Health status is an indicator of the state of a nation’s health. Most OECD countries achieve similar levels of performance on key indicators of health status, particularly life expectancy (OECD 2005). Most countries seek to protect their ranking and actively manage this measure. However, while health status outcomes are impressive for non-Indigenous populations, those for Indigenous populations are often much lower (Bramley et al 2004) and the gap may be widening (Freemantle et al 2007). Table 1.1 below sets out selected indicators of performance on health status for selected peer countries. The data show overall performance and performance on specific life-expectancy measures for the four leading countries in 2004, compared with those of other selected OECD peers.

Health targets, such as health status measures, are important because, theoretically, they focus the system on the effectiveness as well as the efficiency of care. However, the extent to which health status outcomes are linked to health service effectiveness and efficiency is not clear. For instance, the proportion of funds spent on health is not necessarily an indicator of performance on health status outcomes. As an example, Japan has one of the lowest proportions of GDP spent on health (8% in 2003) but a high overall life expectancy (81.8 years), while the US spends a relatively high proportion of its GDP on health (15.2% in 2003) but has a lower overall life expectancy (77.5 years) (OECD 2006b). If health status outcomes are not necessarily linked to the size of health budgets, other factors must explain the difference. Knowing what factors ‘produce’ health and understanding how budgets are spent is therefore important in understanding such differences. But health reform policy and programs aimed at improving health service performance appear to be largely separated from the policies and programs that underpin health status outcomes. Busse & Wismar (2002) describe health targets as the ‘forgotten corner’, because of their separation from the main agenda of health services reform. A comparative study of national and regional health target programs among OECD peers, including the European Union, Australia, Canada, New Zealand and the US, showed that Australia alone linked health targets to health services performance (Busse & Wismar 2002).

Busse & Wismar offer a number of reasons for this separation. Firstly, the policies and programs that support health targets tend to be top-down and the involvement of the general public or their elected parliamentary officials is limited. This means that policy may not contain strategies to encourage grassroots alliances to mobilise health improvement activities in the community and in health services. Secondly, and consequently, there are few incentives for local and professional individuals and groups to be involved. This is important because getting the agreement and commitment of those who produce and co-produce health is a significant factor in improving health outcomes. Yet the evidence is clear that integrating community and professional groups in decisions that affect them is effective. Communities can make difficult decisions about prioritising health services in the face of scarce resources. This is evidenced by the Oregon experiment in the US that sought to involve communities in rationing decisions that took into account the health of vulnerable groups, and the New Zealand experience where professional groups assisted government to develop rationing criteria for elective surgery (Hadorn & Holmes 1997, Klein et al 1996). Importantly, Wiseman et al’s Australian study found that the public overwhelmingly want their preferences to inform decisions about priority-setting and funding allocation (Wiseman et al 2003) (see Mooney, Chapter 13).

For governments, a constant priority is managing the gap between the supply of health services and the demand for them. Governments tend to use the blunt levers of economic policy to contain costs to manage the gap, in preference to more organic processes that involve communities in such decisions. The European Union, Australia, New Zealand and Canada all use economic policy in this way, as do the US and Japan (Abel-Smith & Mossialos 1994, Fujii & Reich 1988, Malcolm 1990, Segal 1998, Dickey 1997, Byrne & Rathwell 2005). Economic policy is a useful lever for governments to manage health services, because it allows control over resource allocation at the macro level at which government decision makers operate. Moreover, the pressures on the health system are likely to continue, hence the pressures on governments to manage the gap between service supply and demand will also continue. Managing the resource consumption of health services will remain an important strategy to contain public spending on health to within manageable levels as populations in developed countries age, the demand for expensive advanced medical technology rises and consumer expectations about choice, access, quality and accountability increase.

These pressures on the health system need to be taken into account when assessing how well they perform. Health service performance is affected by the rising cost of care, the increasing demand for services, the levels of risk to patient safety and the overall quality of care. Each of these pressures is discussed briefly in turn.

First, in terms of the rising cost of care, health services consume a sizable portion of a nation’s resources. Table 1.2 below compares the total expenditure on health as a percentage of GDP (gross domestic product) for selected OECD peers from 1960 to 2004. The expenditure of these countries (Australia, Canada, New Zealand and the UK) is compared with that of countries with relatively low and high expenditure (Japan and the US respectively). The data show that for the six countries selected, all experienced substantial rises in the percentage of GDP spent on health over the 44-year period. National budgets are finite, and as health spending rises, the proportion of GDP spent in other areas of need reduces. Consequently, most governments actively manage their expenditure on health. The high level of public funding facilitates such management, although countries such as the US experience more difficulty, presumably because of a dominant private sector (OECD 2006a) less amenable to centralised cost containment objectives.

