The coordinated care trials were nothing if not ambitious! In 1994, the Council of Australian Government proposed that the organisation, funding and management of health and community services could be restructured into three streams: 1

• a general stream for individuals needing either occasional or uncomplicated healthcare;

• an acute stream for patients needing specialised services for acute illness; and

• a coordinated stream for patients requiring a mix of healthcare services over an extended period. It was assumed that in the last stream patients would benefit by having their care managed and coordinated.

Following a national tender process, nine trials of coordinated care were activated and funded by the Commonwealth (Box 1). A central premise of the trials was that better coordination of the care of people with chronic or complex needs would reduce hospitalisation, and the savings could cover the costs of coordination. The coordinated care trials were the largest and most ambitious experiment of a new method of organising healthcare services ever attempted in Australia.

The trials ran for two years. Evaluation was undertaken both nationally and at the local level. Much of the trial documentation and evaluation reports are available on the Commonwealth Department of Health and Ageing's website.3

Despite this effort and expenditure, the outcomes were disappointing (Box 1). In general, the trials did not demonstrate improved health and well-being of the participants. A significant reduction in hospital admissions in the intervention compared with the control group was seen in only three of the trials, and for most trials an accrued operating deficit was found. Was this ambitious healthcare experiment a failure? It is hard to say because, unfortunately, the design of the trials made it difficult for them to achieve their stated objectives.

The trials included the following design shortcomings.

• Each trial was funded for two years, but the first six months were devoted to recruitment and the last six months were a wind-down phase. Thus, in many trials, the actual intervention was for 12 months or less, a very short period in which to make an impact on complex illnesses.

• Difficulty in recruiting sufficient numbers of participants forced many trials to relax inclusion criteria, with the result that many individuals entered in the trial were inherently unable to benefit from coordinated care, since they were not sick enough, or had insufficiently complex problems to warrant care coordination.4,5

• In many trials, the same intervention was applied to all participants regardless of the severity of their condition or ability to respond to the intervention.

• Interventions varied markedly between trials.

• The chosen measure of well-being, the SF-36, was not optimal to assess the types of intervention, especially over the relatively short trial periods.

Despite these shortcomings, the trials provided a number of benefits. Fundholding allowed trials to fund strategies such as quit smoking interventions that otherwise would not have been possible.6 Because of the evaluation needs, many service organisations received major technology upgrading of information systems. The trials received the full cooperation of the Health Insurance Commission, enabling the use of Pharmaceutical Benefits Scheme and Medical Benefits Scheme (MBS) information for evaluation purposes. New enhanced primary care MBS schedule items were announced shortly before the publication of the final results. Finally, much of the qualitative evaluation showed that participants in the intervention groups appreciated the extra coordination of their care.

The new round of coordinated care trials that has just commenced (Box 2) has taken on board much of what has been learnt from the first round. The trials are for three years rather than two, have better-targeted interventions, and outcome measures have been carefully selected for their sensitivity to the type of intervention.

The possibility remains, however, that the essential premise that better coordination reduces hospitalisation is misguided. It may be that lack of coordination in a complex care system operates as a functioning rationing system, so that better care coordination reveals unmet needs rather than resolving them. Experience in the mental health field implies that this may be so.7,8

With an ageing population and increasing burden of chronic diseases, the government has given priority to increased service coordination, vertical integration and cost containment. The coordinated care trials are just one of several strategies aimed at achieving these objectives. It might well be that the objectives are mutually exclusive and that improved coordination comes at a cost.