05.

All analyses were conducted using STATA SE 12.0. Ethics approval for the project was gained from Monash University Human Research Ethics Committee.

3 Results

3.1 Type of certificate and worker’s condition

Over the 12 months period post-initial certification date there were 71,561 (75.2%) workers with UFW certificates that accessed 2,933,417 (74.9%) services and 20,996 (24.8%) workers with ALT certificates who accessed 978,610 (25.1%) services. Table 1 summarizes HSU by worker condition and medical certificate type.

Over the 12 months post-initial medical certification about one third (∼33%) of all workers with injuries accessed GPs. The second most frequently accessed category of services was hospital (6,563 or 23.1% of all workers with injuries) for fractures, followed by physiotherapy for MSD and back pain and strains – (32,259 or 25.2% and 13,304 or 29.1% workers respectively), and psychology for MHC (5,479 or 25.3% workers with injuries). Similar HSU frequencies were observed in UFW and ALT categories. Workers with fractures and MHC mostly accessed GPs (in total 195,529 or 38.1% and 112,951 or 35.9% HSU respectively); however workers with MSD and back pain and strains most commonly accessed physiotherapy (875,250 or 39.9% and 368,975 or 41.5% services respectively).

Figure 1 depicts the median and interquartile range (IQR) of HSU. The majority of physiotherapy services were accessed by workers with fractures, MSD and back pain and strains; however the median number of services in MSD and back pain and strains was higher in ALT certificates. The least amount of psychology services was accessed in fractures, MSD and back pain and strains. Unsurprisingly, psychology services were accessed most frequently in the MHC group. The median of HSU was similar in UFW and ALT medical certificates for MHCs; however there were more hospitalizations in ALT than in UFW in MHC. Workers with fractures in UFW category accessed more hospital and pharmacy services.

3.2 Most frequently accessed healthcare services (negative binomial regressions from ZINB models)

Table 2 reports the results of the ZINB regression models. The likelihood ratios for all the healthcare services in ZINB models were significant. The Vuong tests revealed the ZINB model as preferred against the standard NB model for all outcomes measures.

The negative binomial regression results of the ZINB models revealed that females were more likely than males to utilize all types of the healthcare services (IRRs ranging from 1.02–1.23). With increasing age workers across all conditions accessed more GPs (IRR = 1.01, 95% CI [1.00–1.02]) and physiotherapy (IRR = 1.01, 95% CI [1.00–1.02]).

Compared to fractures, workers with back pain and strains were significantly more likely to claim for pharmacy (IRR = 1.84, 95% CI [1.68–2.01]). Workers with MSD and back pain and strains were also significantly more likely to use occupational rehabilitation (IRR = 1.12, 95% CI [1.07–1.17] and 1.21, 94% CI [1.15–1.27] respectively) and other type of services (IRR = 1.47, 95% CI [1.38–1.57]; and 1.81, 95% CI [1.69–1.94] respectively). In contrast, workers with MHC were more likely to access pharmacy (IRR = 1.58, 95 % CI [1.42–1.74]) and psychological services (IRR = 1.48, 95% CI [1.34–1.63]). Workers with ALT certificates were also significantly more likely to see GPs (IRR = 1.03, 95% CI [1.02–1.05]) and physiotherapists (IRR = 1.03, 95% CI [1.01–1.05]).

With respect to the residential location, workers with injuries in rural areas were significantly less likely to access the majority of healthcare services; however workers with injuries living interstate were significantly more likely to use GPs (IRR = 1.20, 95% CI [1.14–1.26]) and pharmacy services (IRR = 1.27, 95% CI [1.07–1.48]).

3.3 Most frequent users of healthcare services (logit regressions from the ZINB models)

The logit regression of the ZINB models revealed, that females were significantly more likely to utilise most healthcare services (OR ranging from 0.67–0.79, please note that lower values of OR indicate increasing odds of being a more frequent user). With age (Table 3), the likelihood of using GPs (OR = 0.98, 95% CI [0.97–0.99]) and pharmacy (OR = 0.97, 95% CI [0.96–0.98]) increased in all workers with injuries.

