The economic costs of heart attack and chest pain (Acute Coronary Syndrome)

Transcription

1 The economic costs of heart attack and chest pain (Acute Coronary Syndrome)

2 Acknowledgements Access Economics would like to acknowledge with appreciation the insightful comments and guidance received from various people in the development of this report, including Professor Derek Chew, Associate Professor David Brieger, Dr Ren Tan, Jenny Coutts, Margaret Flaherty, Ineke Bleeker, Dr Alex Brown, Graham Neville, Tony Arvidsson, Rohan Greenland, Dr Andrew Boyden, Kim Goodman, Debbie White, Associate Professor Paul Middleton, Linda Soars, members of the Cardiology Advisory Board (Eli Lilly), Dr Deon Gouws, Paul Dale, Stuart Englund and Fiona Bailey. We would especially like to thank Emeritus Professor Michael Hobbs for providing access to unpublished research information and advice on the epidemiology of ACS and its treatment in Perth. This report aims to enhance the understanding of, and reiterate, the growing impact of Acute Coronary Syndrome on Australia and the need for every effort to be made to resolve the treatment gaps. Copyright Data relating to the WA linked database supplied by Emeritus Professor Michael Hobbs and presented in this report are unpublished from research in progress. They may not be provided to, or published by, third parties without the permission of Professor Michael Hobbs. To obtain a copy A copy of The economics costs of heart attack and chest pain (Acute Coronary Syndrome) can be downloaded from

3 The economic costs of heart attack and chest pain (Acute Coronary Syndrome) June 2009 While every effort has been made to ensure the accuracy of this document, the uncertain nature of economic data, forecasting and analysis means that Access Economics Pty Limited is unable to make any warranties in relation to the information contained herein. Access Economics Pty Limited, its employees and agents disclaim liability for any loss or damage which may arise as a consequence of any person relying on the information contained in this document.

4 Executive Summary Access Economics was commissioned by Eli Lilly to estimate the economic costs of Heart Attack and Chest Pain (Acute Coronary Syndrome-ACS) in Australia for In addition, an objective was to investigate current gaps in ACS treatment and clinical need, and highlight areas of treatment where further investment may result in significant benefits through a reduction in the burden of disease and improvements in efficiency and quality of care. To estimate the economic costs of ACS, this study has used the comprehensive cost of illness framework used throughout the world. In brief, the study consists of the following sections: epidemiology of ACS in Australia; direct health care system costs associated with treatment; indirect financial and economic costs; value of the loss in health associated with morbidity and mortality; and the future of ACS management in Australia. Unless ACS leads to immediate death, patients experiencing an ACS event are hospitalised. Data on hospitalisations and death were used to estimate the number of ACS events in Australia. As the Australian Institute of Health and Welfare (AIHW) data does not account for readmission and transfers, 28 day age standardised separation rates were sourced from the Western Australian linked dataset with the assistance of Emeritus Professor Michael Hobbs. These were extrapolated to the Australian setting using projected Australian population data. It is projected that in 2009 there will be around 79,990 hospitalisation associated with ACS, of which 59% is expected to be due to heart attack (AMI), and the remaining associated with chest pain (unstable angina). Table i shows the projected number of hospitalisations by gender and condition for Table i: Projected number of ACS hospitalisations in Australia 2009 Unstable angina AMI ACS Male 20,224 28,596 48,820 Female 12,228 18,943 31,170 Total 32,452 47,539 79,990 Source: Access Economics calculations Some hospitalisations due to heart attacks are likely to be followed by death. However, deaths following a hospitalisation are expected to account for only 24% of all deaths associated with heart attacks. Most deaths will occur before a person can be admitted to hospital. In total, it is expected that 9,959 people will die from a heart attack in 2009, of which 2,423 are expected to occur within 28 days of an admission. Projected deaths following a heart attack by gender and age for 2009 are shown in Table ii.

5 Table ii: Projected number of deaths following a heart attack in Australia 2009 Males Females Total years years years years , years 1,905 1,570 3, years and over 1,427 2,741 4,167 Total 5,022 4,937 9,959 Source: ABS (2003, 2004, 2005, 2006 and 2007) and Access Economics calculations The projected number of hospitalisations and deaths associated with ACS means the direct health care system costs, indirect costs, and burden of disease imposed on society will be significant. Table iii presents a summary of projected hospitalisations, deaths and economic costs associated with ACS, split into various cost components, for It is projected that the number of ACS hospitalisations and deaths will be 87,526 in 2009 with an associated total economic cost of $17.9 billion. Of this, direct health care system costs (primarily hospital stays and pharmaceuticals) are expected to account for around $1.8 billion. Indirect costs are expected to account for $3.8 billion, primarily due to lost productivity. The largest cost is expected to be the loss in the value of health, otherwise known as the burden of disease due to morbidity and mortality. It is expected that due to disability imposed on individuals, and the loss of life associated with premature mortality, the value in the loss of health will be approximately $12.3 billion in In total, heart attacks are expected to cost around $15.5 billion in The majority of these costs are associated with the loss in the value of health, accounting for around 78%, which is representative of the large amount of premature deaths associated with heart attacks. Total direct health care system costs and indirect costs are expected to total around $3.5 billion in The total cost per heart attack is expected to average $281,000. Unstable angina (chest pain at rest) is expected to cost around $2.4 billion in However the burden of disease only comprises $311 million, or around 13%. The majority of costs are associated with direct and indirect costs, totalling around $2.1 billion. The total cost per unstable angina event is expected to average $74,000.

