The new “Indigenous health” incentive payment: issues and challenges

Rewarding health services that provide quality care through “pay for performance” is a feature of health systems worldwide, but evidence that these payments can reduce health inequalities is lacking.1,2 The Practice Incentives Program (PIP), through Medicare Australia, finances incentive payments to general practices and Aboriginal community controlled health services (ACCHSs) accredited against the standards of the Royal Australian College of General Practitioners (RACGP). One premise is that if, on top of baseline fee-for-service payments, practices are financially rewarded for providing quality care to subgroups of the population (eg, those underscreened or with chronic disease), then health outcomes might improve.

To date, there is little evidence that the PIP has improved health outcomes for the Aboriginal and Torres Strait Islander population (hereafter referred to as the “Aboriginal” population). For example, the Senate Community Affairs Committee could not ascertain whether incentives for cervical cancer prevention had improved Pap smear coverage in Aboriginal women,3 and “very poor uptake” of incentives relating to asthma in the Aboriginal population suggests that the PIP has had little effect on disease management in this context.4,5

In November 2008, the Council of Australian Governments announced a new PIP to encourage better care, particularly from mainstream general practices, of Aboriginal people with chronic disease, and to help “close the gap” by reducing Aboriginal health disparity. The PIP “Indigenous health” incentive is due to commence in May 2010. Funding of $28 million over 4 years6 will provide incentive payments to general practices and ACCHSs caring for Aboriginal patients, and benefits to registered patients with chronic disease who identify as Aboriginal (Box 1).7

This measure aims to (i) mitigate health inequalities by rewarding practices for the additional work required to meet the health needs of Aboriginal peoples; and (ii) encourage practices to better identify Aboriginal patients, and encourage patients to more readily declare their Aboriginality in exchange for better care. However, this measure was not conceived through consultation with Aboriginal community bodies.

As we move towards greater use of blended payment financing schemes for primary health care, identified in recent health reform bids,8,9 this new health incentive will test assumptions that this type of financing can foster equitable and flexible care for those most in need. We identify here some of the contentious issues surrounding the implementation and evaluation of this measure.

There are two eligibility issues. First, although this health incentive should benefit ACCHSs because of their high Aboriginal patient load,10 we estimated that, as of August 2009, a third of the 150 ACCHSs are not accredited against the RACGP standards. This is an improvement; in 2005–06, only 41% of Aboriginal medical services (58/140) with a general practitioner were accredited against the RACGP standards.11 However, a substantial number of ACCHSs will continue to be ineligible for the incentive, as they lack GPs and the necessary capital works and other infrastructure supports required to achieve accreditation. “Locking out” any Aboriginal health service undermines the integrity of the new measure, especially when unaccredited practices are often in regions with higher proportions of Aboriginal people.11

Second, the health incentive would be more likely to be effective if mainstream general practices were “accredited” to receive payments by undertaking “cultural safety training” endorsed by the ACCHSs sector.12 Recent health reform proposals8 and Aboriginal leaders13 have made similar calls for such credentialling of health services and doctors.

The new health incentive is based on identification of a patient’s ethnic background. The RACGP has developed a guide to assist GPs who have difficulty identifying Aboriginal patients,14 and practices have to show that they are working towards such identification to meet accreditation standards.15 Acceptable evidence for proof of Aboriginality,16,17 based on the report of the Constitutional Section of the Department of Aboriginal Affairs in 1981,18 is given in Box 2.

There are arguments that GPs should not have to identify a patient’s ethnicity, as it is viewed as a social and cultural issue, rather than a biological one,19-21 and there are patients who believe that ethnicity is irrelevant to their care.19 However, closing the health gap for Aboriginal people depends on targeting specific health initiatives to improve their access to health services, and prevent disease and early death.

Aboriginal peoples and/or Torres Strait Islanders make up 80%–90% of attendees at ACCHSs,10 compared with 0.7%–1.6% of those attending private general practices, but the latter proportion is considered an underestimate.22 There is a potential for general practices to make an incorrect identification, or for a patient to incorrectly claim to be Aboriginal, leading to inappropriate targeting of the initiative.

