Achieving high-quality and cost-effective care for those with chronic disease requires changes in the behaviour of both doctors and patients. In the past, fragmented policy has led to fragmented management of chronic disease, and there is now an opportunity for change. A new payment scheme for the care of people with diabetes, proposed as part of the federal government’s National Health and Hospitals Network, is centred on patients voluntarily enrolling with a practice and general practitioners being paid in a way that changes their behaviour.1 The proposal is worth $449.2 million over 4 years or up to $10 800 annually per practice, and includes a sign-up payment of $1500 per practice, voluntary patient enrolment, capitation payments ($100 per patient) and annual payments of up to $950 per patient linked to “keeping . . . patients healthy and out of hospital”.

Although there is evidence that changing the way doctors are paid can influence their clinical decisions, evidence of how such a change affects patients’ health outcomes and quality of care is scarce. Systematic reviews reveal only a handful of well designed studies that provide reasonable evidence of effects of changes to payment systems on both doctors’ behaviour and clinical outcomes, although recent evaluation of the Service Incentive Payment (SIP) for care of patients with diabetes showed positive outcomes, including an impressive improvement in the appropriate use of glycated haemoglobin (HbA1c) testing.2,3

The first challenge to the supremacy of fee-for-service in general practice remuneration for patient care came with the introduction of the Practice Incentive Program (PIP) in 1998. The PIP has always involved a form of capitation payment and, since 2001, has included “pay for performance” for diabetes and asthma management, cervical screening and (until 2005) mental health care. PIP payments take the form of SIPs and, for treatment of diabetes, are based on completing cycles of care for at least 20% of the practice’s patients with diabetes. So how might these changes in payment affect the way GPs deliver diabetes care and the outcomes achieved? How will the new payments relate to existing programs and will the scheme extend the role of practice nurses?

First, patient registration has the potential to strengthen the relationship between doctors and the populations they serve. There is good evidence that continuity of care improves patient outcomes,4 especially for those with diabetes. However, limiting registration to specific groups of patients is a piecemeal approach and inefficient. Voluntary registration for all chronic diseases that require longitudinal care would seem a better way forward. Other questions remain to be resolved. What is in it for patients? Will they appreciate the benefits of more intensive diabetes care? For GPs, participation is limited to accredited practices. Presumably, those already claiming payment through the cycles of care program for diabetes will participate — but will their behaviour actually change? Will new doctors be drawn into this scheme?

A key concern is the role of capitation payment. Payments based on patient counts could induce general practices to include in the scheme only those patients whose diabetes is already well controlled, and practices with disadvantaged populations, where treatment is more difficult, may be less likely to participate. However, given that the scheme’s capitation payment of $100 per patient is provided in addition to existing fee-for-service and pay for performance, then incentives for opportunistic selection of patients may be less (depending on the relative size of the performance payments). Care must be taken to avoid opportunities for gaming and other unintended consequences.

Second, the scheme’s authors have not detailed the role of pay for performance. The scheme provides an opportunity to pay for outcomes rather than inputs. Paying for improvements in outcomes, rather than for meeting a specified threshold, should be a key element to encourage participation of practices with currently low outcomes of care for diabetes. But how should performance be measured? The cycles of care programs have the advantage of using Medicare data and not adding to GPs’ burden of data reporting. However, the new scheme, by requiring improvements in HbA1c, blood pressure or serum lipid levels to be measured, will rely on practice-level data. The reporting of data by GPs has proved to be feasible through the Australian Primary Care Collaboratives Program and in standard general practice, but often requires further investment in information technology.5

Third, is the $449.2 million new money or a re-use of the diabetes SIP money? If the latter, and if those currently claiming SIPs move over to the new scheme, then behaviour may not change and quality of care may not improve. There are multiple sources of financing, such as the Enhanced Primary Care items for chronic disease and team-care arrangements. Will these be discontinued and funding redirected into this new scheme? Multiple funding sources create red tape and confusion, further limiting the likelihood of behaviour change.

Fourth, team-based care and the role of nurses have been given a boost with $390 million for the direct employment of 4600 practice nurses. Although the numbers of practice nurses have been growing rapidly,6 their roles have been limited to existing Medicare items. Direct salary support will enable practices to use their nurses’ skills flexibly and fully, including skills in diabetes care.7,8 However, nurses should also be given equitable shares of performance pay, if teams are to function effectively.9

Using management of diabetes as a test case for this new model of funding patient care is welcome, but as with all test cases, careful evaluation is required before the model is expanded. Evaluation should be a priority from the time the scheme is introduced. The most important question is whether the new scheme will be able to drive real behavioural change among primary care teams and patients, rather than being just another way of delivering funding to those already doing a good job.