6 Table iii: Summary of estimated separations, deaths and costs 2009 Heart attack Chest pain ACS Deaths before reaching a hospital 7, ,536 Hospitalisations without death a 45,115 32,452 77,567 Hospitalisations with death occurring later 2, ,423 Total hospitalisations 47,538 32,452 79,990 Total Events 55,074 32,452 87,526 $ (million) $ (million) $ (million) Direct health care system costs 1, ,767 Productivity loss (reduced participation) 1,254 1,073 2,327 Productivity loss (premature mortality) Informal care Deadweight loss Burden of disease (YLD) ,030 Burden of disease (YLL) 11, ,307 Total costs 15,497 2,400 17,895 $ $ $ Cost per separation (direct costs only) b 25,000 18,000 22,000 Cost per event (all costs) b 281,000 74, ,000 Note: (a) Within 28 days of being admitted to hospital (b) Cost per hospitalisation and cost per event have been rounded to the nearest $1,000. Source: Access Economics This study has also highlighted gaps in the treatment and monitoring of ACS throughout Australia. These include: a national ACS registry managed by an independent body that includes comprehensive and consistent data on patients, treatment, and rehabilitation services Australia-wide; which can be used to develop a common set of performance indicators and ACS treatment outcome measures; a national approach to cardiac rehabilitation, including inpatient, outpatient and maintenance care, specific strategies to increase the uptake of women into rehabilitation, further investment to ensure rehabilitation programs are accessible to all regardless of income and geographical location; an increase in the compliance and adherence with medication via the Quality Use of Medicines program; a standardised national program to support employees and employers and the extension of rehabilitation practices. Workplaces can provide an excellent environment to facilitate the ongoing rehabilitation and lifestyle changes to prevent the re-occurrence of ACS events; and

7 further research into the optimal use of existing therapies and identification and promotion of cost effective treatments currently being used within other health systems throughout the world. The focus on ACS at this point in time is particularly important in the context of demographic ageing in Australia, given the increasing age standardisation rates among the older population and the link between health, health care resource utilisation, and quality of life. In 2010, the first of the baby boomers will reach the age of 65 years, where the risk of ACS significantly increases. It is expected that the proportion of the Australian population that is 65 years and older (and therefore at higher risk of an ACS event) will increase from around 14% in 2009 to around 23% in This, coupled with the expected increase in risk factors associated with ACS such as obesity and diabetes, means public and private health care resources to prevent and treat ACS are expected to come under significant pressure in the near future. To mitigate these pressures, investment in cost effective programs should be undertaken now to improve effectiveness and efficiency of ACS treatment in the future. The first step should be to invest in the collection and dissemination of information and data associated with treatment across Australia at a local level. Informed analysis should then be undertaken to identify differences in treatment paths, to determine optimal therapies, and to inform best practice. The goal of ACS management should be to shift resources to cost effective technologies, thereby improving the efficiency of ACS treatment and generating greater health benefits for the Australian community. To ensure any gains made in the hospital are not undone once the patient steps out the hospital door, monitoring of health outcomes, individual behaviours, and the effectiveness of rehabilitation also needs to be measured, continually monitored and supported. Access Economics June 2009

8 Contents Executive Summary The epidemiology of ACS in Australia Definition of ACS Development of ACS Risk factors and comorbidity associated with ACS Projected number of ACS events in Australia Impact of demographic ageing Direct health care system costs Methodology Direct health care system costs Trends in direct health care system costs Direct health care system cost per separation Indirect costs associated with ACS Productivity losses Cost of informal care Private costs associated with rehabilitation Deadweight loss associated with public funding of health care Burden of disease Methodology used for measuring and valuing the burden of disease Burden of disease from ACS Burden of disease comparisons Summary of costs The future of ACS management A multidisciplinary approach to ACS care A national ACS registry Rehabilitation Next generation antiplatelet agents Appendix A: Epidemiology estimates and projections References Charts Chart 1.1 : Share of CHD deaths by risk factors Chart 1.2 : Share of CHD DALYs by risk factors Chart 1.3 : Risk of AMI associated with exposure to multiple risk factors Chart 1.4 : Reduced risk of AMI associated with healthy behaviour Chart 1.5 : Trends in daily smoking for those aged 14 years and over Chart 1.6 : Prevalence of overweight and obese people in Australia Chart 1.7 : Trend in blood pressure amongst Australians aged 25 to

