Area remoteness and the distribution and attrition of the rural health workforce in Australia

Abstract

Background

The health workforce (HW) plays an important role in patient care, and in rural Australia its distribution varies substantially. This paper explores trends in Australia’s HW full-time equivalent (HW–FTE) rates and estimates the risk of HW attrition phenomena using data from local government areas (LGAs) during 2013–2021.

Methods

Trends and spatial analyses were used to understand HW–FTE rates for allied health professionals, medical practitioners, and nurses and midwives in four major types of Australian Statistical Geography Standard (ASGS) remoteness areas. The time-to-event modelling was used to identify HW retention times and probability of HW attrition, by remoteness areas and major states in Australia.

Results

On average the HW–FTE rate at the granular geo-spatial LGA level exhibits variation in trends between States, rurality, LGA and health professional groups over the study period. The increase in the HW–FTE rate over time for medical practitioners and allied health professionals is lower for outer regional, remote, and very remote Australia compared to inner regional Australia. The HW–FTE rate is also consistently lower for rural Australia compared to major cities irrespective of HW professions. The average HW retention time estimated for allied health was highest in major cities (5 years), and lowest in outer regional areas (3 years). States such as NSW and QLD had more than 4 years of HW retention time for medical practitioners. For nurses and midwives, the average retention time was less than 3 years for all states in Australia. There is variation in trends in HW–FTE rate between LGAs within and between States, including markedly contrasting trends between geographically adjacent LGAs.

Conclusions

Our results provide new insight into variation in HW availability, and trends in availability, between major health professional groupings between States, degrees of rurality and local government areas across Australia. This presents new opportunities for understanding and addressing factors that underly the variation in trends for the purpose of refining policy and programs that aim to address the persistent maldistribution and shortages in health worker availability between major cities and regional and remote parts of Australia.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12960-026-01050-6.

Introduction

The health workforce (HW) in Australia has been facing significant challenges over the past decades [1–3], particularly in rural and remote areas, where health inequities exist [4–6]. These challenges are largely due to considerable imbalance in the geographical distribution of healthcare professionals across the nation [7, 8], where rural areas often face shortages of healthcare providers, leading to disparities in access to care and poorer health outcomes among residents [9]. Conversely, major cities and affluent regions tend to have a surplus of healthcare professionals, resulting in better access to services but potential inefficiencies in geospatial resource distribution [10].

In Australia, an array of programs, policies and incentives exist that aim to increase HW recruitment and retention in rural areas [11]. While some programs, such as the rural health multidisciplinary training program, demonstrate success [4, 12], evidence of effectiveness for other initiatives is generally limited [4, 13], and the maldistribution of the Australian HW remains pervasive. The Health Workforce 2025 report forecasts that a “business as usual” approach to workforce planning will not be sufficient to meet future health demands, and that significant reform is needed to address rural shortages [14]. Understanding the distribution of Australia’s rural HW is essential to inform policy which may improve HW maldistribution, and ultimately strengthen Australia’s healthcare system [15–17].

Existing literature demonstrates that the geographical distribution of the HW in Australia varies by profession. The full time equivalent (FTE) rate of medical practitioners, allied health professionals, pharmacists and dentists is higher in major cities compared to rural and remote areas [18–21]. On the other hand, statistics from the Australian Institute of Health and Welfare (AIHW) show that the FTE rate of nurses and midwives is higher in certain rural areas when compared to major cities [18]. AIHW statistics also show that rate of FTE general practitioners (GPs) is higher in rural and remote areas when compared to major cities [18, 21]. However, inability to differentiate between hours logged for GP consultations and the on-call and in hospital work carried out by rural GPs means these findings may not accurately represent availability [22]. AIHW statistics surrounding GP distribution are also contradicted by findings from Watt et al., that demonstrate GPs in rural areas work between 21 and 34% more hours compared to those in urban areas, reflecting possible shortages and increased demand on GPs in rural areas [23].

Trends in HW attrition across rural Australia appear to be more consistent between different health professions. Qualitative studies demonstrate that communities with increased affluence, available housing, adequate childcare, and spousal job opportunities are associated with decreased HW attrition and increased retention [13, 24–27]. On the other hand, factors such as lack of career progression, poor workplace management and community deprivation were found to be associated with increased HW attrition [13, 25, 28–30].

