Despite efforts to more evenly distribute health professionals across Australia, rural and remote areas continue to face unequal access to healthcare services in comparison to their metropolitan counterparts [1
]. For rural communities to optimise healthcare service delivery, it is essential that they have clear information on what types of health professionals can be found in their local community, and their capacity to provide healthcare services. Health workforce planning should provide such information, and yet traditional approaches have typically focused on a supply–demand model, with assessments made between the anticipated number of professionals needed within a community and the actual supply [2
]. Whilst this approach has helped identify potential gaps in service provision and the need for training pipelines to be established, it has also fostered a siloed view of the healthcare workforce. An approach in which significant importance is placed on the availability and accessibility of specific types of health professionals to the neglect of the broader spectrum of healthcare service workforce [4
]. For rural communities, there is merit in moving beyond the potential contribution of individual health professionals as currently categorised to establishing an integrated, multidisciplinary framework that focuses more on the services needed within a community and how best to provide those services with the skill mix available in the existing healthcare workforce [4
]. This aspirational framework recognises the potential untapped capacity of the broader healthcare workforce to provide services that are needed through mechanisms such as role substitution and role expansion [4
]. This may be useful in situations where certain health professionals are in short supply locally, or the size of the population may not be sufficient to warrant, or be able to afford a particular service.
For rural and remote communities, there is hence the need in health workforce planning to move beyond headcounts of healthcare professionals to a more inclusive assessment of the healthcare workforce. However, a review of the literature reveals that previous evaluations of the healthcare workforce have used traditional approaches, focusing on either specific professions [8
] or professional groupings such as medicine [12
], nursing [13
], allied health [15
] or oral health [5
]. The aim of this study was therefore to illustrate a novel approach to health workforce evaluation for rural communities that broadened the scope of workers to better reflect the reality of healthcare service delivery. Specifically, this involved examining the number, type and location of workers involved in the provision of health, welfare and care (HWC) services in Tasmania, Australia between 2011 and 2016.
Tasmania is an island state with a population a little over 540,000 persons dispersed over a geographical area of 68,401 km2
. Being a largely rural state, Tasmania continues its historical struggle to adequately resource the healthcare sector, with a reliance on attracting skilled professionals from interstate and overseas [19
]. Understanding the local HWC workforce more broadly, including the identification of possible service gaps and potential untapped workforce capacity, therefore provides data conducive to effecting innovative and sustainable health workforce policy and planning across the state.
2. Materials and Methods
Data on the size and distribution of the Tasmanian population and the Tasmanian HWC workforce was extracted from the 2011 and 2016 Census of Population and Housing conducted by the Australian Bureau of Statistics (ABS). The Census of Population and Housing was considered the most appropriate data source for the purpose of this study because it includes information on regulated and unregulated workers employed across HWC roles, unlike other health workforce data sources (e.g., National Health Workforce Data Set (NHWDS) and Medical Education and Training (MET)) which limit the reporting of data to regulated health professionals.
To create a customised ABS dataset for analysis, comprehensive review of the Australian and New Zealand Standard Classification of Occupations (ANZSCO) was initially undertaken to identify all workers coded in the Census of Population and Housing as providing HWC services. As illustrated by the example of the breakdown of a Major Occupation Group seen in Figure 1
, ANZSCO uses a series of numerical codes in a five-tiered hierarchical structure to classify occupations, beginning with one-digit Major Group codes through to six-digit Occupation codes [20
]. Following a review of ANZSCO, First Edition, Revision 1 used in the 2011 census [21
], 202 six-digit Occupation codes were identified as HWC workers. A review of the updated ANZSCO, Version 1.2 used in the 2016 census [20
], subsequently produced 207 six-digit Occupation codes that could be identified as HWC workers. A customised dataset was then created by the ABS detailing the head counts and average hours worked for all three-digit Minor Groups and four-digit Unit Groups that included at least one of the identified six-digit Occupation codes by region, by gender and by level of employment participation (i.e., full-time or part-time) for people who specified their workplace location as Tasmania in the 2011 and 2016 Census of Population and Housing datasets.
