Review of Australian Government Health Workforce Programs

Appendix ii: Health Workforce 2025 - summary

Page last updated: 24 May 2013

Health Workforce 2025 (HW2025) is a health workforce projection study undertaken by Health Workforce Australia (HWA) to assist in future workforce planning. The study aimed to model future health workforce supply and demand across a number of possible policy scenarios.

Health Workforce 2025 – Doctors, Nurses and Midwives Volumes 1 and 2, was released in April 2011 and Health Workforce 2025 – Medical Specialties Volume 3 was released in November 2012.

This appendix provides a short summary of the major findings of the HW2025 analysis. It is not intended to replace the nuance and complexity of the three volumes of the HWA report (and apologies are tendered in advance to the authors of that report if this summary does not do their work justice), but may provide a useful reference point in relation to some of the workforce planning and data issues discussed within this review.

Conclusions drawn from the HWA findings

This appendix does not seek to summarise the HWA conclusions in their entirety. However, some relevant observations are as follows:

  • Transformation of health policy settings is required for health services to be sustainable since the comparison (“do nothing”) scenario shows that there will be a continued geographic maldistribution of doctors and a very significant shortage of nurses by 2025.
  • The comparison scenario modelling predicts that there will be a shortage of 109,490 nurses at 2025, which is a 28% shortfall on estimated demand, and a shortage of 2,701 doctors by 2025, a less significant shortfall of 2% below expected demand at that time. (The midwifery workforce data available at the time to HWA was regarded as insufficiently reliable to make reasonable predictions for this workforce segment.)
  • Although the HWA report does not explicitly attempt to specify the projected direct and indirect costs of employing, for example, 110,000 nurses, it is clearly both unaffordable in total budget terms and unachievable in terms of recruitment and training. HWA’s conclusion is that investment must be made in innovation and reform measures, leading to productivity gains.
  • For nurses, HWA modelling projects that measures to improve retention could make a significant impact in reducing workforce shortages. The report notes that there has been increased workforce retention within nursing recently (2007-2008), possibly as a result of the impact of the global financial crisis upon superannuation and retirement funds. If these slowed exit rates were to continue, the expected shortage would decrease by 77%.
  • For the medical workforce, HWA modelled a scenario where demand continues to increase and the working hours of doctors continue to reduce, resulting in an even greater shortage in the projected workforce.
  • HWA concluded that shortfalls cannot be filled by increased training under the clinical training model as currently configured. HWA modelling indicates that this issue is even more highly problematic in the case of medical postgraduate training (specialist training) which will need to be increased in line with increasing medical graduate numbers and graduate demand.

Volume 1

Policy levers

HW2025 presents four policy levers which have the potential to address shortages around workforce reform, training, immigration and geographic distribution. The most promising area, for HWA, concerns innovation and reform measures.

The potential solutions identified in HW2025 include changing models of care, adjustments to skills mix, expanded scope of practice, use of assistants or increased use of technology (i.e. telehealth) which aims to increase the efficiency of the workforce without needing potentially significant investments in education.

HWA’s workforce planning model appears to be particularly sensitive to changes in demand for health services. It is suggested that lowering demand can be achieved through better health promotion and prevention programs and new technology allowing greater numbers of conditions to be treated diagnosed. Options for improving workforce retention (nurses only) included improving the workplace environment (such as provision of adequate equipment and resources), involvement in decision making, leadership support and the ability to practice to the full scope of practice.

Enhanced targeting of international health professionals to areas of need (whether by specialty or geography) is one strategy to ensure adequate workforce numbers and distribution. The third key policy area is education and training, which is generally defined to mean the quantity and funding of education and training places to educate the undergraduate and postgraduate health workforce. Potential policy reforms outlined by HWA include shortening the length of training, improving the work readiness of graduates, streamlining courses and introducing pathways to generalism especially in rural and remote areas.