These drivers of health workforce maldistribution and attrition are reflected in the international literature. Globally, inequitable geographical distribution of health workers has been identified as a persistent challenge across diverse health systems, contributing to reduced access to care and poorer health outcomes in rural and remote populations [31–33]. Studies from high-income countries, including Canada, the United States, the United Kingdom and New Zealand demonstrate similar patterns of workforce concentration in metropolitan areas, driven by professional, economic and lifestyle factors, while rural areas experience chronic shortages and high turnover across multiple health professions [34, 35]. Despite differences in national policy contexts, the consistency of these findings suggests that rural workforce maldistribution is underpinned by structural factors common across health systems [32, 33].

Among general practitioners, ophthalmologists and family physicians, HW attrition increases with increasing area remoteness [7, 20, 36]. One study found that GPs are more likely to move to metropolitan practice than to rural practice, and that risk of attrition is highest in the first 3 years of rural practice [7]. Two sources show that while rural GP retention has improved, possibly indicating success of government policies, it remains that over time there is a decline in GPs working rurally [37, 38]. Bailey et al. further demonstrated that most new GPs left rural practice within 5 years [37]. Humphreys et al. found that among primary health workers, those employed in locations more than 500 km from a major city had approximately 2.5 times the risk of leaving compared to those located closer than 500 km to a major city [39].

Comparable trends have been reported internationally, with studies demonstrating that medical practitioners in rural and remote settings experience higher workloads, professional isolation, and limited career progression, all of which are associated with increased risk of attrition [31, 35]. International evidence further indicates that early career practitioners are particularly vulnerable to leaving rural practice, highlighting the importance of sustained workforce support beyond initial recruitment incentives [32, 35].

Similar findings were present among nurses, midwives and Aboriginal health practitioners. Russell et al. found that in rural and remote towns of the Norther Territory (NT), nurses, midwives, and Aboriginal health practitioners had a 20% chance of remaining in their practice after 1 year [40]. Within 0.34 years of commencement, half of the nurses and Aboriginal health practitioners in this study had ceased working [40]. Another study based on the NT found that in remote Aboriginal communities, the FTE rate of nurses, midwives and Aboriginal health practitioners will increase after a funding boost and then decrease over time, with 1.8–2.0 individuals being required to fill one FTE position per year [41].

International research similarly highlights high turnover among nursing and allied health workforces in rural and remote contexts, with reliance on short-term staffing models and repeated recruitment cycles placing strain on health systems and communities [31, 33]. Mapping reviews and systematic analyses emphasise that without coordinated, long-term workforce planning, short-term funding interventions may lead to transient workforce gains without sustainable improvements in service capacity [35].

The existing literature base demonstrates that inequities in the distribution and attrition of the HW exist in rural and remote Australia. However, the studies discussed above vary in sample size, methodology and overall strength of evidence. In addition, many of the findings discussed were specific to certain professions and geographical areas and may lack generalisability. While international evidence provides valuable comparative insights, national-level analyses remain essential to identify context-specific patterns and inform targeted policy responses [32, 33]. Analysis of Australia’s HW distribution and attrition over a larger geographical area and among a wider range of healthcare professions is needed to strengthen the existing evidence base, and to uncover overarching geographical trends. Understanding variations in these trends across different HW categories and remoteness areas is essential for securing comprehensive healthcare services for diverse patient needs.

Accordingly, this study aims to address the following research questions: (1) how does the geographical distribution of the health workforce in Australia vary across local government areas and remoteness classifications? (2) How do patterns of health workforce attrition and average years of retention differ between rural and urban areas, and across levels of remoteness? (3) To what extent do these patterns vary between major health workforce professional groups, including allied health professionals, medical practitioners, and nurses and midwives?

Hence, in this paper, we analyse data by local government areas (LGAs) in Australia and explore trends in HW distribution, HW attrition and average years of retention. We consider these trends over different remoteness areas and compare rural areas with major cities. Following the categories identified by the AIHW, and based on the available data, we focus on three major HW professional groups in Australia: allied health professionals, medical practitioners, and nurse and midwives. In this paper, we define ‘rural’ as the remoteness areas (RA) inner and outer regional, remote, and very remote, defined by the Australian Bureau of Statistics (ABS) [42]. The urban or city areas are defined as the major cities.

