Report No. 32576-BR Brazil Addressing the Challenge of Non-communicable Diseases in Brazil November 15, 2005 Brazil Country Management Unit Human Development Sector Management Unit Latin America and the Caribbean Region Document of the World Bank ADDRESSINGTHE CHALLENGE OF NON-COMMUNICABLE DISEASESINBRAZIL TABLE OF CONTENTS 1 EXECUTIVE SUMMARY i 2 SCOPE OF THIS REPORT 1 3... ....................................................................................... THE HEALTHTRANSITION AND BURDENOFDISEASEINBRAZIL ..................................................................................... ......3 THEDEMOGRAPHIC TRANSITIONINBRAZIL..................................................................... 3 THEEPIDEMIOLOGICAL TRANSITIONINBRAZIL............................................................... 6 Burden of Disease........................................................................................................ 7 4. THE LINKBETWEENRISK FACTORSAND AVOIDABLE BURDENOF DISEASEINBRAZIL .................................................................................................... 10 Likely riskfactor trends............................................................................................. RISKFACTORPREVALENCEINBRAZIL-THEPRINCIPLEISSUEFORNCDPREVENTION 14 ....13 Smoking...................................................................................................................... 14 Overweight and Obesity ............................................................................................ 15 Healthy Diet .Fruits and Vegetables........................................................................ 16 Physical activity......................................................................................................... . . 16 Conclusions................................................................................................................ 16 PREVALENCEOFHYPERTENSIONAND DIABETESINBRAZIL ............................................ 17 Hypertension.............................................................................................................. 17 Diabetes ..................................................................................................................... 17 Heart Attacks and Coronary Disease........................................................................ 18 Conclusions................................................................................................................ 18 5. NCDSAND POVERTY INBRAZIL .................................................................... 19 Non-communicable Disease Risk Factors and Poverty............................................. 20 Further links between Poverty and NCDs................................................................. 22 Policy Imp1ications .................................................................................................... 23 6. PREVENTIONOFNCDS: INTERNATIONALEXPERIENCEAND CURRENT PROGRAMSINBRAZIL ......................................................................... 25 PRIMARYPREVENTIONOFMAJORNCDS-DECREASING EXPOSURETO RISK FACTORS ....26 Smoking, Physical Activity, Diet and Weight............................................................ 26 Public Health Policies: Anti-Smoking ...................................................................... 27 Public Health Policies to Support Daily Physical Activity ....................................... 29 Public Health Policies to Promote a Healthy Diet.................................................... 32 Comprehensiveprograms shown toprevent cardiovascular disease (CVD) and strokes........................................................................................................................ 33 Diabetes Prevention................................................................................................... 34 CARMEN Strategy..................................................................................................... 35 SECONDARY PREVENTIONOFMAJORNCD ....................................................................... 35 Hypertension.............................................................................................................. 35 1 Diabetes ..................................................................................................................... 37 Cardiovascular Disease (CVD)and Strokes............................................................. 37 Cost-effectiveness of Interventions ............................................................................ 38 Issues Related to Chronic Care................................................................................. 39 Limitations................................................................................................................. Brazil's Arterial Hypertension and Diabetes Mellitus Care Reform Plan................39 40 NCD/RISK FACTORPREVENTIONPOLICIESAND ACTIVITIES INBRAZIL.......................... 41 NCD and Risk Factor Surveillance ........................................................................... 42 Monitoring State and Municipal Perj4ormance on NCD Prevention and Control.....43 7 . FINANCIALAND ECONOMICIMPACT OFEXPANDINGKEY NCD PREVENTIONACTIVITIES -FOUREXAMPLES .................................................. 44 Introduction............................................................................................................... 