HWA notes that any reforms to policies in this area are attended by complexities given the spilt in responsibility between the Commonwealth Education and Health portfolios, state and territory governments as well as the medical colleges, the private and community sectors in funding and providing training to the health workforce. In order to address this issue, HWA has received agreement from Health Ministers to establish a National Medical Training Advisory Network to improve coordination between stakeholders.

Modelling

The principal method used by HWA to develop the projections of the medical, nursing and midwifery workforce numbers is mathematical simulation utilising a “stock and flow” model, where people entering and exiting the workforce (flows) periodically adjust the initial number in the workforce (stock). The workforce is broken down into age and gender cohorts and different flow rates are applied to each cohort. Demand projections are based on service utilisation rates for each population age and sex cohort.

The principle source of HWA’s workforce supply data was the 2009 AIHW labour force survey data, estimates of medical and nursing graduates entering the workforce and immigration information on international medical and nursing professionals. However, any model is necessarily only as good as the quality of the data on which it relies and the robustness of the assumptions used to build the model.

In the case of the 2009 AIHW workforce survey data, the response rates to the nursing (44.4%) and medical workforce (53.9%) surveys were low, casting some doubt on the validity of this data. The model will be updated in the future with newly obtained data from the 2011 medical and nursing workforce surveys for which the response rates have been much better; over 80% in both cases.

Scenarios

For the sake of completeness, and to provide the context of current policy debates in this sector, a short summary is provided of the workforce scenarios modelled by HWA.

HWA modelled all scenarios against a “comparison” scenario which assumes that current policy settings are applied into the future and held constant. This comparison scenario results in the projected 2,701 excess demand (shortage) for doctors and a 109,490 excess demand (shortage) for nurses in 2025. These projected shortfalls have in many ways become the “headline” stories arising from the HWA report, and are frequently cited without reference to the more nuanced and complex work undertaken by HWA.

Innovation and reform scenario

These are scenarios predicated upon changes resulting from various demand management initiatives. They are designed to illustrate how responsive the workforce can be to such changes.

  • The “productivity gain” scenario assumes a 5% gain from various initiatives which HWA believes are possible with current reform projects. This scenario predicts a 2,811 surplus of doctors and an 89,993 shortage of nurses.
  • The “low demand” scenario assumes a 2% reduction in demand due to the effects of preventative health strategies and reform in health care. This scenario envisages that it is possible to achieve an 18,690 oversupply of doctors and a 31,355 shortage of nurses.
  • A scenario based on increasing retention rates for nurses applied the unexpected lower 2007-08 nursing exit rate applied across all future projections. Under such a scenario, a shortage of 24,846 was predicted.

Immigration scenario

These scenarios test the impact of changing current immigration levels to move towards self-sufficiency. Under these scenarios, the demand for services is held constant as the number of entrants into the workforce from international graduates and migrants are reduced. Notably, and unsurprisingly, under “self-sufficiency” scenarios there is projected to be an acute shortage of both doctors and nurses. Key points include:

  • The medium self-sufficiency scenario postulates a 50% reduction in international migration. The inflow of migration would be progressively cut to 50% of its initial level by 2025. This scenario results in a 9,300 shortage of doctors and a 129,818 shortage of nurses.
  • The “high self-sufficiency” scenario would require a 95% reduction in international migration. The inflow of migration is progressively cut to 5% of its initial level by 2025. This scenario results in a 15,240 shortage of doctors and a 148,113 shortage of nurses.

Other impact scenarios

These scenarios examine the potential impact of other shocks on the health system.

  • The “high demand” scenario models the impact of an increase in demand of 2% for health services. This scenario results in a 26,124 shortage of doctors and a 193,122 shortage of nurses. Given past health and technology trends, and trends toward subspecialisation in the medical profession, it is certainly conceivable that demand could accelerate in excess of 2%.
  • The “undersupply” scenario assumes that currently the health system is in a state of workforce shortage, rather than using the 2009 position as a “steady state” baseline. Since no reliable estimate of current shortage could be determined, HWA used an assumption of 5%. This scenario results in an 8,389 shortage of doctors and a 193,122 shortage of nurses.
  • The “capped working hours” scenario for doctors only models the effect of reduced working hours. The total number of hours worked were capped at 50 hours per week and the equivalent FTE modelled to show a decrease in supply. This scenario results in a 5,178 shortage of doctors.