9 Chart 1.8 : Trends in diabetes within Australia Chart 1.9 : Actual and projected age standardised separation rates for AMI Chart 1.10 : Actual and projected age standardised separation rates for unstable angina Chart 1.11 : Actual and projected age standardised separation rates for ACS Chart 1.12 : Projected age standardised separation rates by condition Chart 1.13 : Projected age standardised separation rates for ACS Chart 1.14 : Comparison of projected ACS separations in Australia Chart 1.15 : Projected male separations in Australia Chart 1.16 : Projected female separations in Australia Chart 1.17 : Projected total separations in Australia by condition Chart 1.18 : Projected total separations in Australia by gender Chart 1.19 : Share of AMI and angina pectoris across states and territories Chart 1.20 : 28 day case fatality following AMI Chart 1.21 : Actual and projected deaths following AMI Australia Chart 1.22 : Projected Australian population age structure Chart 1.23 : Projected ACS separations in Australia Chart 2.1 : Distribution of direct health care system costs of ACS Chart 2.2 : Direct health care system costs of ACS by expenditure type Chart 2.3 : Direct health care system costs of AMI by expenditure type Chart 2.4 : Direct health care system costs of unstable angina by expenditure type Chart A.1: Male AMI separation rates and trends Chart A.2: Female AMI separation rates and trends Chart A.3: Male unstable angina separation rates and trends Chart A.4: Female unstable angina separation rates and trends Tables Table 1.1 : Definition of ACS used in this study Table 1.2 : Prevalence distributions for seven lifestyle risk factors by age and sex Table 1.3 : ACS age standardised separations per 100,000 in the Perth Statistical Division Table 1.4 : Projected deaths following AMI by age bracket Australia Table 2.1 : Projected direct health care system costs by age and gender Table 2.2 : Projected direct health care system costs, by expenditure type Table 2.3 : Patient days associated with unstable angina and AMI Table 2.4 : Trend in direct health care system costs associated with ACS Table 2.5 : Direct health care system costs per separation

10 Table 3.1 : Productivity loss due to premature death Table 3.2 : Productivity loss due to working days lost Table 4.1 : Value of YLDs associated with ACS Table 4.2 : YLLs from ACS Table 4.3 : Burden of disease in Australia Table 5.1 : Summary of separations, deaths and costs Table 6.1 : Factors that impact on health and health outcomes Table 6.2 : Recommended medications for ACS treatment Table 6.3 : Status of new antiplatelet agents Table A.1: Male AMI age standardised separations per 100, Table A.2: Female AMI age standardised separations per 100, Table A.3: Male unstable angina age standardised separations per 100, Table A.4: Female unstable angina age standardised separations per 100, Table A.5: Male ACS age standardised separations per 100, Table A.6: Female ACS age standardised separations per 100, Table A.7: Actual and projected ACS separation rates for males Table A.8: Actual and projected ACS separation rates for females Figures Figure 1.1 : Defining ACS over time Figure 6.1 : A model of care for rehabilitation Figure 6.2 : Signalling pathways that activate platelets... 67

11 Glossary ABS ACE ACS AIHW AMI BMI CHD CRA CVD DALY DBP DoFD DoHWA DWL ECG ESC-ACC EMS GBD IHD LLA NHMRC NSTEACS NSTEMI PBAC PCI QALY SBP STEMI TRA VSLY YLD YLL Australian Bureau of Statistics Angiotensin-converting enzyme Acute coronary syndrome Australian Institute of Health and Welfare Acute myocardial infarction Body Mass Index Coronary heart disease Comparative risk assessment Cardiovascular disease Disability adjusted life year Diastolic blood pressure Department of Finance and Deregulation Department of Health Western Australia Deadweight loss Electrocardiogram European Society of Cardiology and the American College of Cardiology Emergency medical services Global Burden of Disease Ischemic heart disease Lipid-lowering agents National Health and Medical Research Council Non-ST-segment elevation acute coronary syndrome Non-ST-segment elevation myocardial infarction Pharmaceutical Benefits Advisory Committee Percutaneous coronary intervention Quality-adjusted life year Systolic blood pressure ST-segment elevation infarction Thrombin receptor antagonist Value of a statistical life year Years of health life lost due to disability Years of healthy life lost due to premature death

12 Definitions Acute coronary syndrome Acute myocardial infarction Angina pectoris Burden of disease Deadweight loss Direct health care system costs Event Health capital Indirect costs Myocardial infarction Separation Separation rate Stable angina Thrombus Unstable angina An umbrella term for conditions resulting from sudden insufficient blood supply to the heart. These include chest pain (unstable angina) and heart attack (AMI). A sudden insufficient blood supply to the heart muscle (myocardium) occurring because of blocked or narrowed arteries. Shown on an ECG as a Non-ST-segment elevation myocardial infarction (NSTEMI) or a ST-segment elevation myocardial infarction. A heart attack. The medical term for chest pain that is due to coronary heart disease. It is a symptom of acute myocardial infarction. Described as uncomfortable pressure in the centre of the chest. Manifested as stable angina or unstable angina. The impact of a disease or condition on the health and mobility of an individual. Inefficiencies created in the economy through distortions created by increased taxes to fund public health care. Public and private costs directly associated with the provision of health care. The occurrence of unstable angina or AMI. It can include a separation, death, or separation and death. The stock of human capital that produces health. This can depreciate with age and ill health, or increase with investment (such as exercise). Costs to the economy associated with flow on effects from reduced health and mobility, such as productivity loss and informal care costs. Reduced blood flow causing damage to the heart muscle. Heart attack. An admitted patient episode of care. A period of hospitalization. The number of separations compared to the number of individuals within the relevant population. Chest pain and discomfort that is instigated by stress or exercise, most commonly caused when the heart is working hard, but not getting enough blood and oxygen. A blood clot that forms inside a blood vessel or cavity of the heart. Reduced blood flow to the heart muscle causing severe chest pain but without damage to the heart muscle. It is usually unexpected and usually occurs at rest.