44 Interventions Includedfor the Expanded NCD Prevention Program ....................... 44 Methodology .............................................................................................................. 44 Results........................................................................................................................ 47 DISCUSSION CONCLUSIONS ..................................................................................... AND 51 8 . CONCLUSIONSAND RECOMMENDATIONS ................................................ 53 RECOMMENDATIONS ....................................................................................................... 55 Develop and implement health strategies to prevent NCDs byfocusing on risk factors ........................................................................................................................ 55 Improve surveillance and monitoring of riskfactors and prevention strategies.......57 Improve secondaryprevention of NCDs through better health care and screening.57 Strengthen research on eflective NCD /riskfactor prevention................................. 58 TABLE8.1. RECOMMENDATIONS KEY STRATEGIESFORTHE SHORT AND MED~UM ON TERM.............................................................................................................................. 59 9 .................................................................................................................. 63 10 ANNEX 2 ..ANNEX 1 .................................................................................................................. 67 METHODS DATAFORECONOMIC AND ANALYSIS........................................................... 67 11 REFERENCES . ........................................................................................................ 73 .. 11 Figures Figure 3.1 Total Fertility Rate. Brazil and LAC. 1950-2000................................................ 3 3 Figure 3.3 Infant Mortality Rate. Brazil and LAC. 1950-2000 ............................................ Figure 3.2 Life Expectancy at Birth. Brazil and LAC. 1950-2000....................................... 3 Figure3.4 Proportion of death by age group Latin America 1950-2050.............................. 4 Figure 3.5 The rapid aging of Brazil's population: 2000-2050 ............................................ 5 Figure 3.6 Age Dependency Ratios. Brazil 2000-2040 ........................................................ 6 Figure 3.7 Epidemiological Transition inBrazil................................................................... . 6 Figure 4.1 Causal chain of illness ......................................................................................... 11 Figure6.1 Relationship between key riskfactors and NCDs ............................................... 25 Figure 7.1 Modelingthe impact of preventive interventions . Interventions risk factors and non-communicable diseases................................ . 45 - Boxes Box 6.1 Agita Siio Paulo ................................................................................................... 30 Tables 7 Table 3.2 Daily rates* and percent for major causes by region........................................... Table 3.1 YLL. YLD. DALY Rates* and Percent. Brazil 1998.......................................... 8 Table 3.3 Top 10 specific causes of DALYs Brazil 1998 ................................................... 8 Table 4.1 Avoidable DALYs (Brazil compared to Amer-A*) ............................................ 10 Table 4.2 Ten mainrisk factors for attributable mortality inthe Americas ........................ 12 Table 4.3 Ten main risk factors for attributable DALYs inthe Americas. by mortality stratum.................................................................................................. 13 Table 4.4 Range of prevalence (9%) of risk factors in 16 Brazilian cities. 2002/3*15 Table 4.5 Prevalence of risk factor by age group. Brazil 2003 ........................................... 15 Table 4.6 Percent of people reporting chronic health problems. Brazil. 2003 .................... 17 Table 4.7 Findings from the 2002/3 Household Survey on N C D Behavioral Risk Factors and Morbidity; Range o f reported prevalence (%) in 16 Brazilian cities* ..........18 Table 5.1 Differences in Age-standardized mortality rates per 10.000 and Relative Risk by socioeconomic zones in Siio Paulo. 1991....................................................... 19 Table 5.2 Prevalence in Obesity in Male and Female Adults by Region and Income Table 5.3 Prevalence of Daily Smoking. by Age. Gender. and Number o f Assets. 2003 ...21 Group. Brazil. 1974/75. 1989. and 1996/97 ........................................................ 22 Table 6.1 Recommendations o f Effectiveness for Interventions to Promote Physical Activity from the Task Force on Community Preventive Services (Centers for Disease Control and Prevention. 2001) ............................................................... 32 Table 6.2 Cost-effectiveness of selected NCD interventions. W H O analyses for 38 Table 6.3 Health Performance Indicators Relatedto NCDs for Goias................................ "Americas-B" region............................................................................................ Table 7.1 Intervention design. scope o f coverage increase. unit costs and lag time ...........43 46 Table 7.2 The estimated effectiveness of preventive interventions..................................... Table 7.3 Baseline scenario .burden o f disease. financial and economic costs..................47 Table 7.4 The costs of scaling up selected preventive interventions [US$. 20001..............48 48 Table 7.5 Scaling up selected preventive interventions reductions in (1) burden of disease; (2) financial costs; and (3) financial and economic costs ...................... 49 111 Table 7.6 Potential net cost savings of scaling up preventive interventions (costs of scaling up less than resulting savings of financial costs) ...................... 50 Table 7.7 Cost-effectiveness ratios [US$. 2000 per DALY] and benefit-cost ratios for selected preventive interventions .............................. 51 iv ACKNOWLEDGEMENTS This report was prepared by Isabella Dane1 (Team Leader), Christoph Kurowski, and Helen Saxenian. Background papers were prepared by Dariush Akhavan, Ken Hoffman, Suzanne Jackson, Roberto Nuiiez, Joaquim Pereira, Mike Pratt, and Marcia Westphal. Peer reviewers included Joy de Beyer, Jesus Maria Fernandez Diaz, and Anne-Marie Pierre-Louis. Additional inputs were provided by Cristian Baeza, Ariel Fiszbein, Gerard L a Forgia and Yasuhiko Matsuda. Lerick Kebeck and Cassia Miranda provided administrative, editing and graphics support. V Vi 1.EXECUTIVESUMMARY 1.1 Non-communicable diseases (NCDs) account for a large and growing share of Brazil's burden of disease. Currently, about 66% of the disease burden in Brazil i s due to non-communicable diseases, compared to 24% from communicable diseases and 10% from injuries. Brazil's shift towards non-communicable diseases i s a consequence of urbanization, improvements in health care, changing lifestyles, and globalization. Most of this disease burden i s not an inevitable result of a modern, aging society, but preventable-often at low cost. The purpose o f this report i s to provide an overview of the changing non-communicable disease burden in Brazil and its root causes, to examine costs and effectiveness o f alternative policy interventions to address this growing burden, and the costs disease and potential returns from expanding NCD prevention and control activities, and to consider policy implication of expanding activities to effectively address the shifting burden. 1.2 The report i s intended to inform decision makers at the federal, state, and municipal level in health, finance, and infrastructure, transport, city planning, and other areas. This i s because an effective non-communicable disease strategy must go well beyond the health sector. It will require broad educational and community interventions, as well as changes in economic policies, changes in the food supply, and changes in transportation policy and urbandesign. 1.3 As Brazil's population continues to age, non-communicable diseases, on current trends, could overwhelm the health system and cause considerable needless suffering and lost productivity. The cost of treating NCDs already accounts for nearly half the cost o f all hospital admissions. Demand will only increase as the health transition continues to evolve. Brazil has programs to prevent non-communicable diseases that are internationally-recognized success stories-such as tobacco control and Agita Sdo PauZo -and itcanbuildon these successes. Brazilcan also draw on evidence of successful interventions in other countries for its own strategy, andthere are many such examples. ChangesinAge Structure and DiseaseBurden inBrazil 1.4 Due to declining fertility and mortality rates, Brazil i s undergoing a very rapid demographic transition. This i s evidenced in the dramatic changes that have already occurred and are projected to occur in the age structure o f the population (see Figure 1). In2000, an estimated 15% of the population was aged 50 and over. This is projectedto increase to 29% by 2005 and 42% by 2050. The increase in absolute numbers is even more dramatic. In 2000, about 27 million Brazilians were 50 and over. By 2025, this i s projected to increase to 63 million, and to 96 million by 2050. The median age o f the Brazilian population was 26 in 2000. It i s projected to rise to 36 by 2025, and 44 b y 2050. 1.5 During the next decades, the aging of the population will also produce a new and probably one-time situation: an optimum demographic ratio between dependent population- children and the elderly - and working age population. This situation, called Figure 1. The rapid aging of Brazil's population: 2000 to 2050 Brazil: 2088 ALE FEMALE 10 8 6 4 2 0 0 2 4 6 8 Population (in millions) ource: U.S. Census Bureau, International Data Ease. Brazil; 2025 lALE FEMALE 1 Population (in millions) :ource: U S . Census Bureau, International Data Base. Brazil: 2050 I FEMALE 770-74 80+ 5-79 65-69 60-64 55-59 50-54 45-49 35-39 40-44 30-34 25-29 20-24 15-19 10-14 5-9 0-4 10 8 6 4 2 0 0 2 4 6 8 10 Population ( i n millions) hurce: U.S. Census Bureau, International Data Base. .. 11 the "demographic bonus",' should succeed in triggering regional development provided social and economic conditions can absorb the large amount of labor. The health status of this population will have a major impact on Brazil's economic and social development. 