To facilitate data analysis, cleaning and reorganisation of the dataset was necessary given that: (a) the ABS had provided additional data for six-digit Occupation codes not included in this study; and (b) ANZSCO codes are grouped according to likeness in training and indicative skill level and not necessarily disciplinary similarity [20
]. First, all six-digit Occupation codes not identified as HWC workers were removed from the dataset. To more accurately reflect the contemporary HWC workforce, the remaining three-digit Minor Group codes, four-digit Unit Group codes and six-digit Occupation codes were then regrouped into one of seven different categories based on the type of service they provide: (a) medicine; (b) nursing and midwifery; (c) allied health; (d) dentistry and oral health; (e) health-other; (f) welfare; and (g) carers (Appendix A
). Health-other was included as a category to capture those workers who were deemed to contribute to the provision of HWC services but who did not fit under any of the six alternative categories (e.g., biomedical engineer and medical laboratory scientist). In some cases, this regrouping required ANZSCO codes from different Minor Groups and Unit Groups to be clustered together; for example, 4114 Enrolled and Mothercraft Nursing was grouped with 254 Midwifery and Nursing to form the nursing and midwifery category. ABS data custodians advised that given the potential for randomisation of small cell values to either zero or three to ensure that identifiable information is not released, where possible, ANZSCO data should be analysed at the highest hierarchical level possible. Therefore, for this study, we chose to use Minor Group or Unit Group codes in situations where all subordinate four digit or six-digit ANZSCO codes were included. In situations where some ANZSCO codes that comprised a Minor Group or Unit Group were relocated or excluded, then the total of each remaining individual six-digit Occupation code was used instead.
For each category, data were analysed to establish: (a) total headcount, which was an aggregate of all individual headcounts from each Minor Group, Unit Group or individual Occupation codes (Appendix A
); (b) total number of FTE positions, which was calculated by summing the total weekly working hours for each included ANZSCO code divided by the average number of hours worked in a full-time job, which for the purposes of this study aligned with the Australian Institute of Health and Welfare’s standard of 40 h per week for medical professionals and 38 h for all other workers [22
]; and (c) annual hours of service per capita, which was calculated by dividing the total annual hours of service for each included ANZSCO code by the Tasmanian population figure. Data for each category were analysed for the state and for ABS statistical area level 4 boundaries (SA4) (Hobart, South East, Launceston and North East, and West and North West) (Figure 2
). Data for each category were also analysed according to gender as well as employment participation, with full-time employment defined as >35 h per week and part-time employment <35 h per week [23
]. If workers were employed but were absent from the workplace for the week preceding census night, they were recorded as ‘away-from-work’ [23
]. Finally, dependency ratios were calculated for statistical area level 4 boundaries, which reflected those aged under 15 years and over 65 years as a proportion of the population aged between 15 and 64 years in that region [24
This study has illustrated a novel approach to health workforce analysis that has included all types of HWC workers to better reflect the reality of changes in healthcare service delivery in rural communities. The overall expansion of the Tasmanian HWC workforce between 2011 and 2016 is consistent with national data illustrating strong growth in all sectors of the healthcare industry over the same period [25
]. Using a broad framework that included all types of HWC workers has, however, illustrated the varying growth in categories in response to changing healthcare demands. For Tasmania, growth has been largest in the carers, medicine and allied health categories over the five-year period, with increased demand for these workers possibly attributable to population demographics and policy initiatives. Tasmania has the highest national dependency ratio [24
], with almost 1 in 5 people aged over 65 years [28
]. With a program of aged care reforms announced in 2011 that included focusing on enabling the elderly to remain at home for longer [29
], this may explain the stronger growth for carers observed in this study. The elderly are also the most prolific users of medical care [30
], which is a further possible contributing factor to the greater growth in the medicine category. However, the introduction of the National Disability Insurance Scheme (NDIS) may also help explain the observed changes in the Tasmanian HWC. With Tasmania having the highest prevalence of disability in Australia [31
], the state was selected as one of four trial sites to introduce the NDIS in July 2013 [32