Putting the onus on disadvantaged and unwell clients to provide proof of Aboriginality creates a perverse situation. However, Aboriginal people are often asked to confirm their ethnic origin, and ACCHSs regularly provide this service for clients and the community. The National Aboriginal Community Controlled Health Organisation (NACCHO) recommends that, for Aboriginality to be recognised, all three parts of the definition in Box 2 should be fulfilled.23 However, Centrelink (the Australian Government’s social security agency) and the Voluntary Indigenous Identifier used by Medicare accept self-identification as sufficient evidence, although this is not the case with all government programs.

Adopting self-identification of Aboriginality as the sole criterion for the incentive payment may be justified, if a pilot study could confirm that the risk of “leakage” is small. Mechanisms to prevent and audit false claims by both practices and patients will be important. Potential funds leakage in two directions will mean fewer resources for Aboriginal people with chronic and complex disease. In addition, privacy constraints might preclude the use of the Medicare Voluntary Indigenous Identifier as an alternative mechanism to inform practices or to calculate payments.

The health incentive does not take into account the heterogeneity of Aboriginal populations, “the great variation nationwide in the characteristics, exposures and genotypes of people identifying as Indigenous”.24 Deprivation is not necessarily defined by ethnicity alone. If health outcomes (eg, Medicare Benefits Schedule [MBS] claims) determine payments, health services with patients from impoverished environments may be less likely to reach the specified outcome thresholds than those with more health-literate patients (irrespective of the quality and quantity of care provided). Given that Aboriginal patients attending ACCHSs have more complex disease than those attending private general practices,10 we may see a differential in payment for the level of effort expended. Moreover, some private practices may be more inclined to direct attention to patients with uncomplicated chronic disease, as “needy” patients may interfere with payment thresholds (inverse care law).

Health outcomes need to be selected carefully; health improvements are influenced as much by social and environmental influences as by health care. Mediating social change and supporting communities is but one facet of the broad activity of ACCHSs.11 It is worth considering whether health services providing care to more complex and “hard to reach” Aboriginal patients could be eligible for outcomes payments at different thresholds to other services.

The Australian Government will pay practices to register Aboriginal clients, but “the maintenance of Aboriginal relationships to places, and to country, and ... the maintenance of social relationships” is dependent on mobility.25 Voluntary enrolment of patients with complex needs with a single health service of their choice (their “health care home”), with grant funding tied to levels of enrolment, has been proposed by the National Health and Hospitals Reform Commission.8

As Aboriginal patients travel between services, often visiting clinics only once, competition for patient registration and outcomes payments (if directed only to registrant services) may not support best practice. Some practices may be discouraged from caring for highly mobile Aboriginal patients, while other practices would be driven by payments to “poach” Aboriginal patients. Rorting of MBS payments has been reported: unscrupulous practices claim for health assessments or care plans with no intention of providing follow-up care for anyone who is not one of their usual patients. The patients’ usual practices subsequently have their MBS claims rejected.26

This situation could be prevented if patients had to declare that the registering practice was their usual care provider and agree to the transfer of their medical records to fulfil the requirements of the health incentive. There should be a complaints mechanism to monitor any inappropriate registration of Aboriginal clients.

ACCHSs, as non-profit organisations, return any funds generated to the Aboriginal community. For private general practices, participation in practice incentive programs is largely a business decision. The current PIP makes up only a very small proportion of general practice income,27 and even if revenue from the new PIP health incentive within mainstream services could be appropriated for Aboriginal programs, the quantum would make little difference to Aboriginal community care.

The current PIP absorbs significant administration costs, which make up 39% of the total government outlay on this initiative.27 Thus, assuming similar levels, about $11 million over 4 years might be spent in administering the new PIP health incentive. As this expenditure would not make its way back to Aboriginal community programs, we would define this as “leakage” at a program level. Moreover, the incentive may not offset the possibly substantial administrative burden on ACCHSs (by virtue of their Aboriginal patient load).