Training scenarios

Using the projections estimated in the various scenarios outlined above, HWA calculated the number of graduates which would be required to meet/balance the gaps between supply and demand. For doctors the assumption is an increase in the 2013 intake (with a 5 year lag for course completion) to produce the extra graduates from 2018. For nurses the increase in student numbers occurs in 2013 with graduates entering the workforce in 2016. Given the larger rate of attrition for nursing, student intake figures and graduate figures were employed for each scenario. The clear conclusion is that bridging the gap through training alone will be very difficult given the number of graduates required.

Vocational medical training

HWA has also attempted to model demand for advanced vocational training positions. This modelling assumed current migration patterns and models the flow of medical graduates through three ‘pools’ of training.

  • Pool 1 – PGY1, PGY2, CMOs and those in ‘basic’ specialist training
  • Pool 2 – ‘advanced’ specialist training
  • Pool 3 – Doctors who have obtained fellowship and are active in the labour force

The results show that if the availability of advanced training places were to be kept fixed at the number required for community needs then there will be an increased pool of pre-vocational (pool 1) doctors who are unable to move through training. If, on the other hand, the number of training places is set in accordance with trainee demand, then specialist training will expand at a much greater rate.

Midwives

As noted above, HWA encountered data limitations with midwifery and hence 2025 projections of workforce supply were modelled using three different datasets. This resulted in three different projected gap results:

  • AIHW data only resulting in an anticipated excess of the midwifery workforce of 721.RI
  • 2006 census “hours worked” applied to AIHW data resulting in an anticipated shortage of midwifery workforce of 346.
  • 2006 census “hours worked” resulting in a shortage of midwifery workforce of 2,030.

Data limitations were also encountered in the projection of demand for services which was unable to factor in the demand for non-birth services (including early pregnancy services, miscarriages, etc). HWA also encountered difficulty in distinguishing between midwives, registered nurses practising with qualifications in midwifery and registered nurses working in midwifery without midwifery qualifications. Given these data limitations and the variance in projection, the HWA modelling was inconclusive for workforce planning purposes.

Registered nurses vs enrolled nurses

HWA undertook modelling for the above scenarios (innovation and reform, immigration, other impact and training) separating out enrolled nurses from registered nurses. Most scenarios saw similar projections for both registered and enrolled nurses with the exception of immigration. Varying the international graduates and migration rates is projected to have a greater impact on workforce gaps for registered nurses than on enrolled nurses. Enrolled nursing is not statistically sensitive to changes in immigration rates due to historically low migration rates.

Geographical Location

The report mapped the number of doctors, nurses and midwives across remoteness areas (Australian Standard Geographical Classification Remoteness Area) using AIHW data. Perhaps unsurprisingly, the results showed that while nurses and midwives were more evenly distributed across the nation, doctor ratios were markedly concentrated in major cities and inner regional areas.

Notably however, nurses in regional, rural and remote areas were older, which would lead to predictions of future geographic maldistribution without appropriate workforce strategies. Given the lack of other health professionals in these areas, HW2025 suggests that the nursing workforce in these areas may come under future pressure.

HWA undertook modelling to show the effect of improving distribution by apportioning the expected increase in doctors (the “comparison/do nothing” scenario) across inner regional, outer regional, remote and very remote areas to gain a 10%, 50% or 100% (even population density) improvement in distribution. The results show an improvement in the ratio of doctors to population in all these areas with a decreased doctor ratio in major cities. In the course of this modelling HWA did not seek to determine which policy levers would be most likely to achieve such changes in distribution.