Methods

Study design and data sources

Health workforce data have been obtained through the Health Workforce Data Tool, Department of Health and Aged Care, Australian Government for the years 2013–2021 [43]. The data set is known as the National Health Workforce Data Set (NHWDS), which is aggregated from several major data sources, such as the Medical Education and Training (MET) data collections, the AIHW and the Australian Health Practitioner Regulation Agency (AHPRA). The AHPRA oversees the national registration process for 15 health professions. This process generates data annually, with information voluntarily provided through a workforce survey (both online and using a paper form) at the time of registration. AHPRA forwards these paper surveys to the Department of Health and Aged Care, where they undergo scanning, cleansing, and integration into the online registration and survey data set. The NHWDS includes demographic and employment details for registered health professionals. The NHWDS also includes geocoding based on the Australian Statistical Geography Standard (ASGS) areas defined by the Australian Bureau of Statistics (ABS).

In this paper, we used LGA level aggregation, and incorporated remoteness areas (RA) defined by the ABS [42]. We use the term ‘rural’ to refer collectively to four types of RAs: (i) inner regional; (ii) outer regional; (iii) remote; and (iv) very remote. We also include the RA category ‘Major Cities of Australia’ in the analysis to compare with the results from data in rural Australia. The RA categories ‘Migratory–Offshore–Shipping’ and ‘No usual address’ were excluded from the analysis. The shape of some LGAs have slightly changed over time during our study period, such as in 2016 and 2020. According to ABS, the LGA structure in 2016 was approximated using mesh blocks, where 90.5% of the LGAs were unchanged. In 2020, the major change was in LGA names and codes and ABS provided a geographic correspondence that enables the translation of data to 2020 LGAs. Hence, we used the correspondence file provided by the ABS to prepare the final data set for analysis [44]. Furthermore, data were not available for all LGAs in Australia; therefore, analyses were restricted to complete cases.

Statistical methods

We use the number of full-time equivalent (FTE) health professionals per 100,000 population following the definition provided by the Australian Institute of Health and Welfare (AIHW), which is also known as the FTE rate [18]. The HW–FTE rates provide a standardised measure of the number of full-time employees working in a particular role or specialty (e.g., allied health professionals, medical practitioners, and nurses and midwives) within the healthcare sector. We calculate the FTE rate for the $j$ th LGA for the $i$ th HW profession using following formula:

$$\text{FTE}{\text{rate}}_{\mathit{ij}}=\frac{\text{Total}{\text{FTE}}_{\mathit{ij}}}{\text{Total}{\text{population}}_{\mathit{ij}}}\times 100,000;i\in \left(\text{AHP},\text{MP},\text{NM}\right)$$

where AHP = Allied Health Professionals, MP = Medical Practitioners, and NM = Nurse and Midwives. The FTE rate plays an essential role for health authorities and policymakers to plan effectively for future workforce needs, ensuring that there are enough healthcare professionals to meet the demands of the population. To understand the trends in HW–FTE rate over time, we use time-series regression [45] adjusted over the RAs for all three major health profession categories considered in the study. The model-based HW–FTE rate change for each LGA is also estimated [46].

To understand the risk patterns of HW attrition by LGA, we used time-to-event modelling [47], which is based on HW–FTE reduction during our study period. We define the HW attrition event at the $j$ th LGA and for the $i$ th HW profession as follows:

$$\text{HW}\text{attrition}{\text{event}}_{\mathit{ij}}=\left\{\begin{array}{c}1,\text{reduction}\text{in}\text{FTE}{\text{rate}}_{\mathit{ij}}\\ 0,\text{otherwise}\\ \end{array}\right);i\in \left(\text{AH},\text{MP},\text{NM}\right)$$

The HW retention time is defined as number of years to the first HW attrition event during the study period. This definition focuses on initial workforce exit, which is of primary policy and planning relevance. Restricting the analysis to the first event ensures a single, well-defined outcome per individual and aligns with the assumptions of conventional time-to-event models by maintaining independence of observations. We explore the risk of HW attrition by remoteness areas and by major states in Australia.

We also explore the percentage change of the HW–FTE rate from 2013 to 2021 by HW profession categories. Furthermore, we explore geospatial models [48] to understand the HW–FTE rate change over years at the granular geospatial areas, including remoteness areas. Statistical analyses were conducted using the R programming language. Survival analysis was performed with the ‘survival’ package, while spatio-temporal models were fitted using the ‘sdmTMB’ package. Details of the statistical methods used in this study are provided in the supplementary section.

