In their Joint Statement of 5 April this year, the Federal and State Health Ministers acknowledged a widely recognised but rarely voiced truth — that past negotiations under the Australian Health Care Agreements (AHCAs) had focused more on health funding than health outcomes. The Ministers' candor took many completely by surprise.

They laid a framework for work toward the 2003–2008 agreements, which emphasised a focus on provision of best care and health outcomes rather than jurisdictional boundaries, with jurisdictions working cooperatively to advance community health and well-being. To this end, they decided the agreements should contain a statement of principles, objectives and proposed outcomes. They also decided that work should be initially organised around a continuum across preventive, primary, chronic and acute care; improvement of the interface between aged and acute care; cross-jurisdictional collaboration on workforce, training and education; the interaction between hospital funding and private health insurance; improvements in Indigenous, mental, and rural health; quality and safety; information technology; research; and "e-health". This work is to precede and inform negotiations about funding within the AHCA.

In that same week in April the Commonwealth Chief Medical Officer, Richard Smallwood, told a health conference in London that Australia's public hospitals are in "varying degrees of dilapidation", and that morale among doctors and nurses was fragile. He was quoted as saying:

"The results of our care and patient experiences of the health care system are too often less than ideal, . . . Our public healthcare systems never seem to have enough resources . . . Access to care, while universal, is too often delayed. The medical workforce is undermanned, maldistributed, or both, and the shortage of nurses verges on the calamitous. In both professions, morale is fragile."1

The indications noted by our Chief Medical Officer are uncontroversial, and could easily be extended. Do they reflect separate causes, capable of treatment seriatim, or are they symptoms of a deeper malaise? The work proposed in the Ministers' Statement has been allocated to nine separate working groups, implying at least some separability.

An underlying ailment seems more likely — this is not the first time Health Ministers have committed themselves to "outcomes", yet the problems besetting the system have not diminished, or even changed much. There has to be more to this than meets the eye.

Across countries in the Organisation for Economic Cooperation and Development (OECD) — that is, the rich world:

• many doctors are disgruntled, overworked and professionally unsatisfied;

• nurses are restless and in short supply;

• richer countries poach healthcare workers from wherever they can get them;

• popular demands for more health spending are universal;

• payers are widely unpopular, whether national governments or United States Health Maintenance Organisations (HMOs);

• many patients feel vulnerable and uncared for; and

• the politics of healthcare is ugly — governments retreat into obfuscation and the difficult issues are systematically avoided.

The explanation can only lie in common and fundamental characteristics of system architecture. Structural questions are not explicitly on the Health Ministers' agenda. They need to be. Opinions as to cause will differ, but here are mine.

A burgeoning international literature on "unhappy" doctors and nurses offers all manner of causal explanations, and proposals for remediation. This literature usually presupposes that the causes lie in the specifics of healthcare.2-4 However, this seems unlikely. Commoditisation, which results from payment of a standard unit price for an implicitly standardised product or service, is a more probable cause. Commoditisation of the product of any industry leads to commoditisation of its labor force, and that always causes "unhappiness", and worse. When, as in healthcare, that labor force is bright, individualistic, highly educated, conscientious and oriented toward professional autonomy, reducing the fruits of their efforts to item numbers and "one price fits all" is clearly counter-indicated.

Australian general practice represents commoditisation par excellence. The squirrel-wheel reimbursement system and narrow tasking imposed on GPs foster personal and professional isolation and disillusionment.

GP work content has narrowed steadily over the years. Opinions differ as to whether the number of GPs is too many, as the Australian Medical Workforce Advisory Committee thought,5 or too few, as Access Economics thinks.6 Regardless of the findings of such studies, the elaborate system used to control doctor numbers in Australia means there are unquestionably too few GPs at prevailing "prices". The absence of unemployment among GPs, and the difficulties across the nation in filling the less desirable posts, puts the matter beyond doubt. Numbers are an important issue in other ways, but, regardless of numbers, general practice under Medicare is purpose-built to create discontent.

