Inequitable provision of optimal services for patients with chronic heart failure: a national geo-mapping study

In response to the high burden of disease associated with chronic heart failure (CHF),1 in particular the high rates of hospital admissions,2 dedicated CHF management programs have been developed in Australia and internationally.3,4 These multidisciplinary programs target recently hospitalised patients in an effort to optimise the continuum of care after discharge.4

A series of recent meta-analyses has confirmed that CHF management programs minimise readmission rates, improve quality of life, reduce costs, and prolong survival.4-6

The most comprehensive of the meta-analyses has shown a significant reduction in the risk of hospital readmission (relative risk [RR], 0.82; 95% CI, 0.72–0.93) and all-cause mortality (RR, 0.80; 95% CI, 0.66–0.98) associated with multidisciplinary programs of care.4

In a 2004–2005 survey of the location and type of CHF management programs nationally, our research team found a disproportionate concentration of services (93%) in capital and major cities.7 The reality in Australia is that the burden of post-discharge care for the 40% of people with CHF living outside of capital cities (about 135 000 patients)8 falls solely onto community-based general practitioners. Unfortunately, there are diminishing numbers of GPs and very few CHF management programs in rural and remote regions.8,9

Given the apparent mismatch between the demand for and availability of CHF management programs, we sought to accurately compare the location and accessibility of current programs and general practice services with the geographic distribution of people with CHF.

Our study combined geographic information systems (GIS) technology, which maps the geographic distribution and demography of the Australian population, with contemporary estimates of the prevalence and distribution of CHF,8 the location of general practices,10 and the known CHF management programs operating in Australia from 1 January 2004 to 31 December 2005.7

Overall, our population and service mapping study suggests that there are many areas of high CHF prevalence (2%–8%) outside capital cities. These areas are typically populated by people over 65 years old or are large Aboriginal and Torres Strait Islander communities. Our CHF population estimates compare well with current published data such as the Cardiac Awareness Survey and Evaluation (CASE) study (13.2% of population aged > 60 years),19 and the Canberra Heart Study (6.3% of population aged > 60 years),20 although neither of these studies have reported on rural, Indigenous, or disadvantaged populations. Our study also suggests that, of the probable 63 000 individuals admitted to hospital with CHF during 2004–2005,2 only 8% (5000 patients) were enrolled in a CHF management program. This indicates that most Australian CHF patients did not receive recommended evidence-based care, which includes regular access to specialist CHF management programs.21,22

The clear disparity between evidence and practice was accentuated in rural and remote regions, with only four CHF management programs located outside major cities. These small centres of care managed about 80 (0.7%) of the potential 16 000 rural patients with CHF discharged from hospital in that year.2

The 62 CHF management programs operating within Australia during the study period were tightly clustered around the capital city regions, predominantly on the eastern seaboard (Box 2). As CHF management programs are a relatively new initiative, this might be expected.

However, we also found a potential mismatch between supply and demand for long-established general practice services. For the typically old and fragile patient with CHF, even 1 km may be a challenge in reaching a GP or an outpatient clinic. The average distances to CHF management programs or, where these are not available, to GP care preclude CHF models of care such as home visiting. Alternative solutions of care, including general practice-based multidisciplinary team approaches or the use of information technology to bridge the gap between very remote communities and centres of care, should be considered.23

Currently, most remote health services in Australia are provided by state and territory governments, with the Commonwealth funding aspects of Indigenous services and primary health care through Medicare. Although there has been some progress in coordination of responsibility by different levels of government in the area of Indigenous health, there is still no clear responsibility and accountability for health service and health outcomes between the different levels of government. This split in Commonwealth–state funding arrangements makes the delivery of integrated care for people with chronic disease (such as CHF management programs) extremely difficult.

Our study has several limitations that require comment. Most importantly, our technique hinges on accurate population and epidemiological data. In the Australian context, we re-emphasise the need for rigorous epidemiological data that clearly describe the particular characteristics of rural and Indigenous populations, to replace our population estimates.24 The limitation of using ABS census collection district data or any statistical method is the presence of outliers, which create extremes and distort the overall results. There were several locations on our maps that appeared to indicate high prevalence over large areas in remote regions. This phenomenon is caused by the required number of households for a collection district (225) being spread over an extensive geographic area, and is why we presented both the absolute number of cases and overall prevalence rates.

Despite these limitations, our data highlight a clear mismatch between where services are provided and where people with CHF are likely to live in the greatest concentrations. There is a need to provide equitable access to CHF management programs to the ageing population regardless of its location. The use of GIS technology in guiding service and resource allocation is scientifically grounded and free of politics and subjectivity. The concepts presented in this article are internationally unique and could have broad application for all chronic disease service research.

1 Estimated numbers of people living with chronic heart failure (CHF) in Australian census collection districts, and locations of CHF management programs

*Natural Breaks (Jenks’) Classification. ABS = Australian Bureau of Statistics. ASGC = Australian Standard Geographical Classification. CD = collection district. GISCA = National Centre for Social Applications of Geographic Information Systems.

2 Estimated prevalence rate of chronic heart failure (CHF) in Australian census collection districts, and locations of general practitioners and CHF management programs

*Natural Breaks (Jenks’) Classification. ABS = Australian Bureau of Statistics. ASGC = Australian Standard Geographical Classification. CD = collection district. GISCA = National Centre for Social Applications of Geographic Information Systems.