2025 Headcount of medical practitioners
Remoteness Area
Current distribution
10% improvement in distribution
50% improvement in distribution
100% improvement in distribution
Major Cities (RA1)
86,153
84,933
80,055
75,974
Inner Regional (RA2)
13,751
14,526
17,627
21,504
Outer Regional (RA3)
5,760
6,132
7,622
9,484
Remote (RA4)
1,102
1,135
1,268
1,434
Very Remote (RA5)
422
461
616
810
Australia
109,205
109,205
109,205
109,205

Volume 2

HW2025 Volume 2 projects both enrolled nursing and registered nursing by areas of practice as well as modelling supply and demand for each of the scenarios across individual state and territories.

HWA modelled scenarios for both enrolled and registered nursing across five areas of nursing practice (acute care, critical care and emergency, aged care, mental health and other nursing) using reported area of practice within AIHW labour force data. For demand side modelling current utilisation rates for each area of practice were applied against population growth.

This modelling may assist policy makers in understanding which areas of practice will be in shortage for both registered and enrolled nursing and therefore in formulating policy responses. State and territory estimates were projected by using the jurisdictional breakdown in AIHW workforce data. Assumptions included constant proportions over time (that is, no interstate/inter-territory movement of workforce), inclusion of all workforces (public, private and community) and the use of modelling to estimate entry and exit rates.

In many cases there were significant differences to the HWA projections and those calculated by individual states, who are the major employers of the nursing workforce. These have been identified as differences in initial staff numbers, data sources, assumptions made on retirement age, demand and exit rates, lack of re-entrants into state workforces and migration. This illustrates the real complexity in trying to bring some element of data rigour to workforce planning.

Volume 3

HW2025 Volume 3 examines the state of the medical workforce across medical specialties. For each medical specialty the assessment includes an overview of training for Australian graduates, and assessment of the additional training required for Specialist International Medical Graduates (SIMGs), modelling data (workforce stock, demand and inflows), scenario results and a workforce dynamic assessment.

The scenarios modelled were limited to service and workforce reform, registrar work value, medium self-sufficiency and capped working hours rather than the more elaborate modelling in Volumes 1 and 2. The “service and workforce reform” scenario examined interventions that would increase productivity and lower demand for the workforce. The “registrar work value” scenario assigned all registrars a value of 50% in the last 2-3 years of the training therefore challenging the increasingly prevalent assumption that trainees do not supply specialty medical services.

A “workforce dynamics indicator scale” was used to highlight the current status of each specialty that may be of concern in the future using the average age of the existing workforce, the percentage of new fellows versus fellows exiting the workforce, dependence on migrant inflows and length of training.

HWA calculated demand for each specialty from Medicare and hospital utilisation data. Supply data included the Medical Training Review Panel data on the number of training places and graduates. Also included is data regarding SIMGs from the Australian Medical Council and the Department of Immigration and Citizenship.

The analysis showed that while the number of medical specialists is increasing the workforce is not evenly distributed. There is a growing trend towards specialisation and sub-specialisation, and an insufficient number of generalists.

The specialties estimated to be in shortage by 2025 are obstetrics and gynaecology, ophthalmology, anatomical pathology, psychiatry, diagnostics radiology and radiation oncology. The specialties of cardiology, gastroenterology and hepatology and surgical specialties are currently meeting demand for health services however projections estimate there will be more of these specialists coming through the pipeline than needed by 2025.

The analysis also concluded that general practice, psychiatry, ophthalmology, radiology and obstetrics and gynaecology are highly reliant on international medical graduates.

One of HWA’s significant conclusions is that the medical training pathway is both poorly coordinated and lacks appropriate planning and incentives to ensure the effective distribution of numbers between generalists, specialties and sub-specialties and on a geographical basis.

Data limitations meant that scenario modelling and what HWA termed “workforce dynamics assessments” were not able to be provided for addiction medicine, medical administration, occupational and environmental medicine, pain medicine, palliative medicine, public health, rehabilitation medicine, sexual health medicine, sports and exercise medicine and some physician sub-specialties.