Mortality due to rheumatic heart disease in the Kimberley 2001–2010
Objective: To determine the mortality burden of rheumatic heart disease (RHD) in the Kimberley.
Methods: A retrospective medical record audit was conducted for patients identified by the Western Australian (WA) RHD Program as deceased between 2001 and 2010. Patients with documented evidence strongly suggesting or confirming RHD were included in the study. Crude and age-standardised death rates were calculated.
Results: A total of 34 patients were identified as having RHD, 15 of whom died of RHD-attributable causes and 93% of whom were Aboriginal. The most common causes of death were arrhythmias and heart failure. The mean age at death of Aboriginal people was 41 years. The age-standardised death rate in Aboriginal people attributable to RHD in the Kimberley was 12.5 per 100,000 people per year, which is 1.6 times the rate of Indigenous Australians nationally. RHD contributed to 342 potential life years lost over the 10-year period.
Conclusion: RHD contributes to significant premature mortality and higher rates of death in Aboriginal people in the Kimberley, which is consistent with other areas of northern Australia. While the recent establishment of the WA RHD Program will endeavour to improve mortality and morbidity due to RHD in the Kimberley, further research and investment is needed to address this disease of socioeconomic disadvantage.
Acute rheumatic fever (ARF) is an acute inflammatory illness that occurs in response to a bacterial infection with group A streptococcus (GAS). ARF predominantly affects the heart, joints, brain and skin. Once the acute illness has resolved, long-term consequences include long-standing damage to the heart, particularly the mitral and/or aortic valves, leading to rheumatic heart disease (RHD). This condition predominantly affects children, adolescents and young adults.1
Approximately 15–19 million people worldwide are affected by RHD, with 282,000 new cases of RHD and more than 230,000 deaths due to RHD per year.2
Even though the overwhelming majority of RHD worldwide occurs in developing nations, some of the highest rates of RHD are found in Aboriginal Australians, which is likely attributable to socioeconomic and environmental factors.1 RHD remains a significant cause of mortality and morbidity in this population. Indigenous people are eight times more likely to be hospitalised for ARF and RHD, and those aged 25–64 years are 20 times more likely to die from RHD-attributable causes than their non-Indigenous counterparts.3
The incidence of ARF in the Kimberley, a remote region of Western Australia (WA), is approximately 375 per 100,000 in Aboriginal children aged 5–14 years, reflecting a high burden of disease in this population.4 This rate is comparable to those in Aboriginal and Torres Strait Islander people living in regional and remote areas of the Northern Territory (NT).1 The prevalence of RHD in Indigenous people in the Kimberley is conservatively estimated to be 1.0%, compared to 1.1% and almost 2% in North Queensland and the NT, respectively.5–6
RHD causes significant mortality, most of which occurs in early adulthood to middle age, thus affecting people in their most productive years.7–9 The age-adjusted death rate for Indigenous Australians is approximately eight deaths per 100,000 people, which is four times higher than that for their non-Indigenous counterparts. Although mortality rates attributable to RHD in the Kimberley have never been recorded, anecdotal reports suggest that people with RHD in this region also experience premature mortality.
This study aims to determine the burden of mortality of RHD in the Kimberley, with specific reference to the average age of death attributable to RHD, the number of years of potential life lost due to RHD and common causes of death associated with RHD.
A list of patients with suspected RHD who had died between 2001 and 2010 was compiled by the WA RHD Program using a system of enhanced case finding across the Kimberley region. Cases were identified using: ARF/RHD patient registers at primary health care clinics; recall lists generated from local electronic health systems (Ferret or Communicare); regional physician and paediatrician letters from 2003–2005; echocardiogram reports provided by the only visiting adult cardiology service in the Kimberley from 2000–2005; and a recall database used by the visiting paediatric cardiology service.10 Patients identified by the WA RHD Program subsequent to 2007 were also included.
All efforts were made to review patient medical records and death certificates. Coroner's reports were accessed where available. These records were accessed at regional hospitals in the Kimberley as well as the two cardiothoracic referral centres in WA (Royal Perth Hospital and Fremantle Hospital).
Inclusion criteria for the study included ‘rheumatic’ mitral or aortic valves documented on echocardiogram, operation report or post-mortem report; history of mitral prosthetic valve replacement, valve repair or valvuloplasty; diagnosis of RHD documented in cardiology specialist correspondence; and mitral stenosis mean gradient ≥4 mmHg as per World Heart Federation (WHF) criteria.11
Exclusion criteria for the study included documented ARF without echocardiographical evidence of RHD; documented history of RHD with normal echocardiogram; isolated tricuspid regurgitation; complex congenital cardiac anomalies; and cardiomyopathy secondary to other causes.
An average yearly crude death rate was calculated for Aboriginal patients using the total number of deaths over the 10-year period and the mid-period population in 2006. This population data was obtained from the Australian Bureau of Statistics.12 Given the relatively young population distribution of the Aboriginal population in the Kimberley, these rates were then age-standardised against the World Standard Population.13 Non-Aboriginal rates were not calculated given the sample included only one non-Aboriginal person. Confidence intervals were calculated at 95% as an estimate of variability due to random chance.
Ethics approval for the study (ref 2011:23) was granted by the Western Australia Country Health Service Human Research Ethics Committee and the Western Australian Aboriginal Health Ethics Committee (ref 364–09/11). This study was also supported by the Kimberley Aboriginal Health Planning Forum Research Subcommittee (ref 2011–014).