1.6 While there i s still an unfinished health agenda in Brazil with preventable disability-adjusted life years (DALYs) related to communicable, maternal and perinatal disease and malnutrition, the burden due to non-communicable diseases and injuries i s much greater, and will only increase, given the projected changes in the age structure o f the population. The burden of non-communicable diseases i s large for both years of life lost (YLL) and years lost to disability (YLD), as well as the combined measure, DALYs (Table 1). Cardiovascular disease i s far and away the most important cause of years o f life lost, with cancer another significant contributor. Neuro-psychiatric disorders predominate the years lost to disability with chronic respiratory disease and diabetes also important. * per 1000inhabitants **Includes benign neoplasias, endocrine, sensory, digestive, genito-urinary, skin, musculoskeletal, oral, and congenital disorders. Source: Brazil Burden of Disease Study. 1.7 Much of the N C D burden i s avoidable. The rate of DALYs due to NCDs i s actually higher in Brazil compared with the more developed countries in the Americas, the United States and Canada. This suggests that there are effective interventions ... 111 available that have lowered the burden in more developed countries. It is important to consider whether some of these interventions could be applied inBrazil. 1.8 For most of the major N C D diseases, including coronary heart disease, stroke, diabetes, and many cancers-the primary cause i s not underlying genetic factors, but environmental and behavioral risk factors. W H O supported a global study that estimated the proportion of deaths and DALYs that could be attributed to different risk factors by region. The study grouped the Americas into three country groupings with similar mortality profiles. For the grouping that includes Brazil, seven modifiable risk factors account for an estimated 53% of all deaths, and 30% of all DALYs. Of course the proportion would be much higher if it referred specifically to non-communicable disease deaths and DALYs. These seven risk factors include blood pressure, overweight, alcohol use, tobacco use, high cholesterol, low fruithegetable intake, and physical inactivity. The Americas country group with lowest mortality (made up of Canada, Cuba, and the United States) had an even higher share of deaths and DALYs attributed to these seven risk factors. It is useful to examine trends inthis lowest-mortality country grouping given that Brazil's diseaseburden i s increasingly similar to this group. 1.9 On current trends, the following can be predictedfor Brazil: increases in the prevalence of poor diet, overweight, and physical inactivity (in part because of the aging of the population, but also because of lifestyle and diet changes) will give rise to increases in high blood pressure, high cholesterol, and diabetes in Brazil, which will in turn cause more heart disease, strokes, kidney disease, cancer, and other NCDs. the current, relatively small share of mortality and burden of disease from smoking will increase as the lag time for its related illnesses catches up with the population (heart disease, strokes, chronic lungdisease, and cancer). 1.10 Although a comprehensive national study assessing the attributable burden of disease from risk factors has not been done for Brazil, there are two recent household surveys that provide important information on several key risk factors. They are the Risk Factor and NCD Survey (WS) carried out between 2002 and 2003 in 16 Brazilian capital cities,2 and the World Health Survey, Brazil, carried out in 2003.3 The RFS survey showed smoking rates among adults 25 or older in capital cities at about 20%, while for young adults, the prevalence was 15%. (The adolescent group i s a key target of anti- smoking interventions as it has been shown that those who do not start smoking in adolescence are much less likely to start smoking in adulthood.) The World Health Survey showed prevalence o f about 18%. Earlier surveys are not standardized, making comparison difficult, but it does appear that prevalence has fallen since the 1980s. Nevertheless, this will still cause considerable preventable harm. 1.11 The World Health Survey showed inactivity levels (defined as "some but insufficient" physical activity--less than 2.5 hours per week) to be around 24% with younger age groups being more active than those over 50 years old. The RFS found the range in inactivity levels between 28-54%, depending on the city. Physical inactivity tends to be a greater problem in large cities, where traffic, low air quality, few parks and iv recreation facilities, and crime make leisure physical activity hard to do. But even in rural areas, sedentary activities, such as television watching, are increasing. Physical inactivity doubles the risk of cardiovascular disease, type I1diabetes, and obesity. It increases the risks of high blood pressure, depression, anxiety, breast and colon cancer, lipiddisorders, and osteoporo~is.