Results

Descriptive

Figure 1a highlights the local government areas (LGAs) used in this study. Out of 566 LGAs (using the 2020 baseline), a total of 368 LGAs were considered in rural Australia. For comparison with rural Australia, we also included 75 LGAs that represent major cities in Australia. Figure 1b provides distributions of the FTE rate by health professions for rural and regional Australia, where we observe that distribution of nurses and midwives is relatively balanced compared to the medical practitioners and allied health professionals. Especially, in very remote areas the distribution for allied health is highly skewed indicating a large variability for FTE rate. Furthermore, from Table 1, we observe that the average FTE rates are highest in major cities compared to rural Australia for all three HW categories considered in this study. In addition, the FTE rate for allied health is higher in inner regional areas compared to outer, remote, and very remote areas. Conversely, the average FTE rate for nurses and midwives is higher in very remote areas compared to remote, inner, and outer regional areas. The overall FTE rate is, on average, 3–5 times higher for nurses and midwives compared to allied health practitioners and medical practitioners. In addition, there is a statistically significant difference in the FTE rate among the four remoteness areas and major cities, across all three HW professions.

Fig. 1.

Map of Australia with Local Government Areas (LGAs). Available data from LGAs used in this study and corresponding remoteness areas are highlighted in (a). NA represents missing values. b Represents violin plots of the FTE rate (per 100,000 population) by health workforce professions and remoteness areas over the study period, i.e., 2013–2021

Table 1.

Summary statistics of the health workforce FTE rate per 100,000 population by Health Workforce categories and Remoteness Areas in Australia (2013–2021)

Remoteness areas	Health workforce
	Allied health		Medical practitioners		Nurses and midwives
	Mean (se)a	PCb	Mean (se)a	PCb	Mean (se)a	PCb
Major cities	799.8 (32.69)	34.8	723.2 (52.36)	4.62	1947.1 (138.0)	0.2
Inner regional	326.8 (5.77)	65.5	231.2 (5.51)	17.6	920.9 (17.30)	7.2
Outer regional	246.9 (5.51)	71.2	178.7 (4.63)	11.6	1030.4 (17.18)	7.9
Remote	218.9 (10.99)	52.8	195.0 (11.20)	−13.7	1324.1 (37.49)	5.7
Very remote	306.0 (24.60)	183.0	240.7 (11.15)	27.1	1445.5 (32.71)	11.9
p value*	< 0.001		< 0.001		< 0.001

^aMean (se)

^bPercentage change (PC) from 2013 to 2021

^*Kruskal–Wallis rank sum test. H0: There is no difference between RAs. p < 0.001 implies we reject H0

In Table 2, we provide the average HW–FTE rate by year from 2013 to 2021 for rural areas and major cities in Australia. On average, the FTE rate increases over time for all three HW profession categories. However, the rate of change is much higher for the allied health workforce, with an increase of about 183% in very remote Australia for allied health professionals (Table 1). Although the average FTE rate for nurses and midwives is much higher than for allied health professionals and medical practitioners, the percentage change in the FTE rate from 2013 to 2021 for nurses and midwives is much smaller compared to that for allied health professionals and medical practitioners. Interestingly, while the overall rural FTE rate for medical practitioners increased over the study period, for remote Australia there is a 14% decrease in 2021 compared to the FTE rate in 2013.

Table 2.

Comparison of the health workforce FTE rate (per 100,000 population) in rural and remote Australia with major cities by health workforce categories and study years (2013–2021)

Years	Health workforce
	Allied health†		Medical practitioner†		Nurses and midwives†
	Rural	Cities	Rural	Cities	Rural	Cities
2013	217 (8.8)	720 (93.4)	192 (9.3)	723 (160.3)	1060 (33.3)	2030 (440.8)
2014	217 (8.8)	726 (90.7)	200 (9.2)	726 (161.5)	1045 (32.8)	1998 (433.8)
2015	231 (12.0)	739 (93.2)	203 (10.0)	705 (153.7)	1036 (32.5)	1918 (411.4)
2016	247 (14.2)	750 (94.0)	216 (10.9)	714 (155.1)	1058 (34.7)	1923 (414.4)
2017	248 (14.9)	761 (96.4)	219 (10.6)	722 (154.7)	1071 (35.7)	1904 (402.2)
2018	266 (16.6)	772 (98.3)	220 (11.2)	722 (157.7)	1099 (35.6)	1923 (412.8)
2019	362 (18.5)	864 (103.5)	225 (10.9)	721 (158.2)	1115 (36.6)	1926 (412.9)
2020	398 (18.1)	896 (100.5)	213 (10.5)	719 (156.8)	1100 (36.1)	1867 (393.5)
2021	390 (14.8)	971 (111.8)	219 (10.8)	757 (163.7)	1145 (37.2)	2034 (424.7)
p value*	< 0.001	< 0.001	0.3	> 0.9	0.5	> 0.9