A case list of 57 deceased patients with suspected RHD was provided by the WA RHD Program. There were 34 cases that met the inclusion criteria (Figure 1). Death details were available for 24 of these cases and, of these, 63% (15 of 24) were likely attributable to RHD (Table 2).
Table 1. Death rates in Aboriginal people from RHD-attributable causes.
|Total||14||12,324||11.4 (5.4–17.4)||12.5 (4.5–15.8)|
All patients who died from RHD-attributable deaths had moderate-severe disease (as classified by RHD Australia).1 Of the patients who died from RHD-attributable deaths, 93% (14 of 15) were Aboriginal and 60% (9 of 15) were male. The mean age of death from RHD-attributable causes in Aboriginal people was 41 years (median 35 years; range 26–73 years). Table 1 demonstrates age-group specific death rates in Aboriginal people from RHD-attributable causes. The overall total crude death rate was 11.4 per 100,000 person-years for Aboriginal people. The age-standardised death rate was 12.5 deaths per 100,000 people per year adjusted for the young age of the population. This is 1.6 times higher than the rate for Indigenous Australians nationally.3
Table 2. Cause of death in RHD-attributable deaths.a
|Complication of valvular surgery|
1× intraoperative haemorrhage
1× intraoperative death
|Prosthetic valvular thrombosis||1|
In Aboriginal patients under the age of 65, there were 342 potential life years lost over the 10-year period attributable to RHD. This is an indicator of premature mortality and equates to a rate of 29 potential years of life lost per 1,000 people. The age-standardised rate of years of potential life lost before age 65 (premature mortality) for Aboriginal people was 2.6 per 1,000 population.
Of the deaths for which details were available, 62.5% (15 of 24) were attributable to RHD. The most common causes of death were arrhythmias and heart failure. One patient died from a cerebrovascular accident with a background of a prosthetic valve and subtherapeutic INR. Two deaths due to other causes may have been related to RHD; however, there was insufficient information to support this at the time. One patient died of sepsis with an unidentified source but previous documented episodes of infective endocarditis, and another patient died out of hospital of an unknown cause.
This is the first study reporting RHD-attributable mortality rates in the Kimberley population. The average age of death from RHD-attributable causes in Aboriginal people was 41 years. This compares to a mean age at death of 36–58 years in other Indigenous populations.7–9 Despite the small numbers and study limitations, the standardised death rate for Aboriginal people was higher than that recorded nationally and there was significant premature mortality attributable to RHD. Due to the small numbers of deaths, the confidence interval was only able to be calculated for the total numbers of deaths.
A major limitation for this study was the state of health records in the Kimberley. There is currently no single coordinated shared electronic health record used within the Kimberley region. Many health records remain in paper files in remote clinics or regional hospitals across a large geographical area. This resulted in incomplete data collection and likely underestimation of rates of RHD-attributable deaths in the region.
The unavailability of formal death registry data was another major limitation. This was particularly relevant for patients who died in the community and did not have information surrounding the circumstances of their death documented in medical notes. This also resulted in incomplete data collection and likely underestimation of rates of RHD-attributable deaths. In addition, it was difficult to determine a location of death trend for this cohort as data was obtained primarily from health care records (particularly regional hospitals) with a significant paucity of data for those who died in the community.
Given the transient nature of the population, high staff turnover rates in the region and a lack of shared electronic records, the retrospective case identification during 2006–07 for patients who had died as early as 5–6 years prior was also likely to underestimate deceased patients with RHD in the region.
Due to the small number of patients, the data likely provide an incomplete picture of the mortality burden of RHD. As such, the authors suggest exercising caution when applying these results to the general Kimberley population.
At least three-quarters of the patients who were excluded from the study were excluded due to inadequate documentation to support a diagnosis of RHD, or because of echocardiographic reports that supported other diagnoses (e.g. complex congenital disease, isolated tricuspid regurgitation or aortic sclerosis). This suggests that several patients may have been labelled with a diagnosis of RHD when it may not have been accurate.
Mortality data in children and adults younger than 20 years of age was particularly scarce in this study. For reasons already discussed, such cases may have been missed and subsequently underrepresented in this cohort, particularly given other case reports from northern Australia.14,15
This study provides preliminary data on the mortality burden of RHD in the Kimberley, although it is likely to be a conservative estimate. RHD and its complications have a significant impact on mortality and efforts should focus on prevention and management of these cases and, most importantly, the prevention of ARF. Efforts should also focus on more accessible, comprehensive and accurate regional data in the future. The recent introduction of the RHD Program in the Kimberley is a welcome step to focus on coordination of diagnosis, secondary prevention and management of RHD in the Kimberley. In the interim, ongoing research to access formal death registry data could develop a more comprehensive picture that could be generalised to the Kimberley.
We acknowledge the contributions of the Aboriginal and Torres Strait Islander community. We thank the medical records staff at the regional hospitals in the Kimberley as well as those at the Royal Perth Hospital and Fremantle Hospital. We thank Dr Jaye Martin, regional consultant physician, for clinical support and interpretation as well as draft feedback. Thank you to Dr Jake Parker at KPHU for assistance with data collection and draft feedback. We also acknowledge Dr David Atkinson at the Kimberley Aboriginal Medical Services Council and Julie Kohring at RHD WA for project support.