Workers with MSD and back pain and strains were significantly more likely to use physiotherapy (OR = 0.39, 95% CI [0.37–0.41]; OR = 0.33, 95% CI [0.31–0.35] respectively) and occupational rehabilitation (OR = 0.78, 95% CI [0.74–0.82]; OR = 0.66, 95% CI [0.62–0.70] respectively). Furthermore, workers with MSD were also significantly more likely to use pharmacy services (OR = 0.75, 95% CI [0.69–0.82]). Workers with MHC were also significantly more likely to use psychology (OR = 0.03, 95% CI [0.03–0.04]) and occupational rehabilitation services (OR = 0.29, 95% CI [0.27–0.31]). Compared to the UFW, workers with ALT certificates were more likely to access GPs (OR = 0.68, 95% CI [0.49–0.94]), pharmacists (OR = 0.74, 95% CI [0.69–0.82]) and physiotherapists (OR = 0.84, 95% CI [0.81–0.88]).

Workers with injuries living in rural areas were less frequent users of the majority of healthcare services. Workers with injuries living interstate were more likely to access general practice (OR = 0.45, 95% CI [0.21–0.95]), physiotherapy (OR = 0.84, 95% CI [0.72–0.97]), occupational rehabilitation (OR = 0.62, 95% CI [0.55–0.70]) and psychology services (OR = 0.67, 95% CI [0.56–0.81]).

3.4 Utilization of hospital and OT services (logistic regressions)

Females, workers with MSD, back pain and strains and MHC were significantly less likely to access hospital services. Alternatively, with increasing age, workers with ALT certificates, living in rural areas and interstate were significantly more likely to access hospital services.

Workers with ALT certificates, particularly females were significantly more likely to access OT services. In contrast, with age, workers across all injury types, living in rural areas and interstate were significantly less likely to visit an occupational therapist.

4 Discussion

Many previous studies [23 , 42–44] have explored HSU patterns after injury and compared predictors of service use by the type of injury or disease. To our knowledge, the present study is the first one to describe HSU in workers with injuries by the type of the initial medical certificate and condition. The results revealed that HSU patterns in the 12 months post-initial certificate varied according to both the type of certificate issued by the GP at the initial consultation and the worker’s condition. Workers with physical injuries were more likely to seek help from a physiotherapist or occupational rehabilitation therapist whereas workers with MHC were more likely to visit psychologists. Workers with MSD and back pain and strains and who received alternate duties certificates from their GP were also more likely to access GPs and physiotherapists, compared to those who had fractures or MHC.

Our findings match previously published research, which suggested that there were differences in HSU across various injuries and diseases [23]. The findings of our study indicate that workers with MSD, fractures and back pain and strains accessed high amount of GP and physiotherapy services, which confirms the findings of Koehorn et al. [27] who also demonstrated increased utilization of similar services in work-related musculoskeletal conditions. An increased amount of medical, physiotherapy and chiropractic services was also observed in population suffering from neck pain [45]. Different patterns of HSU were observed in the MHC group, where workers with injuries tended to access hospital, psychology and pharmacy more frequently [46].

The results of our study also highlighted that age and gender were important factors determining the type of HSU – females were more likely to use all types of medical services, and workers with injuries visited more GPs as they got older. This may be related to numerous factors – age and gender would affect demand for care as they are related to morbidity. High consultation rates were characteristic of the very old and the very young, as well as women in the age group 15–44 years [47]. Females and older claimants had higher rates of HSU throughout Cote et al. study as well [45].

Residential location plays an important role in describing the type of healthcare services. Workers with injuries in rural areas are less likely to access any type of services, which perhaps is related to the limited number of service providers in the region, or long distance travelling. Previous research suggested that, following the injury, those in urban areas averaged much higher HSU [48, 49]. Less frequent usage of services in rural areas could be also associated with psychosocial characteristics, such as coping styles and health attitudes of workers with injuries [50]. Research has demonstrated that people in rural areas tend to be more active in their coping strategies and better able to cope with hardship and injury [51].