~ 1.12 The RFS survey also looked at overweight and obesity. It found that about 45 percent of men and 30 percent of women were overweight. And about 10 percent of adults were obese. The prevalence o f overweight increased with age. Overweight and obesity has increased over time. Monteiro and colleagues found that, for women, the percent obese increased from 6.5% to 12.4% from 1975/76 to 1996/97.5 In men, the numbers are lower but also increasing, from 2.1% to 6.4% over this same time period. The RFS survey also found that 35% of the population (range by city: 19-49%) were consuming low levels o f fruits and/or vegetables, less than 5 per week. 1.13 The poor in Brazil suffer a double burden. They are more affected by communicable diseases, as well as non-communicable diseases. This double burden helps propagate the cycle o f poverty. Non-communicable diseases are sometimes- mistakenly-called "diseases of affluence" because they are associated with urbanization and modernization. However, the association between NCDs and poverty is strong. In some cases, this i s because the poor have markedly higher risk factors, such as smoking rates. In other cases, risk factors may not be higher but N C D disease-related disability and mortality are higher because the poor are much less likely to have early detection and proper management. Disease progression can be much faster, and coping with the disease can pushthe household more deeply into poverty. The links between poverty and NCDs must be considered in policy design. For example, increasing tobacco taxes are effective in reducing tobacco consumption. But policy makers might consider making widely available smoking cessation programs in poorer neighborhoods, as the poor have higher smoking rates and less income to cope with increased taxes. What can be done? 1.14 The growing burden of non-communicable diseases i s increasingly stressing the Brazilian health system. Once they develop, NCDs are generally not curable and most become chronic diseases. The ongoing medical care necessary to treat NCDs has substantial costs for the health system as well as for the individual and their family. 1.15 Fortunately, randomized prevention trials, prospective epidemiological studies, and short-term studies have contributed richly to our understanding o f the dietary and lifestyle determinants o f major NCDs. This research shows that the large share of NCD determinants i s modifiable. Furthermore, low rates for many NCDs-coronary heart disease, stroke, diabetes, and many cancers, can be achieved without drugs or expensive medical interventions. Cynics may question the ability to change individual behaviors or the diet of a population-but other countries have shown that this i s possible. Take the example, often cited, of Finland. Finland had one o f the highest rates of coronary heart disease in the world in 1972. The government introduced a comprehensive program to educate the population about smoking, diet, and physical activity, first in North Karelia, V andthennationwide. It supported anti-smoking legislation, increased availability of low- fat diary products, and improved school meals. Age-standardized coronary heart disease mortality fell over the period 1969/71 to 1995 by a remarkable 73% in North Karelia, where the program was first introduced, and by 65% nationwide.6 1.16 Many of the most effective non-communicable disease control and prevention interventions do not require individual behavior changes because they take a population- based approach. An example would be to reduce-through legislation or voluntary industry action--the salt content in manufactured foods. This lowers blood pressure levels, which can produce significant and long run changes in disease incidence. While population-based approaches can be highly effective, some individual-based approaches are as well, especially if they target those at high absolute risk for a disease rather than those with a single risk factor level (such as hypertension or obesity). 1.17 Prevention interventions can be classified as primary or secondary. Primary prevention refers to actions that prevent the development of NCDs inthe first place. This i s done mainly by decreasing exposure to risk factors. Secondary prevention refers to early detection of a disease to minimize or interrupt its progression, such as blood pressure checks and cholesterol screening. The focus of the analysis in this volume is on primary prevention as the most cost-effective way to address the NCD epidemic. Actions can also be described as populatiodcommunity-based or individual (that i s to individuals through health care providers). Risk reduction tends to be more affordable and often more effective when oriented to the population rather than individuals. 1.18 There i s little information in Brazil about the financial and economic impact of its growing non-communicable disease burden and the costs and effectiveness of alternative policy approaches. To start to fill this gap, this report examined four examples o f primary prevention activities: a comprehensive community campaign to promote physical activity (Agita &io Puulo), the treatment of arterial hypertension with first-line drugs, a tax increase for tobacco resulting in a 10% price increase, and medical counseling of smokers. These activities target three risk factors: physical inactivity, arterial hypertension, and smoking, that are key to the incidence and prevalence of five NCDS, including ischaemic heart disease, cerebrovascular disease, diabetes mellitus, chronic obstructive pulmonary disease, and cancer o f the trachea, bronchi, and lungs. 