The concept of DWS was introduced in 2001 to support the workforce distribution aims of s. 19AB of the Act. As outlined in Box 6.1, OTDs and FGAMS need to apply for an exemption under s. 19AB(3) of the Act in order to provide services that attract a Medicare rebate.
DWS determinations are used to support a range of workforce programs and initiatives, including consequential classification systems employed by state and territory jurisdictions.
The current operation of DWS has been a source of concern for many stakeholders. This is reflected in the Lost in the Labyrinth report which found that there was a lack of transparency in the way that DWS status was determined, that quarterly updates to DWS classifications have produced uncertainty for service providers and OTDs, and that there was significant confusion regarding the interaction of DWS and the state-based Area of Need determinations (see Box 6.3).
With these concerns in mind, the calculation and operation of the DWS system has been considered as part of this review. On 9 November 2012 a working group of key stakeholders was convened to workshop options for reform (referred to henceforth as the DWS working group). In addition to the Department, the following organisations were represented at this meeting:
- The Rural Doctors Association of Australia;
- The Australian Medical Association;
- The Royal Australian College of General Practitioners;
- Rural Health Workforce Australia;
- The Medical Board of Australia/AHPRA;
- The Committee of Presidents of Medical Colleges; and
- The Area of Need Unit of the Western Australian Government.
While there was lack of unanimity on several issues, the majority of participants agreed that there was a continued need for a DWS-type system for identifying areas of Australia that experience the most acute unmet needs for non-government/private medical services, and that the system needs to be evidence-based to avoid the potential for distortions created by a discretionary system.
The most significant concerns that have been raised in relation to the current system, including those discussed at the DWS working group meeting, relate to the process for making DWS determinations for general practice (although in the course of the review generally, particular concerns were raised about the operation of DWS as it pertains to some specialties). The following proposal therefore places a specific focus on the systemic requirements for producing accurate determinations of workforce shortage for general practice.
It is also timely to consider an updated methodology for DWS given that the Australian Bureau of Statistics (ABS) is transitioning to the Australian Statistical Geography Standard (ASGS) in 2013, as noted in Chapter 4. This transition will render the geographic boundaries utilised in the current DWS methodology obsolete.
The methodology proposed in this review for an updated DWS system is outlined in table 6.1 below, followed by a rationale for the core elements.
|Feature||Current system||Proposed System|
|Population data||2004 Census data||2011 Census data|
|Update frequency||General Practice: Quarterly||Annual (all specialties)|
|Other specialties: Annually||Annual (all specialties)|
|Area classification boundaries||General Practice: 2004 Statistical Local Area Boundaries||General Practice: 2011 Statistical Area 2 (SA2) Boundaries under ASGS|
|Other Specialties: 2004 Statistical Sub- District Boundaries||Other Specialties: 2011 Statistical Area 3 (SA3) Boundaries under ASGS|
|Interaction with Remoteness Area classifications||General Practice: No interaction||Stage 1: ASGS-RA3, RA4 and RA5 areas are DWS for other specialities.
Stage 2: ASGS-RA2 and RA3 with populations under 15,000, along with RA4 and RA5, are DWS for all specialties (including general practice).
|Other specialties: ASGC-RA3, RA4 and RA5 areas are DWS||Stage 1: ASGS-RA3, RA4 and RA5 areas are DWS for other specialities.
Stage 2: ASGS-RA2 and RA3 with populations under 15,000, along with RA4 and RA5, are DWS for all specialties (including general practice).
|Interaction with metropolitan area classifications||General Practice: Inner metropolitan areas are automatically classified as non-DWS||Removes interaction with metropolitan area classifications.|
|Other specialties: No interaction||Removes interaction with metropolitan area classifications.|
|General practice workforce measure||Comparison of Population to FTE ratio of each SLA against national average||Three step approach:
|Other specialist workforce measure||Compare FTE specialists per 100,000 persons within SSD to national average||Compare FTE specialists per 100,000 persons within SA3 to national average|
Frequency of updates
The proposed model includes annual updating of the DWS classification for each of the medical specialties, including general practice. This represents a departure from the current system where DWS classifications for general practice are updated each quarter.
A significant concern expressed by stakeholders about the current system is that quarterly updates present a logistical challenge when planning general practice recruitment activities. A system of annual updates to DWS status for general practice, coupled with other proposed reforms, should alleviate this concern.141
While most stakeholders support removal of the quarterly update system, some concern has been expressed that moving to a system of annual DWS updates for general practice will remove capacity to produce a timely measure of unmet medical service needs. While these concerns are recognised the new methodology for determining DWS for general practice should present more accurate workforce shortage classifications, thereby reducing a need for quarterly updates to ensure sensitivity of classifications.