13 1 The epidemiology of ACS in Australia Coronary heart disease (CHD) (also known as ischemic heart disease) is one of the major causes of morbidity in Australia and the largest single cause of death, accounting for around 23,570 deaths in 2005 (AIHW 2007; 2008). It is associated with significant cost to the health care system, individuals, and society in general (Access Economics 2005). Acute coronary syndrome (ACS) is a sub-group of CHD and is associated with unstable angina and acute myocardial infarction (AMI). It includes clinical presentations that span ST-segmentelevation 1 myocardial infarction to an accelerated pattern of angina without evidence of necrotic damage to the heart muscle (myonecrosis) (Chew et al 2005). The common underlying cause of ACS is a build up of cholesterol plaque on the inside of the arteries of the heart muscle (known as atherosclerosis), causing the muscle cells to enlarge and form a hard cover over the area. This narrows the artery, reducing blood supply (and hence oxygen) to the heart. Under normal conditions blood flow may still be adequate but may be insufficient when an elevated blood flow is required (for example, through exercise). This is known as stable angina and is not considered part of ACS. However, if the plaque ruptures from the artery wall it can cause a blood clot within the artery, significantly reducing blood flow or completely blocking blood flow to the heart muscle. This can cause the sudden onset of angina (unstable angina) leading to severe chest pain and potential damage to the heart muscle (acute myocardial infarction). Death can occur if blood flow is not quickly restored to the heart muscle through the use of drugs or catheter procedures. 1.1 Definition of ACS ACS is defined across a range of acute myocardial ischemic states. It encompasses unstable angina, non-st-segment elevation myocardial infarction (NSTEMI), and ST-segment elevation myocardial infarction (STEMI) (Grech and Ramsdale 2003a). Figure 1.1 shows the definition of ACS components over time. An initial electrocardiogram (ECG) is conducted to determine whether ST-segment-elevation is present. If at a hospital, myocardial biomarker levels will also be tested. The ECG results and the myocardial biomarker levels will determine the diagnosis and the treatment path. If STEMI is present on the ECG, patients are diagnosed as having an AMI (heart attack) requiring urgent reperfusion. If ST elevation is not present, then patients may be diagnosed as having a NSTEMI (if biomarkers are elevated) or unstable angina (if biomarkers are not elevated). 1 Recording of electrical activity of the heart over time using an electrocardiograph 12

14 Figure 1.1: Defining ACS over time Source: Aroney et al (2006) For the purposes of this study, ACS is defined as patients that are diagnosed with unstable angina and AMI. Referring to the World Health Organisation (WHO) ICD-10 codes, ACS incorporates I20.0 for unstable angina (a sub-set of angina pectoris) and all sub-sets within I21 (WHO 2007). These are outlined in more detail in Table

15 Table 1.1: Definition of ACS used in this study ICD-10 Group Sub-group I20.0 Unstable angina Angina: I21.0 Acute transmural myocardial infarction of anterior wall I21.1 Acute transmural myocardial infarction of inferior wall I21.2 Acute transmural myocardial infarction of other sites I21.3 Acute transmural myocardial infarction of unspecified site I21.4 Acute subendocardial myocardial infarction I21.9 Acute myocardial infarction, unspecified Crescendo De novo effort Worsening effort Intermediate coronary syndrome Preinfarction syndrome Transmural infarction (acute) (of): Anterior (wall) NOS Anteroapical Anterolateral Anteroseptal Transmural infarction (acute) (of): Diaphragmatic wall Inferior (wall) NOS Inferolateral Transmural infarction (acute) (of): Apical-lateral Basal-lateral High lateral Lateral (wall) NOS Posterior (true) Posterobasal Posterolateral Posteroseptal Septal NOS Transmural myocardial infarction NOS Nontransmural myocardial infarction NOS Myocardial infarction (acute) NOS Notes: (a) Includes: myocardial infarction specified as acute or with a stated duration of 28 days or less from onset. Excludes: certain complications following AMI: I25.2, I25.8, I22-I24.1. Source: WHO (2007) 1.2 Development of ACS ACS begins with a fracture in the protective fibrous cap of an atheromatous plaque (Libby 2001). When these plaques fissure or rupture and core constituents such as lipid, smooth muscle and foam cells are exposed, it leads to the local generation of thrombin and deposition of fibrin (Grech and Ramsdale 2003a). This promotes platelet aggregation and adhesion and the formation of intracoronary thrombus (Grech and Ramsdale 2003b). Downstream 14

