Article Text

Economic burden and the cost-effectiveness of treatment of cardiovascular diseases in Africa
1. T A Gaziano
1. Division of Cardiovascular Medicine, Brigham & Women’s Hospital, Boston, MA, USA
1. Thomas A Gaziano, Division of Cardiovascular Medicine, Brigham & Women’s Hospital, 75 Francis Street, Boston, MA 02115, USA; tgaziano{at}partners.org

## Abstract

Cardiovascular disease is the leading cause of death in those over the age of 45 in Africa. The economic toll from cardiovascular diseases is equally devastating, leading to billions of dollars lost due to healthcare costs and reduced productivity from the disabling and fatal outcomes related to diabetes, hypertension, stroke, valvular heart disease, and heart failure. Much of it is preventable. With reasonable screening programmes and judicious use of scarce resources much of the suffering can be alleviated. This article reviews the economic burden attributable to cardiovascular disease in Africa and many of the potential cost-effective solutions to the large burden. It further outlines many of the areas where we know less and must focus our future research in trying to outline cost-effective solutions.

## Statistics from Altmetric.com

At the beginning of the 20th century, cardiovascular disease (CVD) was responsible for fewer than 10 per cent of all deaths worldwide. Today, that figure is about 30 per cent, with about 80 per cent of the global burden of CVD death occurring in low-income and middle-income countries. In sub-Saharan Africa deaths due to CVD are projected to more than double between the years 1990 and 2020.1 While HIV/AIDS is the leading overall cause of death in this region, CVD is the second leading killer and is the first among those over the age of 45.2

In addition to the large health burden of CVD, there is an enormous financial burden as well. This comes in the form of direct healthcare costs related to the prevention and treatment of CVD and its risk factors. These costs are carried by the individuals, governments, and the private sector. Furthermore, there are numerous indirect costs related to CVD. These include the lost productivity of workers struck by valvular heart disease, stroke, heart failure, and ischaemic heart disease. Other costs include the lost savings and assets that are foregone when families must meet catastrophic healthcare expenditures such as stroke rehabilitation or dialysis, when available.

Compounding the significant health and financial burdens that CVD places on African countries are the limited resources that they have to devote to CVD. The amount of healthcare expenditure as a percentage of gross domestic product (GDP) for African countries is 6.3%, which puts it second only to Europe and Central Asia among the six World Bank low-income and middle-income regions. Again there is a wide range from 2.5% in the Republic of Congo to 12.9% of GDP in Malawi. The actual amount devoted to healthcare is summarised by the World Bank,3 which reports that sub-Saharan African countries spent on average about US$45 per capita on healthcare in 2004. While there is a great range across African nations from as little as$6 per capita in Ethiopia to as much as $390 per capita in South Africa, the amount is still quite small compared with US$ 3727 for high-income countries.

Unfortunately, the number of studies that report these costs is quite small, as few economic analyses relating to CVD have been conducted in Africa. This article reviews the limited literature with regard to the economic burden of CVD, with estimates from the global literature. Further, the article suggests guiding principles for managing hypertension, the leading risk factor of CVD mortality, in a cost-effective manner. Finally, the article reviews cost-effective strategies for managing CVD in Africa.

## COST OF CARDIOVASCULAR DISEASE

The cost of CVD can be divided into two types: direct and indirect. Direct costs are those costs related to the treatment and management of the disease itself. They include the running of clinics and hospitals, salaries for healthcare personnel, medications, rehabilitation where available, and medical supplies and testing. In addition, they include the patient’s time used to seek care. Many of these costs are placed directly on the patients in countries where there is limited government coverage for chronic conditions and their income is too limited to afford private insurance. In general the government’s proportion is approximately 40% of all healthcare expenditures, compared with nearly 75% in western European countries.