As any DWS system relies on an analysis of the level of Medicare-rebated service provision within a defined area, the Department is required to consider how to appropriately define areas for the purpose of a revised DWS system. Making an appropriate determination of an area requires the Department to consider three core issues:
- the interaction between the proposed revised DWS system and remoteness area classifications;
- the need for updated geographic boundaries for DWS classifications; and
- the interaction between the proposed revised DWS system and “metropolitan area” classifications.
Interaction with remoteness area classifications
Under the current DWS system, all areas that are classified under ASGC-RA as outer regional (RA3), remote (RA4) and very remote (RA5) are considered to meet the requirements of DWS for all medical specialties other than general practice.
The use of automatic classifications for certain areas as DWS, based on level of remoteness, has the benefits of efficiency for the administering department and enhanced certainty and stability for those communities and the health workforces within them. There is therefore a strong argument for the introduction of automatic DWS classifications, based on remoteness areas, for general practice.
One option is to extend the current arrangements for specialties (i.e. DWS status for all locations classified as RA3 to RA5) to general practice (noting this would be based on the new ASGS methodology). However, this may provide DWS status for general practice to larger regional centres, such as Cairns and Townsville, which might not otherwise meet the criteria for DWS, and arguably, do not experience the same level of difficulty in attracting medical practitioners as smaller, more remote areas.
Some stakeholders have also presented opposition to this option, on the basis that such an approach would not account for the unique models of medical service provision that have been established within individual regional, rural or remote communities. In particular, it has been argued that the proposed change may encourage corporate practices, often staffed by OTDs, to operate in towns where this is not currently feasible due to the inability to attract sufficient medical professionals. Stakeholders have raised concerns that this could lead to corporate practices “cherry picking” general practice work, leaving after-hours and on call work to already hard pressed local practitioners.
An alternative may be to draw upon the amended geographical classification system proposed in Chapter 4 (the ‘modified Monash model’), which would allow for automatic DWS status to be better targeted towards regional and rural areas with smaller populations. Using the ‘modified Monash model’, the following areas could be classified as DWS:
- RA2 and RA3 areas with populations less than 15,000;
- RA4; and
This categorisation would address the relative disadvantage of small towns that was identified in Chapter 4. It is also likely to reduce the potential for corporate “cherry picking” discussed above.
It should be noted that any classification system will have its disadvantages. The DWS system is simply one tool designed to produce classifications based on the level of access to Medicare-subsidised services within an area. No tool has been devised which would produce classifications according to a preferred or desired model of medical service provision, which need to be achieved using other policy levers.
In revising the model, it is proposed that the application of automatic DWS status for other medical specialties be aligned with that of general practice, in order to simplify the system. While the implementation of the model outlined above will result in some RA3 areas (those with populations above 15,000) no longer having automatic DWS status for other specialties, the impact of this is likely to be minimal, given the overall shortage of specialists in non-urban areas. It is also noted that those areas experiencing a genuine shortage will be assessed as DWS according to the methodology outlined in Table 6.1.
The classification of some regional, and all remote and very remote areas as DWS for general practice and other specialties will:
- provide equality of treatment for regional, rural and remote communities in terms of DWS classifications;
- provide follow-on efficiencies for workforce programs that use DWS classifications, such as the Rural Locum Relief Program and the AoN Program; and
- provide DWS status to all rural and remote public hospitals and other public health services that have been granted an exemption to Section 19(2) of the Act, which allows them to offer some primary care services that are eligible for Medicare rebates.142 DWS status would ensure that these services may effectively access the intended provisions under the s19(2) determination.
There is no question that there will need to be significant development work undertaken to implement a new rural classification methodology under the modified Monash model, and the administrative complexities in applying this to the DWS system should not be underestimated. While supported in a theoretical sense, the practicalities of applying automatic DWS status for areas in RA2 and RA3 with populations under 15,000, and for RA4 and RA5 under the new ASGS will need to be considered in the light of the implementation arrangements for the modified Monash model.
Updated geographic boundaries
As stated above, the current SLA and SSD boundaries that are used for the purpose of DWS classifications under the current system will become obsolete upon the transition to the ASGS in 2013. This creates a need to identify appropriate geographic boundaries for the purpose of establishing DWS classifications for general practice and the other medical specialties within areas that have a major city (RA1) or inner regional (RA2) remoteness area classification.