16 embolisation from friable coronary thrombus may occur, leading to focal cell necrosis and the release of cardiac troponins (Heeschen et al 1999). STEMI usually occurs when thrombus forms on a ruptured atheromatous plaque and blocks an epicardial coronary artery. Patient survival depends on several factors, the most important being restoration of blood flow, the time taken to achieve this, and the sustained patency of the affected artery (Grech and Ramsdale 2003b). NSTEMI is a form of myocardial infarction and these types of patients differ from STEMI only through the absence of ST elevation on the presenting ECG. Although there is no universally accepted definition of unstable angina, it has been described as a clinical syndrome between stable angina and AMI (Grech and Ramsdale 2003a). Unstable angina can be recognised by ischemic-type chest pain that is more frequent, severe or prolonged than the patient s usual angina symptoms, occurs at rest or minimal exertion, or is difficult to control with drugs. Recent onset angina is also classified as unstable (Maynard et al 2000). 1.3 Risk factors and co-morbidity associated with ACS Most known risk factors of ACS can be reduced by specific preventative methods such as pharmacotherapy and lifestyle changes (Patel and Adams 2008). These include smoking, high blood cholesterol, physical inactivity, diabetes, high blood pressure, being overweight or obese, and depression and social isolation (Heart Foundation 2009). However, there are also some risk factors of ACS that cannot be reduced, namely age, gender (being male) and a family history of coronary heart disease (Heart Foundation 2009). As part of the Global Burden of Disease (GBD) Study, the World Health Organization (WHO) developed a method for risk quantification to assess the health implications of certain risk exposures and provide a degree of conceptual and methodological consistency and comparability across risk factors (Ezzati et al 2004). Using this methodology, which was established as part of the Comparative Risk Assessment (CRA) study, Vos and Begg (2007) determined that seven risk factors explain 81.5% of CHD deaths and 85.2% of CHD disability adjusted life years (DALYs). 2 Although their study did not specifically focus on ACS, unstable angina and AMI make up around 57% of Australian separations associated with CHD. 3 As such, risk factors associated with CHD outlined by Vos and Begg (2007) can be used as a good proxy for ACS. Table 1.2 provides the prevalence distributions of the seven modifiable risk factors recognised by the WHO s CRA study as having an impact on the prevalence of CHD. Blood pressure, cholesterol, Body Mass Index (BMI) and fruit and vegetable intake are reported at their mean levels (and standard deviations) in the Australian population. Physical inactivity, tobacco and alcohol are provided as the percentage of the Australian population that falls into each category. 2 A DALY is a summary measure of health developed as the measurement unit to quantify fatal and non-fatal health outcomes, labelled the burden of disease and injury, on populations around the world for the Global Burden of Disease Study (Murray and Lopez, 1996). DALY weights are measured on a scale of zero to one, where a zero represented a year of perfect health and one represented death. Other health states associated with specific conditions are attributed values between zero and one. For example, a DALY weight of for unstable angina means a patient who has unstable angina has lost 23.8% of their total health. 3 Derived from accessed 03 April

17 Table 1.2: Prevalence distributions for seven lifestyle risk factors by age and sex 2003 Blood Pressure (mmhg) Cholesterol (mmol/l) BMI (kg/m 2 ) Fruit and vegetable intake (g/day) Physical inactivity (% population) Tobacco (% population) M F M F M F M F M F M F Mean SD Mean SD Mean SD Mean SD Vig Mod Insuff Inact Smoker Nonsmoker Abstain Alcohol (% population) Low Hazard Harmful Note: Vig = Vigorous, Mod = Moderate, Insuff = Insufficient, Inact = Inactive, Hazard = Hazardous Source: Vos and Begg (2007) The relative impact of each of the seven risk factors on CHD deaths and DALYs are illustrated in Chart 1.1 and Chart 1.2 respectively. 4 Blood pressure and cholesterol levels have the greatest influence on the number of deaths attributable to CHD, and alcohol and tobacco have the lowest. In contrast, cholesterol levels and blood pressure also have the most significant impact on DALYs, but tobacco and low fruit and vegetable intake have the smallest effect. 4 The actual percentages are reported in the sections below. 16

18 Chart 1.1: Share of CHD deaths by risk factors 2003 Blood Pressure Cholesterol BMI Low fruit and vegetable Physical inactivity Tobacco Alcohol Source: Vos and Begg (2007) and Access Economics calculations Chart 1.2: Share of CHD DALYs by risk factors 2003 Blood Pressure Cholesterol BMI Low fruit and vegetable Physical inactivity Tobacco Alcohol Source: Vos and Begg (2007) and Access Economics calculations Similar findings on the contribution of risk factors to CHD morbidity and mortality presented in Vos and Begg (2007) have been found throughout the world. In a study on potentially modifiable risk factors associated with AMI in 52 countries (including developed and less developed countries), Yusuf et al (2004) found that tobacco consumption and high cholesterol were the two strongest risk factors, followed by psychosocial factors, abdominal obesity, history of hypertension, and history of diabetes. Daily consumption of fruit and vegetables, moderate to strenuous exercise and consumption of alcohol more than three times per week reduced the risk of AMI. The odds ratio associated with exposure to multiple risk factors and 17

19 the reduction in risk associated with healthy activities from the study are shown in Chart 1.3 and Chart 1.4 respectively. Chart 1.3: Risk of AMI associated with exposure to multiple risk factors Note: Smk = Smoking, DM = diabetes mellitus, HTN = hypertension, Obes = Abdominal obesity, PS = Psychosocial, RF = Risk factors. The odds ratios are based on current vs never smoking, top vs lowest tertile for abdominal obesity, and top vs lowest quintile for ApoB/ApoA1. Source: Yousef et al (2004) Chart 1.4: Reduced risk of AMI associated with healthy behaviour Note: Smk = Smoking, Fr/vg = fruits and vegetables, Exer = Exercise, Alc = Alcohol. Odds ratios are adjusted for all risk factors Source: Yousef et al (2004) 18