There is also the potential for the new health incentive to fuel discriminatory attitudes within Australian society to the detriment of Aboriginal peoples, as, within private general practices, ethnicity-based benefits and financial relief to Aboriginal patients would be denied to other patients.

The government expects that around 4500 practices per year will sign up for the health incentive payment,6 and 55% of the adult Aboriginal and Torres Strait Islander population will receive a health assessment over the next 5 years,28 with treatment of illnesses delivered mainly through the mainstream health system.10 Will these expectations be met?

Given that there are currently 4798 practices participating in the PIP,29 and 70% of private practices do not see a single Aboriginal client,10 it is unlikely that the expected number of practices will sign up for the new health incentive.

In 2008–09, 6.5% of all Aboriginal or Torres Strait Islander adults (15–54 years) had a health check for which an MBS rebate was claimed (the rate is calculated based on 2006 census population estimates for the 15–54-year-old Indigenous Australian population [282755]).30 This rate is much less than the 21% of older Australians (>75 years) who received a health check in 2006–07.11 However, of 4386 Aboriginal and/or Torres Strait Islander patient encounters within private general practices over the period 2004–2008, there was not a single encounter for which any of the MBS rebates for Aboriginal health assessments were recorded.31 Of the general practices in the Australian Capital Territory Division of General Practice surveyed in 2006–07, few knew of these rebates and none had claimed for them,32 suggesting attribution of most Aboriginal health assessments to the single ACCHS in the ACT. Many ACCHSs also undertake health assessments without claiming the Medicare rebate.33

While Divisions of General Practice have been unable to influence general practice performance, as measured by claims for cervical, asthma and diabetes service incentive payments (2000–2005),34 Divisions will be funded for up to 80 project officers to assist practices with the new health incentive.35

To estimate any gains from the new initiative, an evaluation must be based on outcomes disaggregated by type of service provider. The 2008–09 National Performance Indicators for Divisions require them to report on the number of MBS health checks for Aboriginal people (Items 710 and 708) provided “by GPs within the Division”. However, the indicators’ technical specification does not differentiate health checks undertaken by ACCHSs from those done by general practices,36 as Medicare Australia does not capture data by type of service provider and is unable to extract data by using practice names.

The proposed outcome measures for the new PIP health incentive are similar to those of the current PIP: the number of PIP practices signed on to the incentive; the number of practices receiving payments for registering patients; the number of Indigenous clients registered with a PIP practice for chronic disease management; and the number of Aboriginal and/or Torres Strait Islander people receiving an MBS adult health check.35

Registration and outcomes data could be disaggregated by service type, if this was supported. However, proposed sources of health assessment data, such as from sentinel sites35 or Healthy for Life services (an Australian Government program to improve the health of Aboriginal mothers and children), of which 70% are ACCHSs, could introduce selection bias. New performance indicators for Divisions of General Practice are needed to measure their impact on Aboriginal health outcomes.

The “Indigenous health” incentive may reduce Aboriginal health disparities caused by poor access to mainstream health services, but “closing this gap” requires much more than improvements to clinical care and increased MBS claims. Comprehensive primary health care, fostering Aboriginal community governance and action, which is the function of ACCHSs, is critical to all “close the gap” measures. However, funding for the new incentive will be channelled largely through general practices. This means that careful attribution of outcomes from these services will be vital to assess the effectiveness (or otherwise) of blended payment financing to mitigate health inequalities.

Certain assumptions underpin the rationale for this health incentive:

• that it will increase Aboriginal people’s attendances for general practice care;

• that the increased attendances will be from Aboriginal people with chronic and complex diseases who are not currently accessing the health system; and

• that each consultation will deliver better care.

If these assumptions are not fulfilled, the incentive will merely fund practices to coordinate the care of Aboriginal people who are already regular users of the health system (thereby diminishing returns), and make little difference to those Aboriginal people this Council of Australian Governments measure is meant to assist.