The strengths of this study are as follows: (1) the use of longitudinal study design and (2) the inclusion of initial medical certificate information. The limitations need to be noted as well. First, in this study we described HSU patterns since the date of the first medical certificate only, but we have not specifically considered subsequent certificates. Second, we examined only the first 12 months after the certificate was issued, which means we did not capture long-term patterns of HSU by those workers off work for more than 12 months. Third, we have not described the costs of the services utilised. Finally, the VWA database does not include any morbidity information, which could have provided insight into concomitant conditions that may have impacted the type of services worker with an injury had accessed.

4.1 Implications for research and practice

To the best of our knowledge no previous research has addressed HSU patterns post-initial medical certification in compensable injury. The findings of our study enable better understanding of HSU in workers with injuries across various medical conditions, workers’ occupation, age, gender and residential location in the first 12 months after the initial medical certification. The analyses presented here extend our understanding of the consequences of medical certification practices in primary care by helping to identify those groups of workers with injuries that are more likely to access increased number of health services as well as those providers whom might need to be targeted for more education on the health benefits of RTW and safe work. These findings have implications for more appropriate resource allocation and strategic thinking on optimal use of particular health services and professionals. For example, workers with ALT certificates access more GP services, which suggest that although they are back at work on alternate duties, they still require additional medical consultations and treatment on a regular basis.

5 Conclusions

Significant focus should be given to workers with MHC. Existing research on mental health treatment utilization has focused on identifying which patient characteristics are associated with receiving needed treatments [16 , 53]. Our findings suggest that a more balanced approach, incorporating both worker and health provider-level factors, is essential to developing policies and programs that can improve the unmet mental health needs of workers suffering from MHC. Future research may be necessary to better understand how to optimize post-injury mental healthcare service use.

Conflict of interest

The authors have no conflict of interest to report.

Footnotes

Acknowledgments

Funding for this project, Dr Ruseckaite’s and Dr Collie’s involvement in the study was provided by a research grant from WorkSafe Victoria.

References

Anema

, Schellart

, Cassidy

, Loisel

, Veerman

, Van der Beek

. Can cross country differences in return-to-work after chronic occupational back pain be explained? An explanatory analysis on dissability policies in a six country cohort study. J Occup Rehabil 2009;19:419–26.

Campolieti

. Multiple state hazard models and workers’ compensation claims: An examination of workers compensation data from Ontario. Acc Anal Prev 2000;33:197–202.

Garcy

, Mayer

, Gatchel

. Recurrent or new injury outcomes after return to work in chronic disabling spinal disorders: Tertiary prevention efficacy and functional restoration treatment. Spine 1996;21:952–9.

Vedsted

, Christensen

. Frequent attenders in general practice care: A literature review with special reference to methodological considerations. Public Health 2005;119(2), 118–37.

Bollag

, Rajeswaranc

, Ruffieuxc

, Burnandc

. Sickness certification in primary care - the physician’s role. Swiss Med Wkly 2007;137:341–6.

Collie

, Ruseckaite

, Brijnath

, Kosny

, Mazza

. Sickness certification of injured and ill workers by General Practitioners in the state of Victoria, Australia: A retrospective population-based cohort study. Med J Aust 2013;199(7), 480–3.

Deyo

. The role of the primary care physician in reducing work absenteeism and costs due to back pain. Occup Med 1988;3(1):17–30. PubMed PMID: 2963389. Epub 1988/01/01. eng.

Diaz-Ledezma

, Urrutia

, Romeo

, Chelen

, Gonza'lez-Wilhelm

, Lavarello

. Factors associated with variability in length of sick leave because of acute low back pain in Chile. Spine 2009;9:1010–5.

Driscoll

, Hendrie

. Surveillance of work-related disorders in Australia using general practitioner data. Aust NZJ Public Health 2002;26:346–51.

10.

Feuestein

, Shaw

, Lincoln

, Miller

, Wood

. Clinical and workplace factors associated with a return to modified duty in work-related upper extremity disorders. Pain 2003;102.

11.

Flach

, B

, Groothoff

. Determinants of sick-leave duration: A tool for managers? Scaand J Pub Health 2008;36:713–9.

12.

Labriola

. Conceptual framework of sickness absence and return to work, focussing on both the individual and the contextual level. Work 2008;30:377–87.