^†Mean (se)

^*Kruskal–Wallis rank sum test. H0: there is no difference between years. p < 0.001 implies we reject H0

Model-based

To understand the effect of remoteness on HW–FTE rates, we implement spatial cluster-based modelling and observe that rural Australia consistently has lower FTE rates compared to major cities for all HW categories (Table 3). Breaking it down specifically for rural areas, we find that outer regional, remote and very remote areas have lower FTE rates compared to inner regional areas for allied health professionals and medical practitioners. However, for nurses and midwives, the FTE rates are notably smaller for inner regional areas compared to outer, remote, and very remote Australia. On average, remote areas in Australia have 597 fewer FTE per 100 k population for allied health professionals, 573 fewer FTE rate for medical practitioners, and 650 fewer FTE rate for nurses and midwives compared to major cities in Australia. The situation differs in inner regional Australia for nurses and midwives, where there is an average reduction of 993 FTE rate compared to major cities (Table 3).

Table 3.

Model-based adjusted effects of remoteness areas on the health workforce FTE rate (per 100,000 population) compared to major cities by (i) allied health, (ii) general practitioners and (iii) nurses and midwives (2013–2021)

	Allied health	Medical practitioner	Nurses and midwives
	$\beta$(se)	$\beta$(se)	$\beta$(se)
Remoteness
Major cities	(reference)	(reference)	(reference)
Inner	−473.9 (58.3)*	−486.9 (83.1)*	−992.8 (225.6)*
Outer	−564.4 (63.0)*	−574.4 (89.9)*	−890.7 (244.0)*
Remote	−596.7 (81.4)*	−573.1 (116.2)*	−649.6 (315.2)*
Very remote	−505.4 (79.7)*	−520.0 (113.8)*	−558.0 (308.7)*
AIC	52821.5	51221.2	57714.0

$\mathit{\beta }$ refers to slope parameter of the regression model. ‘se’ refers to standard error

^*All estimates have p < 0.10. AIC Akaike information criterion

Figure 2 provides the model-based spatial representation of FTE rate change across 443 LGAs. We categorised the change into four categories for visualisation. Overall, we observe that 392 (88%), 303 (68%) and 299 (67%) LGAs show increased FTE rates over the years for allied health professionals, medical practitioners, and nurses and midwives, respectively. Among those, 69 (18%), 55 (18%), and 42 (14%) LGAs are from major cities for the three HW professions, respectively. In addition, we observe that about 36 (8%) rural LGAs for allied health professionals, and about 32 (7%) rural LGAs for nurses and midwives have an increase of 50 + FTE rate. Some rural LGAs also have negative FTE rate, such as about 45 (10%) rural LGAs showing negative FTE rates for allied health, 120 (27%) rural LGAs for medical practitioners, and 111 (25%) for nurses and midwives.

Fig. 2.

Categorised estimates for the FTE rate change using model-based approach. The FTE rate changes are spatially visualised at the LGA geographical areas

To understand the risk of HW attrition during the study years, we implement time-to-event modelling. Figure 3 provides the cumulative risk for HW attrition probabilities for the three workforce categories by remoteness areas and states in Australia. We observe that outer regional areas have the highest cumulative risk of HW attrition for allied health professionals and medical practitioners. In contrast, inner regional Australia has the highest cumulative risk of HW attrition for nurses and midwives. Furthermore, for allied health professionals and medical practitioners, Queensland has the lowest cumulative risk for attrition compared to other states in Australia. We also observe a relatively large variation of HW attrition between states and between rurality categories for allied health and for medical practitioners compared to variation for nurses and midwives. Table 4 shows that the estimated average retention time for allied health professionals and medical practitioners is the lowest in the outer regional Australia. Whereas, for nurses and midwives it is the lowest in remote Australia. Average retention times are consistently smaller for nurses and midwives compared to medical practitioners and allied health professionals, across all remoteness areas and states.