Less is known about how the direct costs of CVD care vary on a country-by-country basis. A study conducted in South Africa is the only one to estimate the cost of CVD in a sub-Saharan country.4 In 1991, the authors reported that the overall costs of heart disease and stroke were between $550 million and$700 million, which would be nearly double those figures today assuming a 4% rate of inflation. Using these values South Africans spent about $20 per capita on CVD care compared with$500 to $1000 per capita in developing countries.5 These figures did not include rehabilitation or follow-up costs. About 75% of the costs were assumed by the private sector. At that time the cost of CVD accounted for 2–3% of the GDP of the country.6 Only recently have economists begun to evaluate the indirect costs of cardiovascular diseases in developing countries, and, while the methodology has limitations, it also provides us some measure of the impact of ill health beyond the consumption of healthcare resources. Indirect costs from ill health include loss of income resulting from lost productivity or employment due to major disabilities, such as stroke or heart failure. Other sources of indirect costs include the loss of future earnings from assets that are sold off for chronic and/or catastrophic care, loss of income from other family members who must provide care,7 and the future loss of income from children who drop out of school to provide care for adult members with CVD. One potentially significant source of indirect costs related to CVD is productive work years lost. In A Race Against Time,8 its authors suggested that South Africa lost nearly 300 000 years of productive life among adults. This number is projected to increase by over 30% by the year 2030. The authors also point to one of the most concerning trends for developing countries. In low-income and middle-income countries, the age at which CVD mortality occurs is much younger. Currently South Africa has double the annual age-adjusted CVD mortality rate among men aged 35–64 (97/100 000) compared with the United States, and nearly triple for women (68.2/100 000). In the future, it is projected to worsen. For example, it is projected that, out of all CVD deaths in 2030 in South Africa, 41% of them will occur among those aged 35–64 compared with 12% in the United States, thus striking down those in the prime of their productive working years. The method of assessing indirect costs due to number of years lost during the normal working age assumes near-full employment rates. Given widely varying and often high unemployment rates in many African countries, the magnitude of the losses may be hard to quantify exactly. More recently, economists have sought a measure beyond productivity to assess the value of illness or loss of life. The full-income method seeks the value of health as a good in itself as opposed to just a means for increased production. Using the full-income method, the World Health Organization evaluated the economic costs attributed to CVD,9 taking into account not only worker productivity losses, but also the value that people place on health. Under these assumptions, economists estimated the projected loss of national income over a 10-year period for selected countries. From 2005 to 2015, Tanzania and Nigeria are estimated to lose US$2.5 billion and $7.6 billion respectively due to heart disease, stroke, and diabetes. ## COST-EFFECTIVENESS OF INTERVENTIONS There are few cost-effectiveness analyses that have been conducted at the country level. The majority of the estimates for the sub-Saharan African region come from the WHO-CHOICE (CHOosing Interventions that are Cost-Effective) programme and the Disease Control Priorities Project (DCP2) in Developing Countries.10 The estimates provided in both of these analyses assumed average costs across the region, given that there are few countries in the region with exhaustive country-level information for many of the cost inputs. The interventions are divided into population-based and person-based strategies. Most of these analyses evaluate the costs only in terms of direct costs related to the healthcare system. ## POPULATION-BASED Most population-based interventions require public promotion or policy changes at the state level to effect change. However, they can be further divided into two categories. The first category of interventions encompasses those that do not require interaction between individuals and healthcare providers. They may be cost-effective, even if the population effect is small, if the societal cost to achieve a change in one or more risk factors remains low as it is spread across the population. Unfortunately, the data on precise cost estimates for such mass interventions are limited. One of the analyses evaluated by the WHO-CHOICE programme was a combination of legislation, voluntary industry participation, and mass media to reduce salt consumption.11 This combination of interventions was found to have a cost-effectiveness ratio of about$48–60 per disability-adjusted life year (DALY) averted. However, the ratio is sensitive to the estimated costs expected to achieve the intervention and the blood pressure reduction anticipated from the interventions. Table 1 displays results of a cost-effectiveness analysis of a legislative and media campaign to reduce salt intake in the diet for sub-Saharan Africa assuming a range of the assumptions.12 The methods for the analysis are described in detail elsewhere using a model to reflect changes in blood pressure.13 Essentially, the model evaluates the likely reduction in blood pressure from reduced salt intake in both developing and developed countries1416 and the costs that are necessary to achieve that reduction through mass education and legislation. The INTERSALT14 study found that for a reduction of sodium chloride intake by 6 g there is about a 3–6 mm reduction in systolic blood pressure. The meta-analyses15 16 suggest that the reduction is graded depending on the starting blood pressure. The authors estimate that reductions in salt intake can lead to a reduction from 1 mm Hg systolic in non-hypertensives up to 5 mm Hg systolic in hypertensives. Using these estimates, the intervention could be cost-saving (reducing morbidity and mortality and reducing healthcare costs), or it could cost up to $600 per DALY averted, assuming a small reduction in blood pressure with a relatively large cost to achieve the intervention. Later analyses of INTERSALT suggest that perhaps even a greater reduction of up to 9 mm Hg in systolic blood pressure could be achieved with a reduction in daily sodium consumption of 100 mmol per day, further enhancing the cost-effectiveness of such an intervention.17 Table 1 Cost-effectiveness ratios (US$/DALY) of salt reduction depending on expected blood pressure reduction and cost to achieve through media and legislation