It is proposed that the revised DWS system would use Statistical Area (SA) 2 boundaries when determining DWS for general practice and SA3 boundaries when determining DWS for the other medical specialties.143 This approach to establishing updated geographic boundaries is suggested for the following reasons:
- SA2 areas are similar in size to the SLAs that are used when determining DWS for general practice;
- SA3 areas as similar in size to the SSDs that are used when determining DWS for the other medical specialties; and
- there is no evidence to suggest that the size of current SLA or SSD boundaries produces either unintended or inequitable outcomes when applying the DWS process.
Interaction with “metropolitan area” classifications
If an area is identified as an inner metropolitan area it is currently automatically designated as non-DWS for general practice. The rationale for this has been that inner metropolitan areas have better access to a broad range of public and allied health services when compared to outer metropolitan, regional, rural and remote areas. These services are not captured within Medicare billing statistics used to determine DWS as they are delivered by salaried health professionals. The non-DWS classification for general practice in inner metropolitan areas therefore went some way towards creating parity with the level of medical service provision within other areas of Australia.
This was a policy decision that was implemented primarily to ensure equitable determinations under the previous Preliminary Assessment of District of Workforce Shortage (PADWS) application process. A designation of ‘inner metropolitan’ meant that medical practices were restricted from accessing the special circumstances provisions available under the PADWS to employ an OTD.
It is proposed to discontinue the practice of basing DWS classifications for general practice on “metropolitan area” classifications as a feature of the proposed new system. This change is proposed for three reasons:
- the workforce measure proposed to support the modified system is robust and is likely to identify where there is a genuine general practice workforce shortage within areas that hold an inner metropolitan classification;
- “metropolitan area” classifications, as currently defined and operationalised, are based in part on population data circa 1992 and are therefore out-dated; and
- the value of continuing to consider metropolitan classifications as part of the process of providing DWS determinations has diminished after the discontinuation of the PADWS application process on 28 September 2012.
The proposed refined DWS system therefore abandons the use of a blanket non-DWS classification for metropolitan areas. This change will ensure that general practice workforce shortage needs are identified and appropriately classified against a national average measure, no matter where they are located.
The transition to the use of the ASGS will provide an updated analysis of the composition/dispersion of the Australian population based on 2011 census data.
As the revised system will be based on the ASGS system, the population data that is used to inform the Medicare billing statistics and DWS classifications will be updated accordingly. This proposal does not seek to make any additional changes to the population data that would be used for DWS classifications beyond the updates that will be achieved upon the transition to the ASGS.
Proposed workforce measures
A mechanism for measuring the composition of the medical workforce will continue to be a core requirement of any revised DWS system. As with the current system, this proposal is based on two workforce measures:
- a measure of the general practitioner workforce; and
- a measure of the specialist medical workforce.
The possibility of including some measure of local hospital staffing within the methodology for a revised DWS was discussed during the DWS working group meeting, as the level of access to primary care delivered through smaller local hospitals will often impact upon the workload of local medical practitioners.
However, it is proposed that the workforce measures should continue to be based on an analysis of medical services that are subsidised by a Medicare rebate within the relevant specialty, given the complexity involved in developing and maintaining meaningful and up to date measures of (state-based) hospital staffing in a multitude of locations, and the potential for creating perverse incentives. Medicare billing statistics, while often criticised, are at least a quantifiable source of information in relation to medical service provision within a geographic area.
It is also noted that DWS classifications do not directly affect employment of medical practitioners within positions that do not require access to a Medicare provider number.
General practice workforce measure
The current DWS classification system produces classifications that are based on a full-time equivalent (FTE) measure of the portion of the medical workforce who are providing Medicare-rebated medical services. The FTE measure is applied to all medical specialties for two reasons:
- a single medical practitioner cannot be counted more than once within a local area; and
- the partial contributions of medical practitioners who practise on a less than full-time basis within a local area may be accurately accounted for.
Concerns have been raised in regard to the efficacy of continuing to base DWS solely on an FTE measure of the medical workforce who provide Medicare-rebated medical services. These questions have been focused on the fact that the use of an FTE measure, as derived from Medicare billing statistics, does not provide a means to consider the actual workload of individual medical practitioners within a local area. In particular, it does not account for situations where a medical practitioner is working extended hours (beyond a standard full-time load). This issue has been raised with a particular focus on the specialty of general practice.