13.

Leigh

. Employee and job attributes as predictors of absenteeism in a national sample of workers: The importance of helath and dangerous working conditions. Soc Sci Med 1991;33(2):127–37.

14.

Leigh

, Robbins

. Occupational disease and workers’ compensation: Coverage, costs, and consequences. Milbank Q 2004;82(4):689–721.

15.

MacEachen

, Kosny

, Ferrier

, Chambers

. The “toxic dose” of system problems: Why some injured workers don’t return to work as expected. J Occup Rehabil 2010;20(3):349–66.

16.

McManus

, Mowlam

, Dorsett

, Stansfeld

, Clark

, Brown

, et al. Mental health in context: The national study of work-search and wellbeing, DWP Research Report No 810 Available at: http://researchdwpgovuk/asd/asd5/rports2011-2012/rrep810pdf. 2012

17.

Dasinger

, Krause

, Thompson

, Brand

, Rudolph

. Doctor proactive communication, return-to-work recommendation, and duration of disability after a workers’ compensation low back injury. J Occup Environ Med 2001;43(6):515–25. PubMed PMID: 11411323 Epub 2001/06/20. eng.

18.

Gimeno

, Benavides

, Benach

, Amick

III . Distribution of sickness absence in the European Union countries. Occup Environ Med 2004;61:867–9.

19.

Skaner

, Nilsson

, Arrelov

, Lindholm

, Hinas

, Alexanderson

. Use and usefulness of guidelines for sickness certification: Results from a national survey of all general practitioners in Sweden. BMJ Open 2011;1(e000303.): doi: 10.1136/bmjopen-2011-000303

20.

Soler

. Sick leave certification: A unique perspective on frequency and duration of episodes - a complete record of sickness certification in a defined population of employees in Malta. BMC Family Practice 2003;4(2) http://www.biomedcentral.com/1471-2296/4/2

21.

Wynne-Jones

, Mallen

, Mottram

, Main

, Dunn

. Identification of UK sickness certification rates, standardised for age and sex. Br J Gen Pract 2009;59:510–6.

22.

Ruseckaite

, Collie

, Bohensky

, Brijnath

, Kosny

, Mazza

. Trends in sickness certification of injured workers by general practitioners in Victoria, Australia. J Occup Rehabil 2013; [Epub ahead of print].

23.

Cohen

, Krauss

. Spending and service use among people with the fifteen most costly medical conditions, 1997. Health Affairs 2003;22(2):129–38.

24.

Roehrig

, Miller

, Lake

, Bryant

. National health spending by medical condition, 1996–2005. Health Affairs 2009;28(2):w358–w67.

25.

Beattie

, Nelson

, Basile

. Differences among health care settings in utilization and type of physical rehabilitation administered to patients receiving workers’ compensation for musculoskeletal disorders. J Occup Rehabil 2013;347–60.

26.

Gerritsen

, Deville

. Gender differences in health and health care utilization in variuos ethnic groups in the Nedherlands: A cross-sectional study. BMC Public Health 2009;9(109): doi: 10.1186/471-2458-9-109

27.

Koehoorn

, Cole

, Hertzman

, Lee

. Health care use associated with work-related musculoskeletal disorders among hospital workers. J Occup Rehabil 2006;16:411–24.

28.

Burns

, Kendrick

. Care of long-term mentally ill patients by british general practitioners. Psychiatric Services 1997;48(12):1586–8.

29.

Iversen

, van Staden

, Hughes

, Browne

, Greenberg

, Hotopf

, et al. Help-seeking and receipt of treatment among UK service personnel. The British Journal of Psychiatry 2010;197(2):149–55.

30.

Kendrick

, Burns

, Freeling

, Sibbald

. Provision of care to general practice patients with disabling long-term mental illness: A survey in 16 practices. British Journal of General Practice 1994;44(384):301–5.

31.

Côté

, Baldwin

, Johnson

. Early patterns of care for occupational back pain. Spine 2005;30(5):581–7.

32.

Fleury

, Grenier

, Bamvita

, Caron

. Professional service utilisation among patients with severe mental disorders. BMC Health Services Research 2010;10.