Table 2 lists the remaining population strategies for targeting risk factors for ischaemic heart disease, stroke and heart failure evaluated in DCP2. As with salt reduction, the results are published in ranges from low to high estimates based on underlying assumptions used in the models. Listed are the low and high estimates for each intervention from the DCP2 analyses18 for sub-Saharan Africa. The first tobacco intervention is a 33% price increase of tobacco through increased taxation on the product. This intervention has a range of ratios of $2–26/DALY averted depending on the expected price elasticity and benefits from smoking cessations assumed. The non-price interventions such as advertising bans, labelling of cigarette packaging and mass education campaigns have higher ratios of$42–570/DALY averted.

Table 2 Cost-effectiveness of population-based interventions

Finally, table 2 lists the cost-effectiveness ratios of substituting polyunsaturated fats for saturated and trans fats. There are two ranges based on different estimates of expected reduction in ischaemic heart disease (IHD). Initial estimates19 suggested that replacing 2% of energy intake from trans fat with polyunsaturated fat would reduce IHD by about 7%. A later experience in Poland, when government subsidies for animal-based fats were reduced, encouraging the uptake of vegetable-based oils, suggested the reduction could be up to 25%.20 21 Others report up to a 40% reduction in women.22 Based on these estimates and the range of costs to achieve such a reduction, these interventions could be anywhere from $53–1344/DALY averted, assuming the lower IHD reduction to be as low as cost-saving, or only$184/DALY averted, assuming the 40% reduction is achievable with the substitution. Both estimates assume $0.50 per capita to achieve the intervention at the low end up to$6 per capita to achieve the change.

### Risk factor guidelines

The other population category includes screening-based interventions, which require some interaction with individuals and the healthcare system, but are generally aimed at the entire adult population. While the screening is population-based, once individuals are identified as high-risk they require personal interventions for the management of their elevated risk. Two approaches to screening have been used globally: treating those above a certain threshold value for a single risk factor, or those above a certain absolute CVD risk based on multiple risk factors. These latter guidelines will be cost-effective, even with modest costs, if the approach targets high-risk individuals and relies on easily identifiable screening tools such as limited laboratory testing and opportunistic screening initially.