Fig. 3.

Cumulative risks estimated by years for remoteness categories and states in Australia

Table 4.

Estimates of the health workforce FTE (per 100,000 population) retention year for remoteness and major states in Australia by (i) allied health, (ii) general practitioners and (iii) nurses and midwives (2013–2021)

	Allied health	Medical practitioner	Nurses and midwives
	Estimated average time (in years) of HW retention or sustainment
	Mean (se)	Mean (se)	Mean (se)
Remoteness
Major cities	5.00 (0.35)	3.96 (0.30)	3.35 (0.25)
Inner	4.51 (0.24)	3.97 (0.19)	2.88 (0.11)
Outer	3.08 (0.13)	3.61 (0.17)	2.92 (0.15)
Remote	4.09 (0.39)	3.83 (0.37)	2.57 (0.14)
Very remote	4.48 (0.35)	4.66 (0.39)	3.38 (0.21)
States
NSW	3.27 (0.17)	4.14 (0.25)	2.56 (0.10)
VIC	3.44 (0.19)	3.67 (0.27)	2.41 (0.13)
QLD	4.51 (0.35)	4.51 (0.38)	2.78 (0.18)
SA	3.00 (0.18)	3.02 (0.18)	2.41 (0.08)
WA	3.96 (0.21)	3.91 (0.23)	2.89 (0.14)
TAS	3.24 (0.31)	3.16 (0.28)	2.84 (0.39)

‘se’ refers to standard error

All estimates have p < 0.05

Discussion

Our findings highlight inequities in the distribution of the HW in Australia while also prompting reflection on how equity in workforce distribution is defined and assessed. We found that the FTE rate of medical practitioners and allied health professionals is higher in major cities compared to rural areas, as is reflected in current literature [18–21]. For these professions, we see that inner regional areas had higher FTE rates compared to outer regional, remote, and very remote areas, suggesting that further disparities exist between levels of remoteness in rural LGAs. These patterns raise questions about whether equal geographic distribution is an appropriate benchmark, given differing population needs, service models, and scopes of practice across regions.

Excluding major cities, our findings show that the FTE rate of nurses and midwives was highest in very remote areas and decreases with decreasing level of remoteness. This is consistent with the current literature [18], and contrary to our findings among medical practitioners and allied health workers. In each remoteness category, nurses and midwives also exhibit FTE rates that are 3–5 times higher than that of medical practitioners and allied health professionals. The increased FTE rate of nurses and midwives in areas of increased remoteness, where there are significantly lower FTE rates of medical practitioners and allied health workers, does not necessarily reflect adequate health service delivery [49], and may even indicate that nurses and midwives are having to substitute for other professionals [50]. This highlights the importance of considering workforce “fit-for-purpose” rather than assuming equivalence between urban and rural service configurations. One of the ways to aid this using ‘rural generalism’, which refers to the need for rural health professionals to be generalists, rather than specialists in specific areas [51]. This includes the need for nurses to provide basic allied health care that in locations with better availability of health workers would be provided by allied health professionals.

Nurses and midwives exhibit the smallest HW retention time in each remoteness area and in each state, when compared to medical practitioners and allied health workers. This implies that nurses and midwives have increased attrition when compared to medical practitioners and allied health workers, which may increase expenses [52], and hinder quality of care [53]. These findings reflect those of Russell et al., who observed very high levels of turnover among nurses and midwives in remote areas of the Northern Territory [40]. While the distribution of nurses and midwives over remoteness areas appears to be more equitable than that of medical practitioners and allied health professionals, these findings suggest that distribution alone may be an insufficient indicator of workforce adequacy, and that the retention of nurses and midwives is poor, and this may need to be an area of stronger focus for policy makers. Further research could explore whether targeted incentives, retention programs, or professional support schemes were implemented during the study period and how these may have influenced these retention patterns.