20 1.3.2 Tobacco consumption Tobacco consumption, particularly the human carcinogens and other toxic properties inhaled through cigarette smoking, is causally related to an increased risk in mortality from many medical conditions, including CHD (Ezzati and Lopez 2004). This link has also been established in reverse, with the risk of AMI and death from CHD decreasing by half one year after quitting, and, after several years, approaching that of non-smokers (Patel and Adams 2008). Vos and Begg (2007) estimated that tobacco consumption accounts for 1.5% of CHD deaths and 1.2% of CHD DALYs in Australia. The smoking rate for the Australian population has been steadily declining within the last 50 years. More recently, between 1985 and 2007 the prevalence of daily smoking declined by around 15% and 11% for males and females respectively (AIHW 2008b). Trends in daily smoking for those aged 14 years and over are shown in Chart 1.5. In 2006, Australia had the second lowest prevalence rate of smoking amongst OECD (Organisation for Economic Cooperation and Development) countries at 16.8%, the lowest being Sweden (AIHW 2008b). Chart 1.5: Trends in daily smoking for those aged 14 years and over Source: AIHW (2008b) High cholesterol Cholesterol is a fat-like substance produced by the body which is found in the blood stream and all other parts of the bodies including organs and nerve fibres. Most cholesterol in the body is made by the liver from a variety of foods, but especially from saturated fats. The main factors that can influence an individual s level of cholesterol include a diet high in saturated fat content, heredity, and various metabolic conditions such as type II diabetes (Lawes et al 20024b). Cholesterol is thought to accelerate atherosclerosis, and thus influence CHD. However, the exact process remains uncertain. Clear and consistent positive associations between CHD and cholesterol level have been observed in cohort studies, and clinical trials of cholesterol lowering treatments have provided evidence of reversibility (Lawes et al 2004b). 19

21 Cholesterol is defined as total serum cholesterol expressed in millimoles per litre of blood (mmol/l). Vos and Begg (2007) estimate that high cholesterol accounts for 10.1% of CHD deaths and 5.3% of CHD DALYs in Australia. Average blood cholesterol levels of adults aged between 25 and 64 years were relatively unchanged between 1980 and 2000 (AIHW 2008b) Body mass index The body mass index (BMI) provides a general relationship between weight and health (James et al 2004). Excessive body-weight gain results in abnormalities in blood lipids, leading to an increased risk of developing CHD. In particular, the distribution of body fat appears to be an important determinant of the risk of coronary disease and death as patients with abdominal obesity experience the greatest risk (Krauss and Winston 1998). Vos and Begg (2007) estimated that elevated BMI accounted for 3.7% of CHD deaths and 2.5% of CHD DALYs in The prevalence of overweight and obese people in Australia continues to increase. In around 2.5 million adults were obese and a further 4.9 million were estimated to be overweight but not obese (AIHW 2008a). This means around 7.4 million people were estimated to have been above the BMI associated with healthy weight. Trends in overweight and obesity prevalence between 1995 and are shown in Chart 1.6. In nearly every age bracket there has been a steady increase in the prevalence of overweight and obese people in Australia. 5 A recent study on adults attending general practice shows that the prevalence of overweight and obese people in Australia has increased from 51% in to 58.5% in (AIHW 2008b). Chart 1.6: Prevalence of overweight and obese people in Australia Source: AIHW (2008a) 5 Prevalence of overweight and obese people aged between 65 and 74 years decreased slightly between 2001 and

22 1.3.5 Hypertension It is generally accepted that blood pressure plays a significant role in accelerating atherosclerosis of the blood vessels and thereby increasing the risk of cardiovascular disease (Lawes et al 2004a). A variety of prospective cohort studies and overviews have demonstrated a strong, continuous temporal association between blood pressure and CHD (APCSC 2003; MacMahon and Rodgers 1993). The standard unit for measuring blood pressure is mmhg and each 10mmHg below-usual SBP is associated with a 26% (95% confidence interval of 24-29%) lower risk of CHD (Lawes et al 2004a). According to Vos and Begg (2007), high blood pressure accounts for 10.7% of CHD deaths and 4.8% of CHD DALYs in Australia when analysed independently from the other risk factors. In recent years blood pressure amongst Australians has been trending down. AIHW (2008b) notes that the prevalence of high blood pressure in males and females aged between 25 and 64 years has more than halved. This is shown in Chart 1.7. Chart 1.7: Trend in blood pressure amongst Australians aged 25 to 64 Source: AIHW (2008b) Diabetes Diabetes mellitus is a chronic metabolic disease resulting from reduced levels of insulin in the blood, or through ineffective insulin. The consequence is a high level of glucose in the blood that can lead to a number of conditions, including CHD. People with diabetes are much more likely to have disability from cardiovascular disease than those without diabetes (Franklin et al 2004). According to Vos and Begg (2007), diabetes accounts for around 0.6% of the disability associated with CHD and 3.6% of the years of healthy life lost. Furthermore, around 2.1% of the total burden of disease associated with CHD was attributed to diabetes. 21