These strategies have been outlined in guidelines for hypertension and cholesterol. There have been few published guidelines regarding the management of hypertension in sub-Saharan Africa. The region as a whole published its most recent guidelines23 in 2003, and these share many similarities to the WHO-ISH guidelines.24 These guidelines in particular focus on an arbitrary cut-off value above which certain interventions, including pharmacological therapy, should be initiated. One problem with this approach is that the arbitrary cut-off point continues to change. Blood pressure is associated with a continuous and graded risk of stroke and ischaemic heart disease over a long range down to at least 115 mm Hg.25 Since there appears to be no clear cut-off at which the risk is zero above 115 mm Hg, then any level above it is arbitrary. Indeed, most of the disease burden resulting from blood pressure, lipids, and excess weight occurs in the large majority of the population with nonoptimal levels but without hypertension, hypercholesterolaemia, or obesity as defined by the arbitrary cut-offs in multiple guidelines.26

The choice of a cut-off has real policy and cost implications (fig 1). In sub-Saharan Africa, if the definition for treatment eligibility were 160/95 mm Hg, which was the cut-off for guidelines in South Africa27 in 1995, then approximately 4% of the adults over the age of 30 would be eligible for treatment. Under the current WHO-ISH guidelines, which use 140/90 mm Hg as a cut-off, nearly 22% of the population is eligible for treatment, which is nearly a sixfold increase. If the cut-off were eventually dropped to 120/80 mm Hg, which the United States Joint National Commission on Hypertension now defines as pre-hypertension, then nearly 70% of the adult population would be eligible for intervention. Even if the interventions do not include drug treatment, these are large increases in the proportion of the population that would require lifestyle interventions and follow-up, which could not be sustained given the current human resources devoted to healthcare in Africa.

Figure 1 The percentage of adults aged 30–74 in sub-Saharan Africa with estimated systolic blood pressures.

Not only are current guidelines based on targeted blood pressure levels likely to be infeasible to implement, but also they are likely to be cost-inefficient. This is because the major decision on whether to initiate interventions is based on only one risk factor, hypertension. However, the use of multiple risk factors to assess someone’s overall CVD risk is much more precise. For example, a 45-year-old man with a blood pressure of 150/90 mm Hg, who is a non-smoker, non-diabetic, with a total to HDL-cholesterol ratio of 4, has about a 2% chance over 10 years of developing CVD, but would ultimately be eligible for treatment according to the sub-Saharan African hypertension guidelines. In contrast, a 59-year-old man who is a smoker and has a total to HDL-cholesterol ratio of 6 but a blood pressure of 139/84 would not be treated according to the same guidelines. However, his risk of CVD over the next 10 years is 25%, 10 times that of the man who would be treated. Thus, blood pressure-targeted guidelines lead to potential costly over-treatment of those at relatively low risk and under-treatment of those at high risk for CVD. The same could be said for cholesterol guidelines based on LDL-cholesterol levels alone.

The cost-effectiveness of these two strategies, blood pressure thresholds and absolute risk thresholds, was estimated in a modelling study of the two approaches in South Africa.13 Both strategies assumed the standard use of medications, both generic and non-generic, at their current rate of use in 1998.28 The strategies based on blood pressure targets cost more and saved fewer lives than the strategies based on absolute risk. In that analysis, it was projected that the current South African guidelines cost an additional $30 million annually compared with an approach of treating those with a 10-year absolute risk of CVD of 15 per cent or greater. Further, the current approach based on a target blood pressure of 140/90 mm Hg would result in 5000 fewer quality-adjusted life years for the adult population over 10 years. If only generic agents were used, the strategy of treating all those with a 10-year risk of CVD >35% would be cost-saving compared with no drug use at all. There will be two challenges, however, for those implementing risk-based guidelines in African countries. The first will be to educate current physicians who were trained to target individual risk factors. The second will be to find simplified measures of assessing risk, given the high cost and impracticality of measuring cholesterol in many regions of Africa. This author, along with others, has been involved in evaluating the use of non-laboratory-based risk assessment tools using age, gender, and smoking status along with blood pressure to assess risk. ### Personal interventions The list of possible interventions for advanced cardiovascular disease in table 3 includes interventions for treating or reducing ischaemic heart disease, congestive heart failure, and stroke.29 Interventions regarding valvular heart disease have not been fully evaluated in developing countries. The cost-effectiveness analysis for congestive heart failure is based primarily on studies of left ventricular dysfunction from IHD and hypertensive heart disease and thus may not represent the large number of patients who suffer heart failure from the many other causes outlined in an earlier paper in this series.30 Interventions for rheumatic heart disease are discussed separately. Table 3 Personal interventions for CVD in sub-Saharan Africa ### Ischaemic heart disease While the incidence of acute myocardial infarction (AMI) is low compared with other developing regions, it is expected to rise over the next 15 years.1 For countries with facilities in place to make a rapid diagnosis of AMI, relatively cheap and effective interventions exist. The use of aspirin and β-blockers in the setting of AMI is relatively cost-effective with a ratio of$11/DALY averted. For urban centres or day hospitals with well-trained staff, the use of the generic streptokinase can cost about $600 per DALY averted in addition to the use of aspirin and beta-blockers. The use of patent-protected tissue-plasminogen activator (t-PA) is not an attractive option at$15 900 per DALY averted.