1.19 The analysis estimated the costs of scaling up these preventive interventions, their impact on disease, and the financial and economic cost savings relative to a baseline scenario. The analysis found that providing anti-hypertensive treatment to 25% of those with hypertension, scaling up comprehensive community physical activity campaigns similar to Agita SZo Paulo, and a tax increase on cigarettes would, as a package, reduce Brazil's NCD burden by 845,000 DALYs, equivalent to a 5% decrease of the baseline DALY disease burden. It would generate about $1 billion in savings in treatment costs, equivalent to about 3% of the costs of treating NCDs. And it would save about $3.1 billion in treatment and productivity losses. The annual medical counseling o f smokers turned out to be a less desirable intervention in terms of costs and impact. For both comprehensive community physical activity campaigns and the cigarette tax increase, the savings in treatment costs well vi outweighed the intervention costs. The tobacco tax increase was extraordinarily cost-effective at about $1 per DALY gained. (The cost-effectiveness refers to the incremental intervention costs per DALY averted). Agitu Slio Paulo was also a remarkably cost-effective "buy" at only $246 per DALY averted. Anti-hypertensive treatment had lower cost-effectiveness, at $1,498 per DALY, but still very worthwhile. The medical counseling of smokers was less cost-effective compared to the others, at $9,360 per DALY averted. (Interventions with costs per DALY of less than twice a country's per capita GDP are, as a rule-of-thumb, considered highly cost-effective). 1.20 Overall, the analysis showed that scaling up community campaigns to promote physical activity, the treatment of uncomplicated hypertension with first-line drugs, and tobacco tax increases should be financially and economically attractive, and will help lessen Brazil's burden of non-communicable disease in the future. However, much more could be done beyond scaling up the three interventions used as examples. There are many other effective interventions that have been applied in developed countries. A good place to start would be considering which of these interventions mightbe effective inBrazil as well. 1.21 Table 1.2 presents recommendations for the short to mediumterm for reshaping Brazil's health system more fundamentally towards NCD prevention and control. The first broad recommendation i s to further develop and implement health strategies to prevent NCDs by focusing on risk factors. This includes measures to strengthen capacity in N C D policy formulation, regulation, monitoring, implementation, and research. Specific measures to be considered for support as soon as possible include the preparation of a national plan for N C D and N C D risk factor prevention, supported by a multisectoral national advocacy body, and clear focal points in the MOH. Regulatory andlegislative action will be needed on tobacco tax levels, food content and labeling, and smoke-free environments. Effective communication of health messages i s important in N C D risk factor prevention but this requires improved capacity in health communication and social marketing. As shown by the economic analysis, population-based interventions are cost-saving in the long term but they require up front investment. These interventions could, in part, be self-financing if increases in taxes related to risk factor control such as for tobacco and alcohol are implemented andearmarked for health. 1.22 Much more could be done to reduce exposure to risk factors over the coming years. The report has focused on diet, physical activity, and smoking as these risk factors contribute so much to Brazil's non-communicable disease burden. Brazil has implemented food labeling to help raise consumer awareness, but much more could be done at a population-based level to improve diet. For example, legislation to reduce the use of partially hydrogenated fat and the salt content of food could be considered. 1.23 Increases in cigarette taxes could be helpful in reducing smoking prevalence even further. This could be complemented b y measures to reduce smuggling and sales of tobacco to minors, and more support for smoking cessation programs, particularly among the poor and less educated. vii 1.24 In terms of physical activity, scaling up comprehensive community physical activity campaigns similar to Agitu Sa"o Puulo throughout Brazil could have a huge impact as the analysis in this report has shown. Among other measures in Table 1.2, better city planning and infrastructure design could profoundly change patterns of daily living and physical activity levels (better public transportation, car-free zones in cities, separate and safe bicycle paths, well-lit sidewalks, and other measures). 1.25 While population-based interventions tend to be more cost-effective, primary care providers can also deliver NCD and risk factor prevention / health promotion activities to individuals under their care. Health promotion i s one of eight priorities for the FHP, but this area has not been adequately developed. An assessment of which interventions should be financed by the FHP is needed. They should then be included in an FHP health promotion strategy and activities monitored using appropriate performance indicators. 1.26 Better information i s required to address the NCD challenge in Brazil. The second broad recommendation i s to improve surveillance and monitoring of NCD and risk factor prevention strategies. Regular risk factor surveillance needs to be supported in order to monitor trends, provide input for policy and program design, and monitor impact. Appropriate NCD indicators need to be devised for inclusion in the bipartite and tripartite "puctos" that monitor performance at the state and municipal levels. 