A narrow majority of doctors and the substantial majority of nurses work in hospitals rather than in general practice and other forms of primary care.7,8 Public hospitals are kept alive on a lean diet of funding by governments, and are under intense pressure to maximise patient throughput. Governments are obsessively interested in hospital-throughput statistics, and in minimising indicators of excess demand, such as waiting lists and the frequency of ambulance bypass. These make headlines — quality of care and patient satisfaction does not. Facing an institutionalised payer, hospital managers respond by institutionalising working arrangements, and directing their efforts to the performance targets by which they are measured.

Input controls and statistical output targets provide little defense against horrors such as those that emerged at the King Edward Memorial Hospital in Perth.9 However, these were no more than a minor departure from mainstream public hospital practice in professional deployment and quality management practice across Australia. Institutionalised and depersonalised working arrangements are the absolute enemies of professional satisfaction, high morale, and respectful, high quality care. It is not surprising that many doctors and nurses are unhappy.

The surprising thing is that Australian healthcare, operating under such compromised arrangements, manages to be as good as it is. Good people can beat bad systems, at least for a time, but at enormous cost to themselves. Many are burned in the process. Institutionalised production systems destroy leadership talent and break the hearts of professionals of skill and integrity. Vital, customer-responsive organisations breed professional talent, active, committed leadership and display the high professional morale that, for most good people, is its own reward.

A remote third-party payer is in itself a guarantee of alienation and quality problems. Australia did away with agricultural marketing boards in the 1980s, with good reason, and to impressive effect. The Western world largely abandoned nationalised industries in the 1980s and 1990s. The Union of Soviet Socialist Republics (USSR) never managed to produce decent consumer goods. Healthcare is the ultimate individualised product — no centrally planned and controlled production system ever succeeded at mass-customisation.

Third-party payer systems are bad for the payer as well as the consumer. Western governments, other than that of the United States, meet 72% to 97% of all health expenditures, and this accounts for 12% to 18% of all public outlays.10 A blowout in health spending means a blowout in the Budget. Regardless of the feelings of Health Ministers, the overall interest of governments in healthcare centres on expenditure control, not health outcomes. The Health Minister's duty to the government is to protect the Budget and hose down the political consequences of doing so, not to open up difficult questions.

Whenever the compact between the customer and the producer is nullified, there are dire consequences. Large, bureaucratised, third-party payers are a universal feature of modern healthcare the world over, and the consequential symptoms of malaise are also ubiquitous.

None of this goes against the aims of universality and equality of access to healthcare, which all civilised nations embrace. However, the payer must stand behind the patient, and not between the patient and the provider. The starting point on the way back must lie in the realm of system design, not clinical reform — the professions will see to the clinical side if structures are right.

Remedies for the immediate roadblocks on the path to outcome-oriented healthcare lie not in the clinical realm, but in general properties of constrained optimisation. Instead of producing services in response to demand, the system is controlled by layer upon layer of regulatory constraints, operating both on aggregate-level inputs (dollars and doctor numbers) and product-level outputs (item numbers, descriptions and reimbursements) as follows:

• Budget caps constrain public hospital spending and most of the smaller healthcare and aged-care programs.

• Non-transparent restrictions on the supply of doctors, through limits on medical school intakes and doctor immigration, are used to control Medicare outlays.

• Pharmaceutical Benefits Scheme outlays are constrained by more complicated, though more rational and direct, measures.

Capital investment is even more tightly rationed than "recurrent" inputs to force every possible dollar into statistically measured outputs rather than into investments in system improvement. New structures are not properly maintained — hence Smallwood's hospitals in "varying states of dilapidation". Governments may find the dollars for showy new equipment, often "opened" with elaborate political ritual, but the recurrent funding required for its operation and maintenance is another story. Investment in information and communications technology (ICT), being largely invisible, is even more restricted, with serious consequences for efficiency, quality and safety of care.