For survivors of AMI, analyses have assessed a two-drug regimen29 and a four-drug regimen.31 A regimen of aspirin and β-blocker taken for secondary prevention would be considered cost-saving in sub-Saharan Africa in a setting where hospitals were available. In a setting where hospital access was limited, secondary prevention with the same combination would cost $389 per DALY averted, since the intervention would reduce repeat AMI and mortality but would not reduce hospitalisation costs associated with such events. The use of a four-drug regimen comprising aspirin, a β-blocker, an angiotensin-converting enzyme inhibitor (ACEI), and statin would yield a cost-effectiveness ratio of$350/DALY averted compared with no secondary prevention. Use of the same four-drug regimen in patients without AMI but at global risk of greater than 15% would yield about $900/DALY averted. These results also assume limited access to hospitals, so the ratio may be even better in settings where hospital savings can be achieved. When coronary artery bypass graft surgery (CABG) in those with left-main coronary artery disease or three-vessel disease with reduced systolic function was compared with the four-drug regimen in secondary prevention, it had a cost-effectiveness ratio of nearly$27 000 per DALY in sub-Saharan Africa.

### Rheumatic heart disease

Most of the cost-effectiveness analyses on rheumatic heart disease have been conducted in a developed-country setting. There is general consensus that secondary prevention with benzathine penicillin injections is cost-effective.34 35 However, the debate regarding primary prevention is more controversial.34 Studies in the developed countries show high cost-effectiveness ratios for primary prevention.36 However, the western estimates are based on a much lower prevalence level of group A streptococcal infections than in developing countries such as sub-Saharan Africa, where the incidence may be 10-fold higher. Estimates suggest that primary prevention37 in a high endemic area, treatment of a sore throat with a single intramuscular penicillin injection, would prevent one case of rheumatic fever for about \$46. Until adequate cost-effectiveness analyses can be conducted regarding the use of clinical scoring mechanisms versus antigen testing or cultures or the use of the various treatment modalities, treatment decisions will need to be made at the local level.

### Conclusion

The economic burden of cardiovascular diseases in sub-Saharan Africa is significant. CVD will cost the continent billions of dollars in the next decade. Hypertension remains the number one cause of significant financial burden, including the cost of caring for stroke, ischaemic heart disease, congestive heart failure from hypertensive heart disease and systolic dysfunction. Peripheral vascular disease and end-stage renal disease requiring dialysis or transplantation were not covered here, but are likely also to be significant drivers of costs to the healthcare systems of African nations.

However, there exist several cost-effective interventions for the effective prevention, treatment, and management of cardiovascular diseases in Africa: first, appropriate risk assessment for those at high risk for cardiovascular diseases such as stroke, ischaemic heart disease and hypertensive heart disease. Less is known about the cost-effectiveness for certain valvular disorders, but it needs evaluation. Secondary prevention, and possibly primary prevention, of rheumatic heart fever could, however, go a long way towards preventing rheumatic valvular disease, a major cause of heart failure and disability in Africa.

## Footnotes

• Competing interests: None.

• Funding: Thomas A Gaziano, MD, MSc is funded through the Fogarty International Center grant from the National Institutes of Health: 1K01TW007141-01.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.