The largest increase over time in the FTE rates in both rural areas and major cities were seen among allied health professionals, as is consistent with national statistics [18]. This may indicate that policy surrounding recruitment and retention is effective for allied health professionals, or reflect a general growth in the allied health sector as patient needs become increasingly complex [54–56]. While our findings demonstrated a general increase in the FTE rates over time for all professions, many individual rural LGAs demonstrated a decrease in the FTE rates for each HW category, and remote LGAs had an overall 14% decrease in the FTE rates of medical practitioners. These findings indicate that the growth of the HW over time is not equitable, and that certain rural areas are not experiencing the same increases in the HW–FTE rates that are being seen elsewhere. This suggests that system-wide gains may mask important localised workforce challenges. Understanding which policies or programs contributed to higher retention among medical practitioners and allied health professionals in very remote areas could inform similar strategies in other underserved regions globally.

For each profession, our results showed that major cities had the highest years of retention compared to inner regional, outer regional and remote areas. In addition, for all professions, areas with the highest risk of attrition were outside of major cities. These findings indicate that HW attrition is more prominent in rural areas compared to major cities, as reflected in current literature [7, 20, 36]. However, for allied health professionals and medical practitioners, years of retention were higher in very remote areas compared to major cities and other rural areas. These findings may reflect success of government policy and programs, and further highlights the importance of recognising heterogeneity within rural areas rather than treating non-urban regions as a single group. The lower cumulative risk of attrition in Queensland suggests that state-specific policies or workforce initiatives may have been effective, and initiates further investigation to identify best practices that could be applied in other regions or countries.

There are various limitations which should be noted when interpreting the results of this study. Due to lack of individual level data, HW retention time was estimated based on the number of years until FTE rates decreased. This means that attrition events may have been overlooked if they occurred, while overall FTE rates remained constant. As a result, this report may overestimate HW retention time and underestimate attrition. In addition, only the first attrition event was recorded in our time-to-event analysis, and all subsequent attrition events were not recorded. This approach limits our ability to consider changes in HW retention times as the years progress and may mask patterns in attrition that arise after the first attrition event. In addition, our geospatial models did not adjust for the clustering effects of repeated observations [54], which may also decrease the reliability of the results found.

While this study focuses on Australia, these findings provide insights relevant to other countries facing similar rural workforce challenges. Future research could explore comparative analyses of rural workforce distribution and retention policies internationally, and evaluate which strategies are most effective in different contexts. This would help inform global policy debates and guide evidence-based interventions to strengthen rural healthcare systems worldwide.

Conclusion

This paper contributes to ensuring equitable access to healthcare services, optimising healthcare delivery, and improving health outcomes for populations in rural regions. In this study we explored the FTE rates of three major health profession groupings across different remoteness areas in Australia. We also analysed HW trends over time to uncover changes in the composition of the HW, and variations in workforce attrition across different regions. Our findings demonstrate that inequities exist in both the distribution and attrition of the HW in rural Australia, and that trends differ between HW categories. Our findings will enable decision makers to tailor policy by identifying areas with shortages or surpluses of specific healthcare professionals and adjust resource distribution accordingly. In addition, our results will be valuable for identifying emerging workforce challenges and implementing targeted interventions to address them.

These findings provide practical guidance for policymakers and healthcare organisations. Specifically, they highlight the need for targeted workforce planning strategies, such as directing incentives and training programs to regions with persistent shortages, strengthening retention initiatives in high-attrition areas, and aligning workforce supply with local population needs. Our results can also support place-based workforce policies, improved recruitment strategies, and long-term planning to ensure a sustainable rural health workforce.

Understanding HW distribution in rural and remote areas is fundamental to achieving equitable access to healthcare, improving health outcomes, and advancing the overall efficiency and effectiveness of healthcare systems. By identifying and addressing disparities in workforce distribution, policymakers, healthcare organisations, and stakeholders can work towards building resilient, inclusive, and patient-centred healthcare systems that meet the needs of all individuals and communities.

Author contributions

KSB, RB, PR, and NH were responsible for the initial conceptualisation, while KSB and PR managed the data curation and formal analysis. Methodological development was a collaborative effort between KSB, PR, NH, and RB. KSB provided supervision and the writing process was a team effort with contributions from EM and KSB. All authors reviewed the manuscript ensuring a thorough and polished final product.

Data availability

Department of Health and Aged Care, Australian Government. Data link: https://hwd.health.gov.au/ Australian Bureau of Statistics. Data link: https://www.abs.gov.au/statistics/statistical-geography

Declarations

Competing interests

The authors declare no competing interests.