Second, the demand for health services is rising, and is likely to continue to rise in line with the ageing of the population in developed countries. As an example, Figure 1.1 below shows the increase in admitted patients in Australian hospitals from 1998–99 to 2004–05.

Governments and health services must manage this increasing demand, and health resources must be rationed to manage it within allocated budgets. Reducing bed numbers and patients’ lengths of stay are common strategies to contain costs within budget limits while maintaining patient throughput. Table 1.3 below shows data for seven selected countries, as in Table 1.1, for 2004 on three key indicators of efficiency, namely waiting times for health services, average lengths of stay and proportion of hospital beds per 1000 population. The data vary widely. Of the two countries with lower lengths of stay (Sweden and New Zealand) and bed ratios (Sweden), each experiences waiting times for rationed hospital services to contain the proportion of GDP spent on health: for Sweden to 9.1%; for New Zealand to 8.4% (OECD 2006b). The two countries that report no waiting times for hospital services (Switzerland and France) also have the highest proportion of beds, and in the case of Switzerland, the highest average length of stay. Thus, the trade-off for maintaining access to health services with no waiting times appears to be a higher proportion of GDP spent: for Switzerland 11.6% in 2004; for France 10.5% (OECD 2006b).

The extent to which reducing bed numbers is a sustainable strategy to manage throughput and cost emerges in Figure 1.2 overleaf. The data show that as an example the consistent reduction in bed numbers in Australia from 1998–99 to 2002–03 was not sustainable, with bed numbers rising again from 2003–04.

Third, risks to patient safety have become a priority for many health services in developed countries (see Warburton, Chapter 9 and Merry, Chapter 11). The methodology to quantify adverse events developed in the 1990s has revealed a pattern of patient risk associated with acute inpatient admissions within a relatively consistent range across selected countries for which data are available, from 7.5% in Canada to 17.7% in the US, as Table 1.4 shows.

Three issues arise from these data. First, a large majority of the adverse events reported caused significant disability or death and were considered highly preventable: 57% in the Australian study (Wilson et al 1995). Second, the cost of adverse events was high, estimated to be in excess of $800 million per annum, again in the Australian case. This represents a significant proportion of the health budget that potentially could be directed to meeting the supply–demand gap. Third, these figures raise questions about their cause, whether individual clinicians are at fault, or inadequate systems. This issue is important when decisions about health service priorities and funding for improvements are being considered.

Errors occur in all health and hospital services and must be managed as routine. For instance, Cullen compared adverse drug events in ICUs with events in non-ICUs (Cullen et al 1997) and when adjusted for volume, the rates were found to be comparable. This study is significant, because it revealed that preventable and potential adverse drug events occurred in units that functioned normally, involving caregivers working under reasonably normal circumstances and not necessarily or not only in units at the extremes of workload and stress, such as ICUs. Cullen’s estimate of cost of between $5.6 and $2.8 million in a 700-bed teaching hospital suggests that eliminating error is not only good for patients, it is also good for budgets.

In addressing the issue of cause, Baker et al believes that poorly designed systems are at fault. This view notwithstanding

The implications of this for managing clinical processes arise in terms of the types of solutions devised to address adverse events, and whether they should focus on improving the skills and competencies of individual clinicians, or rectifying underlying systems problems.

Fourth, the overall quality of health services has concerned policymakers, health service managers, clinicians and patients, and it had done so well before the high levels of adverse events came to light. The problems with quality are serious and extensive. Chassin (1998) maintains that they arise from treatment underuse (failure to provide a treatment when needed), overuse (when harm exceeds the benefit of a treatment) and misuse (when complications occur from the treatment). He maintains that the need for rapid change in health services is urgent because

Policymakers and health service managers have generally responded to the problems in overall quality with broad quality improvement programs such as continuous quality improvement (CQI). Just how effective such programs are is uncertain, because of a paucity of evaluation. Some clinicians believe that the quality of care has actually deteriorated in the countries where they have been implemented (Ferlie & Shortell 2001). A possible reason for this is that the broad scope of programs is too ambitious as they attempt to cover the myriad systems that appear to impact on quality. Thus, the links between cause and effect have been difficult to establish. What these programs do reveal is further evidence of wide variations in the processes and outcomes of care in technologically advanced systems.

Ferlie & Shortell (2001)