It is proposed that the revised system would be based on a three step process for measuring the general practice medical workforce within areas that are classified as major cities (RA1) or inner regional (RA2), as outlined in Table 6.2.
|Step 1:||Compare the Medicare billing statistics local area with the national average
|Step 2:||Apply a 10% buffer to the billing statistics
|Step 3:||Apply FWE-to-GP ratio
There are two key differences in the revised process. The first is the application of a buffer in the comparison of an area’s access to GPs with the national average. Approximately 30% of all areas (SLAs) have GP access rates that fall within 10% of the national average. Using a 10% buffer would mean these areas are identified with each annual update and are subject to additional scrutiny before being classified as a DWS or non-DWS area. By applying this buffer, an area will not be classified as non-DWS solely due to having marginally better access to services than the national average.
Any area that has a better FTE-to-GP ratio than the national average and that does not fall within the buffer zone is classified by the Department as a non-DWS area. These areas have substantially better access to Medicare-rebated general practice services when compared to the national average.
The second key difference in the revised process is the application of a full-time workload equivalent (FWE)-to-GP ratio to those areas falling within the 10% buffer zone. This will identify those local areas that would otherwise be classified as non-DWS for general practice as a result of local GPs providing substantially more than full-time services.
If areas falling within the 10% buffer zone have a FWE-to-GP ratio greater than 1.3, the area will be classified as DWS. This value has been identified as a cut-off because it:
- means that each GP within the area is on average providing up to a third more than the recognised full-time equivalent level of medical services (i.e. substantially more services);
- is unlikely to be produced by seasonal changes that affect some local areas; and
- is unlikely to be negated by the presence of short-term locum medical practitioners.
The proposed methodology continues to use a comparison with the national average as a basis for producing classifications. Most stakeholders represented at the DWS working group were supportive of the continued use of a national average measure as part of the process for identifying relative medical workforce shortage. This methodology uses this basis and a buffer to effectively identify marginal areas, and examines Medicare billing behaviour of general practitioners who practise privately within these areas.
The Department currently collects all necessary Medicare billing information to complete this proposed workforce measure.
The specialist medical workforce
As described in Box 6.2, DWS for each of the medical specialties is currently determined according to the use of ABS population data and the Medicare billing statistics. A FTE measure of the specialist medical workforce is used.
Any medical specialty that has a national average of less than three FTE specialists per 100,000 persons is considered to be in acute shortage. All areas of Australia are considered to be DWS for specialties that are considered to be within acute shortage, with the objective being to increase the number of specialist medical practitioners practising privately within this specialty.
It is proposed that the revised DWS classification system continue to adopt this measure for the specialist medical workforce. This measure continues to be appropriate when considering the relatively small numbers of practitioners practising privately within each specialty compared to general practitioners.
The majority of concerns relating to DWS classifications for specialties other than general practice are based on an argument that such classifications are only appropriate when considering general practice. It is not proposed to remove DWS classifications from any medical specialty as the acute shortage provisions continue to ensure that DWS does not adversely affect those specialties that have small numbers of doctors.
Impacts of revised general practice workforce measure
It is anticipated that the revised methodology for determining DWS will result in a higher number of areas being declared DWS for general practice. Prior to the introduction of the ASGS and obtaining updated population data and statistical area boundaries, it is impossible to determine the exact nature of the overall changes to the number of DWS areas.
However, the proposal to classify some regional, and all remote and very remote areas as DWS is likely to increase the overall number of DWS areas. It is also anticipated that there will be a slight increase in the number of DWS areas located within areas that would be classified as outer metropolitan due to the proposed use of a buffer when determining DWS.
The key features of this methodology appear to respond to the major concerns raised by external stakeholders. While it is unlikely that the proposal will win universal acceptance, it is anticipated that the proposed methodology will produce robust workforce shortage determinations and therefore achieve acceptance amongst stakeholders.
Further policy considerations
Future system name
Several external stakeholders, including a majority of the representatives of the DWS working group, have expressed the view that the term “District of Workforce Shortage” is itself unhelpful and in some cases may be actively misleading. The term DWS does not provide the best reflection of either:
- the information (specifically Medicare billing statistics) taken into account when making DWS determinations; or
- the statistic modelling that governs the process for making DWS determinations.
Based on these concerns, and given the scope of the proposed reform of the scheme, consideration could be given to renaming the revised system. For example, the term “targeted workforce areas” could be a suitable title for a relaunched scheme as it more accurately reflects the intent of the scheme.144 Other options could reflect the Medicare data basis of the classification, for example, ‘Lower Medicare Access Areas’.
While there are clearly some identifiable benefits to renaming the revised system, these need to be weighed against the flow on administrative costs of reconfiguring programs that currently utilise the DWS classification, for example, the Bonded Medical Places (BMP) scheme.