1.27 The third broad recommendation i s to improve secondary prevention of NCDs through better health care and screening. The report only briefly touches on some key issues. More work i s needed to assess the strengths, weaknesses, needs and gaps of the health system in this area. Table 1.2 recommends measures to scale up screening and increase demand for treatment of hypertension and diabetes. The FHP teams need improved capacity to treat and control NCDs. Issues that should be addressed include long-term adherence to treatment, continuity of care, integrated care, and the ability of patients to self-manage their own disease. 1.28 The final recommendation i s to strengthen research on effective NCD / risk factor prevention in Brazil, particularly among the poor. Many o f the interventions require changes in lifestyle and behaviors or are applied in settings like schools, communities and workplaces where the cultural context i s important. There i s very little information specific to Brazil on the effectiveness of prevention interventions, and it i s needed if scarce resources are to be used efficiently. Research i s particularly urgently needed on the effectiveness o f prevention interventions among the poor, who carry a higher burden of NCDs, since many interventions - such as those using mass media, written information, screening, and taxes - often either do not reach the poor or have a deleterious impact on the poor. 1.29 On current trends, the burden of NCDs in Brazil will generate enormous costs in terms of disability and suffering, premature mortality, lost productivity, and health care expenditures. Fortunately, a large share of the determinants of NCDs i s modifiable- through appropriate public policies, programs, and partnerships. Brazil has already had some remarkable successes, for example, in reducing smoking prevalence and in v111 ... increasing physical activity in S2o Paulo. But much more needs to be done. The more quickly Brazil can reshape and strengthen its health system and develop multisectoral policies for more effective N C D prevention and control, the more quickly Brazil's NCD epidemic will take a lower trajectory. i x a, 0 C Q.' L d ii > b L 0 U m 5i 5 C 3 i) - t E cd C x -.- a0j C cd a, 2 2b Y 20 % .-L a, L>E c .3 x 2 L. .3 x .-c M C E LA e e s2 m a, c m a, .e * U 0 2. SCOPE OF THIS REPORT 2.1 Non-communicable diseases account for a large and growing share of Brazil's burden of disease. Non-communicable diseases account for a large and growing share of Brazil's burden of disease. Currently, about 66% of the disease burdenin Brazil i s due to non-communicable diseases, compared to 24% from communicable diseases and 10% from injuries. Brazil's shift towards non-communicable diseases i s a consequence of urbanization, improvements in health care, changing lifestyles, and globalization. Most of this disease burden i s not an inevitable result of a modem, aging society, but preventable-very often at low cost. The purpose of this report i s to provide an overview of the changing non-communicable disease burden in Brazil and its root causes, to examine the economic costs of the disease burden, and costs and effectiveness of alternative policy interventions to address this growing burden. Chapter 3 reviews the latest demographic, epidemiological, and burden o f disease trends in Brazil and places these trends in the broader context of Latin America. Chapter 4 examines how much of the disease burden i s avoidable, and examines the evolution of the major risk factors, such as smoking, obesity, and diet, in Brazil. Chapter 5 assesses how the poor are faring in the epidemiological transition, given that they are hit by a double burden of both excess communicable and non-communicable diseases. Chapter 6 reviews the interventions that Brazil might consider to address non-communicable diseases, based on international and Brazilian experience. Chapter 7 examines four interventions that address physical activity, tobacco use, and hypertension in more detail, developing cost- effectiveness estimates for Brazil of a scaled up response. Chapter 8 i s a summary of the conclusions and policy recommendations of the report. 2.2 Non-communicable diseases i s a huge topic. It includes a multitude of chronic diseases that affect every organ in the body. And there are many different ways to address these diseases. NCDs are receiving more attention because they are a growing global health threat. There are several recent reports that examine the many interventions available to address N C D prevention and control. This report draws on these studies' findings. 2.3 There are two major challenges in the field o f NCDs: preventing them from occurring in the first place (primary prevention) and assuring that people who have NCDs receive adequate treatment thus preventing complications and disability (secondary prevention). This report reviews the situation inBrazil, and summarizes interventions that have great potential to address the problems in Brazil. The report emphasizes primary prevention because of the greater social and economic benefits that are often associated with prevention. However, given the magnitude of the existing burden already, the report also briefly discusses some of the more important issues and interventions related to the treatment and control of NCDs. Finally this report considers some of the implications o f the NCD burden for the Brazilian health system. 