23 The prevalence of diabetes in Australia has increased significantly in the last 20 years. Based on National Health Survey, prevalence has increased from around 1.3% of the population in to 3.4% in (AIHW 2008b). This equates to around 700,000 Australians with diabetes in Given the trends in the number of people with diabetes and the growth in the highest risk age bracket (65-74 years) due to demographic ageing, prevalence of diabetes could increase significantly in the future, with subsequent impacts on the incidence of ACS. Trends in the prevalence of diabetes are shown in Chart 1.8. Chart 1.8: Trends in diabetes within Australia Source: AIHW (2008b) Alcohol Alcohol consumption is linked to long-term biological and social consequences through three outcomes: intoxication, dependence and direct biochemical effects. The direct biochemical effects can influence IHD in both a beneficial and harmful way. Moderate alcohol consumption reduces plaque deposits in arteries, promotes blood clot dissolution and protects against blood clot formation (Zakhari 1997). On the other hand, alcohol increases the risk of high blood pressure (Apte et al 1997) and hormonal disturbances (Emanuele and Emanuele 1997). Consequently, when estimating the burden of alcohol consumption, the overall deaths attributable to alcohol is an underestimation of the true relationship between alcohol consumption and IHD (Rehm et al 2004). Vos and Begg estimated that the net impact of alcohol consumption accounts for 0.8% of CHD deaths and 2.3% of CHD DALYs Fruit and vegetable intake Studies have found that fruit and vegetables provide a protective effect against ischemic heart disease (IHD) (Law and Morris 1998; Ness and Powles 1997). In particular, numerous studies have consistently shown that individuals who eat more fruits and vegetables have a reduced risk of AMI (Rimm et al 1996). The mean dietary intake of fruit and vegetables (excluding potatoes) is estimated to be 600g/day in adults, 480g/day in children aged 5 to 14 years, and 330g/day in children aged 0 to 22

24 4 years (Lock et al 2004). Vos and Begg (2007) estimated that low fruit and vegetable consumption accounts for 2.4% of CHD deaths and 1.4% of CHD DALYs in Australia Physical activity The apparent protective effect of being more active has been extensively documented with a significant amount of literature quantifying and qualifying the role of physical inactivity as a risk factor of CHD (Bull et al 2004). There is evidence of a strong inverse correlation of leisure time activity and energy expenditure, habitual exercise and fitness with risk of coronary disease and death (Patel and Adams 2008). The effect appears to be proportional to energy expenditure; the greater the degree of physical activity the lower the risk of coronary events. Vos and Begg (2007) estimated that physical inactivity accounts for 6.6% of CHD deaths and 3.4% of CHD DALYs in Australia. Data from the National Health Surveys for 1995, 2001, and show there has been little change in the level of physical activity in the Australian population. The proportion of adults (18 years and over) that undertook less than 100 minutes of exercise in the two weeks prior to the surveys has fluctuated between 30% and 35% (AIHW 2008b) Depression Depression has been recognised as a common co-morbidity among cardiac patients and an independent predictor of adverse outcomes (Amin et al 2008; Reddy et al 2008). Approximately 20% of patients with a recent ACS have major depression, and almost 20% have minor depression (Carney and Freedland 2008). Numerous studies have documented that depression in patients with ACS is associated with a higher incidence of mortality, recurrent cardiovascular events, and healthcare utilisation (Rozanski et al 2005). Parker et al (2008) determined that only depressive episodes that commenced after an ACS admission were associated with a poorer cardiovascular outcome. Amin et al (2008) found that reduced levels of omega-3 fatty acids in blood cell membranes, an emerging risk factor for both ACS and depression, could help explain the relationship between depression and adverse ACS outcomes. Furthermore, other studies have found psychobiological processes to underlie the emotional triggering of ACS (Steptoe and Brydon 2009). Patients with advanced atherosclerosis may be triggered into ACS by acute anger, stress and depression. Vos and Begg estimated that CHD accounts for 3.3% of DALYs attributed to depression (Vos and Begg 2007). 1.4 Projected number of ACS events in Australia There are two primary Australian data sources on ACS treatment that were available. The AIHW provides an estimate of the number of separations by event type, gender, and 10 year age brackets, with the latest data being (AIHW 2009). However, there is a possibility that this data may over estimate the real number of separations because it does not adjust for transfers and readmissions related to the same event. The second data source is based on population-based linkage of health records in the Perth Statistical Division, Western Australia. The linked data is created by determining connections between core Department of Health Western Australia (DoHWA) data collections and other administrative data sources and research collections, based on probabilistic linkage created through the use of similar demographic information(for example, name, sex, date of birth and address). The data collections linked for the purposes of this study are hospital admissions, 23

25 emergency presentations, and death records associated with the Registry of Births, Deaths, and Marriages. ACS separation rates per 100,000 people based on 28 day episodes within the Perth Statistical Division are shown in Table 1.3. A breakdown into AMI and unstable angina is shown in Appendix A. The data are derived from hospital admissions regarding diagnoses for AMI or unstable angina in any diagnostic field and includes fatal and non-fatal cases. They relate to residents of Perth aged 35 to 79 years and do not necessarily include persons admitted to hospitals in Perth. Male Table 1.3: ACS age standardised separations per 100,000 in the Perth Statistical Division , , , , , , , , , , , , , , , , , , , , , , , , , , , , Female , , , , , , , , , , , , , , , , Source: Emeritus Professor Michael Hobbs, pers. com. 07 May 2009 As patient records are linked, the Western Australian data linkage information has the capacity to avoid the inflationary effects of transfers and readmissions as it allows a patient to be followed within the hospital system. To provide an estimate of the number of separations within Australia, separation rates for AMI and unstable angina derived from the linked data were applied to the Australian population by age bracket and gender. These were then compared to the separation data supplied by AIHW. Although the data accounts for re-admissions within a 28 day period there are some limitations in their use for this study. As the most recent data are for 2004, and there is a downward trend apparent in the data between 1998 and 2004, using 2004 data is likely to overestimate the number of separations for In order to adjust for possible over 24