33.

Wong

, Schell

, Marshall

, Jaycox

, Hambarsoomians

, Belzberg

. Mental health service utilization after physical trauma: The importance of physician referral. Medical Care 2009;47(10):1077–83. doi: 10.1097/MLR.0b013e3181a80fc9

34.

Field

, Briggs

. Socio-economic and locational determinants of accessibility and utilization of primary health-care. Health and Social Care in the Community 2001;9(5):294–308.

35.

Green

, Pope

. Gender, psychosocial factors and the use of medical services: A longitudinal analysis. Social Science & Medicine 1999;48:1363–72.

36.

Parslow

, Jorm

, Christensen

, Jacomb

, Rodgers

. Gender differences in factors affecting use of health services: An analysis of a community study of middle-aged and older Australians. Social Science & Medicine 2004;59:2121–9.

37.

ABS. Australian Bureau of Statistics [cited 2013], Available from http://www.abs.gov.au/

38.

Ruseckaite

, Collie

. Repeat workers’ compensation claims: Risk factors, costs and work disability. BMC Public Health 2011;11.

39.

Ruseckaite

, Clay

, Collie

. Second Workers’ compensation claims: Who is at risk? Analysis of WorkSafe Victoria, Australia compensation claims. Can J Public Healthe 2012;24(103):e309–13.

40.

Moon

, Shin

. Health care utilization among Medicare-Medicaid dual eligibles: A count data analysis. BMC Public Health 2006;6(88) doi: 10.1186/471-2458-6-88

41.

Voung

. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica 1989;57:307–3304.

42.

Labriola

, Christensen

, Lund

, Nielsen

, Diderichsen

. Multilevel analysis of workplace and individual risk factors for long-term sickness absence. JOEM 2006;49(9):923–9.

43.

Maher

, Williams

, Lin

, Latimer

. Managing low back pain in primary care. Aust Prescr 2011;34:128–32.

44.

Mahmud

, Webster

, Courtney

, Matz

, Tacci

, Christiani

. Clinical management and the duration of disability for work-related low back pain. JOEM 2000;42:1178–87.

45.

Cote

, Yang

, Kristman

, Hogg-Johnson

, Van Eerd

, Rezai

, et al. The association between workers’ compensation claims involving neck pain and future health care utilization: A population-based cohort study. J Occup Rehabil. 23(4):547–56. doi: 10.1007/s10926-012-9415-8

46.

de Lusignan

, Chan

, Parry

, Dent-Brown

, Kendrick

. Referral to a new psychological therapy service is associated with reduced utilisation of healthcare and sickness absence by people with common mental health problems: A before and after comparison. J Epidemiol Community Health 2012;(66:e10. doi: 10.1136/jech.2011.139873)

47.

Banks

, Beresford

, Morrell

, Waller

, CJ

. Factors influencing demand for primary medical care in women aged 20-44 years: A preliminary report. Int J Epid 1975;4:189–95.

48.

Dixon

, N

. Researching the rural-metropolitan health differential using the ‘social determinants of health’. Aust J Rural Health 200;8:254–60.

49.

Parslow

, Jorm

. Who uses mental health services in Australia? An analysis of data from the National Survey of Mental Health and Wellbeing. Australian and New Zealand Journal of Psychiatry 2000;34:997–1008.

50.

Young

, Cifuentes

, Wasiak

, Webster

. Urban-rural differences in work disability following occupational injury: Are they related to fifferences in healthcare utilization? JOEM 2009;51(2):204–12. doi: 10.1097/JOM.0b013e3181917d8e

51.

Harvey

. Understanding Australian rural women’s ways of achieving health and wellbeing—a metasynthesis of the literature. Rural Remote Health 2007;7:823.

52.

Cherry

, Burstyn

, Beach

. Mental ill-health and second claims for work-related injury. Occup Med 2012;62:462–5.

53.

Koopmans

, Roelen

, Bultmann

, Hoedeman

, Van der Klink

, Groothoff

. Gender and age differences in the recurrence of sickness absence due to common mental disorders: A longitudianl study. BMC Public Health 2010;10(426):1–9.