While there are no direct program costs resulting from the proposed DWS changes, there may be flow-on impacts on the Medicare Benefits Scheme. The model proposed above is likely to increase the number of areas classified as DWS for general practice, meaning that some practices that have struggled to fill positions with Australian GPs will be able to employ OTDs subject to the ‘ten year moratorium’, leading to a higher rate of billable service provision. There may be an increase in Medicare expenditure if the revised DWS scheme increases the attractiveness of some locations within Australia as a work destination for OTDs.
Impacts on distribution
There is a potential risk that if the proposed system increases the number of areas classified as DWS in metropolitan or regional areas, this could exacerbate existing difficulties experienced in the most remote areas in recruiting medical practitioners. However, the increasing numbers of participants in the BMP Scheme attaining fellowship and becoming eligible to complete their return of service obligation in DWS areas may offset this. It is also the case that the practice of locating newly recruited OTDs in the most remote areas of Australia has been strongly criticised, including in submissions to this review. The impact of the revised system on workforce distribution will need to be carefully tracked.
Timing of transition to proposed new system
If approved, it is suggested that an appropriate time for the rollout of the first stage of this proposal would be following transition to the ASGS. The ABS data to support the ASGS was due to be provided to the Department in early 2013 with a more formal date for the transition to be advised. The possible implementation of a second stage, involving the “modified Monash model” of geographical classification proposed in Chapter 4, is dependent upon the arrangements for that system, which is likely to take at least 18 months, given the need for changes to information technology and other infrastructure.
Grand parenting arrangements
There is a potential that the introduction of a revised methodology for DWS may result in some locations losing DWS status. The Department may therefore need to consider a period of grand parenting arrangements to ensure that local communities are not subject to perceived disadvantage upon the transition to the revised classification system.
|Recommendation number||Recommendation||Affected programs||Timeframe|
|Recommendation 6.7||The Commonwealth should introduce a revised system to replace the current districts of workforce shortage (DWS) classification system. It should be introduced in 2 stages.|
Under the first stage, the geographic classification requirements of the revised system should be based on the Australian Statistical Geography Standard (ASGS), these requirements being:
The revised system should abandon the use of the additional overlay of the “metropolitan areas classification system” for general practice.
The revised system should use a modified general practice workforce measure within major cities (RA1) and inner regional areas (RA2) comprised of:
If the “modified Monash model” of geographical classification is implemented and its methodology can be applied to DWS in an administratively efficient manner, the following areas should be granted automatic DWS status for both general practice and other specialties:
An implementation working group should be established.
|DWS, allocation of MBS provider numbers, BMP||Short term – transitional arrangements, further discussions with stakeholders and the development of communication and data system should commence immediately post-Review.|
Medium term – it is likely that the work outlined above will necessitate medium term implementation of the full DWS changes.
141 The revised Health Insurance (Section 19AB Exemptions) Guidelines 2012 have gone some way towards addressing these concerns by allowing for the delegate to consider the DWS status of an area at the time recruitment activity commences when making a decision on an application for a s. 19AB exemption.
142 Under Section 19(2), Medicare benefits are not payable for public hospital and other public health services unless the Minister otherwise directs. The COAG Improving Access to Primary Care Services in Rural and Remote areas – s. 19(2) Exemptions Initiative was introduced in 2006 and allows Medicare benefits to be claimed for some state-remunerated professional non-admitted services (including eligible nursing and midwifery services) and eligible allied health and dental services provided in emergency departments, outpatient and community clinics at approved rural and remote public hospitals. Queensland, Northern Territory, Western Australia and New South Wales currently participate in the initiative. To be eligible, a locality must have a population of less than 7,000 people, not be in a major city, and be in an area of workforce shortage.
143 The ASGS is structured with six hierarchical levels of geographic region. The levels are based on population size, and each level directly aggregates to the level above:
- Mesh block: the smallest geographic region captured by AGSS, made up of 30 to 60 dwellings (347,000 units covering the whole of Australia);
- Statistical Area Level 1 (SA1): areas with populations in the range of 200 to 800 persons (54,805 regions);
- SA2: areas with populations in the range of 3,000 to 25,000 persons (2,214 regions);
- SA3: areas with populations in the range of 30,000 to 130,000 persons (351 regions);
- SA4: areas with populations in the range of 100,000 to 500,000 persons (106 regions);
- States and territories.
For further information, see Australian Bureau of Statistics http://www.abs.gov.au/ausstats/abs@.nsf/mf/1270.0.55.001
144 Note that this term has previously been discussed with key stakeholders in relation to an alternative DWS proposal.