26 estimation, linear trends were projected to 2009 (by age group, condition, and gender) to estimate a more recent measure of separation rates. Actual and projected age standardised separation rates for AMI, unstable angina and ACS for age groups 35 to 79 years are shown in Chart 1.9, Chart 1.10 and Chart 1.11 respectively. In summary: AMI in males has decreased from 425 separations per 100,000 in 1998 to 362 separations per 100,000 in It is projected that AMI in males will be approximately 326 separations per 100,000 in AMI in females has decreased from 164 separations per 100,000 in 1998 to 147 separations per 100,000 in It is projected that rates for AMI in females will be approximately 145 separations per 100,000 in Unstable angina in males has decreased from 518 separations per 100,000 in 1998 to 382 separations per 100,000 in It is projected that unstable angina in males will be approximately 275 separations per 100,000 in Unstable angina in females has decreased from 252 separations per 100,000 in 1998 to 184 per 100,000 in It is projected that unstable angina in females will be approximately 109 separations per 100,000 in ACS in males has decreased from 943 separations per 100,000 in 1998 to 744 per 100,000 in It is projected that ACS in males will be approximately 601 separations per 100,000 in ACS in females has decreased from 416 separations per 100,000 in 1998 to 331 separations per 100,000 in It is projected that ACS in males will be approximately 254 separations per 100,000 in The faster decline in unstable angina compared to AMI is consistent with the hypothesis purported by Sanfilippo et al (2008) that the uptake of troponin testing between 1998 and 2004 has increased diagnosis of AMI at the expense of unstable angina. According to Emeritus Professor Michael Hobbs (pers. comm. 06 May 2009) it is likely that this trend will continue. However, there is evidence that separation rates for ACS are still declining (as shown in Chart 1.11), which is consistent with the long term decline in 28 day case fatality following AMI (Sanfilippo et al 2008) and CHD (AIHW 2006). The declining trend in age standardised separation rates for ACS is consistent with AIHW findings that separation rates for CHD have been declining since its peak in the late 1960s (AIHW 2009a). Specifically, AIHW (2009a) showed that CHD rates were separations per 100,000 and separations per 100,000 for males and females respectively in 1968, but had declined significantly to and 76.6 in

27 Age standardised rates per 100,000 Age standardised rates per 100,000 The economic costs of heart attack and chest pain (Acute Coronary Syndrome) Chart 1.9: Actual and projected age standardised separation rates for AMI 500 Male 400 Female Actual rates Projected rates Note: Based on age groups between 35 and 79 years old in Perth Statistical Division, Western Australia Source: Emeritus Professor Michael Hobbs, pers. comm. 07 May 2009 and Access Economics calculations Chart 1.10: Actual and projected age standardised separation rates for unstable angina Male Female Actual rates Projected rates Note: Based on age groups between 35and 79 years in Perth Statistical Division, Western Australia Source: Emeritus Professor Michael Hobbs, pers. comm. 07 May 2009 and Access Economics calculations 26

28 Age standardised rates per 100,000 The economic costs of heart attack and chest pain (Acute Coronary Syndrome) Chart 1.11: Actual and projected age standardised separation rates for ACS 1, Male Female Actual rates Projected rates Note: Based on age groups between 35 and 79 years old in Perth Statistical Division, Western Australia Source: Emeritus Professor Michael Hobbs, pers. comm. 07 May 2009 and Access Economics calculations One other limitation with the WA linked data for this study was that separation rates were available only for those between 35 and 79 years. Although AIHW data suggests that less than one per cent of all separations for ACS are for those aged less than 35 years (AIHW 2009), evidence suggests that the burden of cardiovascular disease falls particularly heavily on those above the age of 80 (Begg et al 2007; Vos and Begg 2007). Consequently leaving these age groups out of the analysis would underestimate the true number of ACS separations in Australia, and underestimate the costs associated with those separations. In order to capture separation rates for patients over the age of 79, separation rates between 25 and 79 years were fitted with trends (by condition and gender) and then projected to age groups beyond 79 years 6. Separation rates and fitted trend lines are shown in Appendix A. Projected separation rates for AMI and unstable angina are shown in Chart 1.12 and projected separation rates for ACS are shown in Chart In summary, age standardised separation rates associated with: AMI (males and females) and unstable angina (females) follow an exponential growth curve with separation rates significantly increasing beyond the age of 80 years; unstable angina (males) follow a polynomial curve, increasing with flatter growth (compared to AMI) beyond the age of 80 years; AMI are larger for males compared to females; unstable angina are larger for males between 35 to 94 years, but female rates become larger than males for 95 years and above; and 6 Unfortunately separations recorded by AIHW are truncated at age 85+ so a comparison of projected separation rates used in this study for those 85+ could not be made. 27