The average age of the capital stock in Australian healthcare is 16 years, compared with 10–12 years in most of the private sector.11 According to Gartner Corporation, a major international ICT consultancy, spending on ICT is equivalent to about 3% of turnover in the centrally planned and controlled European and Australian healthcare systems.12 In the market-exposed United States health sector, ICT spending is equivalent to between 5% and 6% of turnover.12

The first objective en route to an outcome-enabled health system must be to relieve the constraints that bind inputs and distort the health services' "production" system. Ubiquitous input controls also distort the pattern of outputs, as, in the presence of input rationing, outputs reflect the supply of constrained inputs rather than the demand for health services. This is why the interminable academic debate about supplier-induced demand is just that — academic. If reform is to be possible, the system must be given some additional degrees of freedom. Eventually output must come to reflect demand rather than supply to make an outcome-oriented health system possible. Other measures will, of course, be needed, such as universal personal health records and other supports for quality and safety enhancement, before reform is complete, but the initial steps must centre on capacity building.

The recent Wanless Review of the United Kingdom health system came to somewhat similar conclusions regarding input constraints.13 Wanless (Group Chief Executive of NatWest for seven years until he retired in 1999) carried out the Health Trends Review at the request of the Chancellor of the Exchequer in 2001–2002 as a key contribution to the Blair government's planning for a major increase in spending on the United Kingdom National Health Service. His report found that improving the use of information and communication technology in the Health Service is a key issue in improving quality and productivity. Wanless also concluded that there is scope for major changes in skill mix and the way professionals work in the healthcare service, although he was diplomatically silent as to how the existing distortions may have come about, and what might be done about ongoing causes. He stopped well short of questioning the structural features of the UK's monolithic National Health System.

If Ministers wish to entrench quality, safety and patient choice as the primary drivers of care, they must take measures to move the locus of control away from the input end of the production chain and toward the output end — that is, toward the consumer. This does not imply dismantling equitable and universal health insurance. It simply means that government must move towards a health financing system that insures the patient. It will take time, and must be done openly and carefully, if the effort is to endure.

There is a growing consensus that Australia, like the UK, should be spending more on healthcare. However, if we were simply to increase spending the effect would be immediate inflation of healthcare costs. First we must feed out some slack in doctor numbers, settle the major nursing issues, and relieve the pinch points that stem from a history of capital rationing — a 5–10-year recovery program. While this may sound backward looking, there is no prospect that we can accurately see the system of the future while the system we already have does not match the demands of the present.14

Five to 10 years is also long enough to build a base for a universal longitudinal patient record, and to do so in the only way that can succeed — from the bottom up. The foundation of a national system must be electronic patient record systems in routine and ubiquitous daily use by providers. At present, only general practice makes widespread use of electronic records and orders. Practice grants brought about the revolution in GP computing. Tied grants must be offered to hospitals so that they can overcome their ICT investment backlog.

Many of the elements of the wider program of reform require that hospitals begin the transition to enterprise-level electronic clinical support systems and electronic patient records as soon as possible. There will be no outcome-driven healthcare system until the system recognises the whole patient — a person with a past, a present, and a future. Ambitions for major gains in safety and quality of healthcare will come to very little until every patient encounter with any healthcare provider is supported by the patient's personal health record.

Whether we stay with the present third-party payment system or move gradually to patient-based funding, as I believe we should, the injection of funds must not run ahead of real capacity or the whole project will end up in disrepute. Governments are right to dread healthcare cost inflation.

We should look to Health Ministers neither for radicalism nor for a financial bonanza. Barriers to continuity of care and patient-based outcomes must be carefully dismantled. None of the conceptual, practical or financial requirements for fundamental reform is beyond us. The path will be long; there are 50 years of history invested in the existing Australian system and its counterparts elsewhere. The system has allowed people in all walks of life to achieve standards of health and longevity unimaginable 50 years ago. The next broad advance will take the form of healthcare which is precisely matched to the wishes of the individual. A mass-oriented public health system has taken us a long way, but it cannot do what is now required.