]. The subsequent progressive rollout of the NDIS has likely increased the number of carers and allied health professionals in the state, with both categories of workers in high demand to support NDIS participants [33
]. Tasmania also has some of the highest rates of chronic health conditions nationally [34
]. Growth in the allied health category may therefore also be associated with the increasing utilisation of allied health services to support chronic disease management [35
The only category of the Tasmanian HWC that declined was welfare. Despite a small gain in headcount, the overall number of FTE positions and per capita services provided by this category fell over the five-year period. This finding is in contrast to national data indicating solid growth in the collective welfare workforce over the same period [36
]. However, some professions comprising the welfare category, such as welfare workers, have declined in headcount over the past five years nationally [37
], suggesting that specific roles within this category are changing or may be less attractive. Certainly, the decline in FTE despite growth in headcount observed in this study suggests that there may be fewer full-time jobs or workers may not want to work as many hours in their current roles. However, the observed decline may also reflect a changed job market, where investment has shifted away from welfare roles in favour of other HWC categories. Policies such as the Better Access initiative, introduced in 2006 and then refined in 2009–2010, have likely fostered this shift [38
]. Designed to improve the provision of mental health services to the community, the initiative introduced a range of new items to the Medicare Benefits Schedule to better remunerate general practitioners, psychologists, social workers and occupational therapists for their time providing mental health services [38
]. With mental health care provided by medical and allied health professionals now more affordable, this may in part explain the declining FTE and per capita welfare services observed in this study [38
By including a broader range of HWC workers, this study moved away from examining the distribution of the healthcare workforce based on the number of registered professionals per 100,000 population [12
] and computed the annual hours of service per capita provided by individual categories and the total HWC workforce. This approach recognised that populations that are widely distributed, especially rural and remote communities, will rarely attract their notional ‘per capita’ allocation of a particular professional discipline. These circumstances may require other HWC workers to either increase service provision, or alternatively change roles, to compensate for a particular service not being represented locally. By determining both the breadth of the HWC workforce, together with the total hours of service provided to communities, a more holistic estimate of local healthcare service capacity was possible. This could help identify opportunities for role substitution and expansion to build local capability and increase the flexibility of the existing workforce to meet local healthcare needs when the recruitment and retention of some professions may be very difficult. Such circumstances underscore the growing recognition of the health-welfare interface, resulting from major national reforms in hospital services, primary care, aged care and disability support that promote and require integrated cross-sectorial care [40
]. However, given that unmet need may be provided by unregulated workers, authorities need to generate a stronger framework for regulation across healthcare more broadly to ensure the quality and safety of services provided [40
For rural communities such as Tasmania, whose population is widely dispersed across small, geographically distinct localities, the use of this per capita hours of service model would therefore be transformative for local health service managers facing the challenges of uneven distribution of the HWC workforce and service availability across the state. In the SA4 Hobart area, for example, where there are more total hours of service per capita than any other region of the state, health service managers can plan to deliver a range of healthcare services using a varying mix of HWC workers. The Hobart region includes Tasmania’s capital city and is home to just under half of the state’s total population. Some centralisation of services could be expected, particularly specialist medical services that are notably clustered in metropolitan areas across Australia [41
]. Although evidence of service decentralisation emerged over the five-year period, with greater proportional increases in annual hours of service per capita in both Launceston and the North East and the West and North West regions compared to the Hobart region, all other SA4 regions of the state will need to consider exactly what types of services they can practically offer from the HWC workforce available. In the West and North West region for example, where long-standing difficulties recruiting medical professionals are expected to persist [42
], an increase in all non-medical HWC workers is likely to continue to ensure healthcare is provided to the community.
Consistent with the profile of the broader Australian HWC workforce [25
], this study observed the Tasmanian HWC to be highly feminised, with around three quarters of workers being female. Medicine, dentistry and oral health, allied health, health-other and welfare also appear to be becoming increasingly feminised, with the proportion of females in each of these categories increasing over the five-year period. This mirrors national data [25
], confirming the long-standing observation that more females are entering some traditionally male dominated professions. Of note is that female HWC workers typically work fewer hours per week in comparison to their male counterparts [15
]. Therefore, as the proportion of female HWC workers continues to grow, so too will recruitment efforts given that additional workers will be required to achieve the meaningful change in FTE needed to keep pace with increasing service demands.
Unlike all other HWC categories, the Tasmanian care workforce showed growth in the number of male care workers over the five-year period which is consistent with national reports illustrating a gradual increase in the number of male care workers in recent years [26
]. An important factor behind this rise may be that industries such as agriculture and manufacturing have endured substantial job losses over the past decade [45
], resulting in men having to consider alternative employment options. Part of the attraction to care work may be the relatively short commitment to vocational study to achieve a relevant qualification. Further, care roles are typically dominated by part-time work arrangements [26
], which may appeal to men wishing to balance work and family commitments.
Although some categories of the Tasmanian HWC including medicine, allied health, welfare and dentistry and oral health continue to be dominated by full-time roles, this study observed a clear shift toward part-time employment for all workers comprising the HWC workforce. This trend appears characteristic of the systemic change in workforce participation observed both in Tasmania [46
] and nationally across all industries and sectors, not just healthcare [47
]. The reasons for the change are underpinned by both supply and demand factors, such as a move toward flexible and more cost-effective job offerings to the personal decision of workers to engage in part-time capacity to balance work commitments with study and family care [47
]. Although this personal choice may be thought to centre around those workers with young children, the literature illustrates that older workers nearing retirement [41
], as well as workers under the age of 40 years [44
], may also shy away from full-time roles. Given that both age cohorts are strongly represented in the Tasmanian population, and hence the Tasmanian HWC, this may account in part for the systemic shift toward part-time work. Future health workforce evaluation will need to capture both headcount and FTE to monitor the impact of declining workforce participation on the ‘total care’ provided to local communities [48
The present study also found evidence of the ‘disappearing working man’ phenomenon [46
], with more male workers taking up part-time employment than female workers over the five-year period. This trend challenges the traditional cultural norm of a ‘breadwinning male’ within each family and highlights that flexibility is likely to be equally important in recruiting workers, both male and female, to all health, welfare and care roles. While the move to part-time employment may in some circumstances be by choice, it has also been highlighted that around one quarter of part-time workers would like to work additional hours [47
]. Certainly, this study has illustrated the capacity of part-time HWC workers to work additional hours, with a greater percentage increase in FTE compared to headcount of part-time workers across most categories. This may well reflect greater economic rationalisation, with employers’ preference for a casual or part-time workforce so that hours can then flexibly increased or decreased in line with organisational pressures. If this is the case, the move away from secure employment arrangements has substantial implications for attracting HWC workers to the state, who may be reluctant to relocate from interstate for anything less than permanent full-time employment, particularly for those professions who are not supported by a local supply of graduates and are hence reliant on interstate and international migration [19
While this study has illustrated changes in the Tasmanian HWC, the findings must be interpreted with caution. Namely, census data is obtained through self-report and is not corroborated. Further, headcounts of smaller occupations may also be somewhat inaccurate given the random adjustment of cell values to ensure identifiable information is not released. In an effort to control for this, this study has used aggregated totals for professional groups where possible, thus helping to maintain the integrity of the dataset. Other features of the census also present as inherent limitations to reporting health workforce data, including the ‘away from work’ coding of census respondents. The current dataset observed large numbers of HWC workers recorded on census night as ‘away from work’, and with no way to determine their average hours of work, their contribution to HWC per capita hours of service across the state remains unmeasured. Finally, census data also fails to reflect the time spent on non-clinical versus clinical duties by HWC workers. Should administrative demands have grown over time, then the actual hours of direct clinical care provided to the community has likely been overreported in this study. Despite these limitations, the Census of Population and Housing provides a valuable resource for tracking both regulated and unregulated HWC workers, particularly in the absence of an alternative comprehensive dataset.