2.4 The report pays special attention to two diseases out of the large domain of NCDs: hypertension and diabetes. These two diseases have been identified by the MOH as top priorities. Two areas that are important for Brazil but not examined in detail in this 1 report simply due to time constraints are mental healthand alcohol as a risk factor. These are important topics and need to be addressed insubsequent studies. 2.5 What about the unfinished agenda of MCH and communicable diseases? While there i s still an unfinished MCHand communicable disease agenda inBrazil, most o f these problems are being successfully managed. The remaining issues in these areas are related to inequities and gaps within and between states rather than knowledge / program content at the national level. 2.6 The World Bank has included in its Country Assistance Strategy for Brazil achievement of a longer, healthier life for Brazilian people. The findings and recommendations from this report can contribute this goal, andinform activities inWorld Bank health projects including the VIGISUS, Family Health and state health projects. Adolescence i s an important time for forging lifelong habits related to healthy lifestyles, andthis report dovetails some of the recommendations made by the Report on Youth at Risk inBrazil. The work on the quality of health care (QUALISUS) will expand on some of the findings about the need for a shift in the way health care services are provided to people with chronic diseases. 2.7 The Brazilian M O H has taken steps to address the challenge of NCDs. It has included hypertension and diabetes as priorities in its national health plan. It has also included the screening and treatment of these two diseases as priorities within the Family Health Program, Brazil's principle vehicle for health care delivery. In the area of primary prevention, Brazil has begun to assess the prevalence o f risk factors, and smoking prevention and tobacco control activities have been underway for some time. This report aims to provide evidence, economic and epidemiological, to support the further expansion of these activities, particularly those related to primary prevention, and to summarize some of the key issues to be tackled for this effort to have the most impact. 2 3. THE HEALTHTRANSITIONAND BURDENOF DISEASEINBRAZIL THEDEMOGRAPHIC TRANSITION INBRAZIL 3.1 Brazil i s undergoing a rapid demographic transition due to falling birth and mortality rates, and increasing life expectancy. This transition differs from the classic model based on European experience in that it i s occurring over a very compressed time frame. A transition that took approximately two hundred years in Europe is taking place over a few decades in Brazil. The rate of progress of the demographic transition influences the epidemiological transition and burden of disease and will therefore be briefly reviewed here. In 1950, Figure 3.1 Total Fertility Rate, Brazil and LAC, 1950-2000 the total fertility rate (TFR) in Brazil was about 6.2 children per woman, compared with about 5.9 for LAC (Figure 3.1).73s Starting about 1960, the fertility transition has been very rapid. The current TFR 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 for Brazil i s estimated at 2.1 children, compared to 2.5 children for LAC as a whole. I 3.2 Mortality rates Figure 3.2 Life expectancy at birth, Brazil and LAC, 1950-2000 bo- decreased and life expectancy ~~~-=-11 steadily improved between - - - - - ' 1950-2000in Brazil. Brazil's life expectancy was 51 in 1950 and rose to 68 in 2000-slightly below the 0 LAC average (Figure 3.2).7 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 This trend should continue +Brazil --.-LAC with further improvements duringthe next 50 years. Figure 3.3 Infant Mortality Rate, Brazil and LAC, 1950-2000 3.3 The main explanation 160 for the rapid rise in LEB during 140 the past fifty years i s the 120 100 dramatic reduction in infant and 80 tBrazil child (under 5) mortality rates. 60 40 Figure 3.3 shows the evolution 20 of the Infant Mortality Rate 0 (IMR) which decreased more 1950195519601965 1970197519801985 I99019952000 than 75 percent since 1950, 3 from 140 per 1000 live births (LB), to 30 in 2000.' The IMR has been one of the main beneficiaries of improved healthconditions to date. As the roomfor improvement in IMR declines, gains in LEB will have to come from decreasing mortality in older age groups. 3.4 Figure 3.4 depicts this transition in LAC overall by showing the share of deaths for each age group with respect to total deaths.' This indicator i s based on two factors: the evolution in mortality and changes in the populations' age structure. In 1950, children under 15 accounted for more than half o f all deaths (54 )ercent), and only 17 percent Figure 3.4. Proportion of deaths byage group. Latin America. were 65 years or older (65+). Currently (2000-05), W r +0-14 these shares are reversed: 17 percent o f deaths for +15-6 children and 46 percent for the elderly. The share o f child deaths will further decrease in the future (5 percent by 2050), and three- 1 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 fourths will be for people 65 I Year and older. 3.5 Figure 3.5 depicts the dramatic changes that have already occurred and are projected to occur in the age structure of the population. In 2000, an estimated 15% o f the population was aged 50 and over. This is projected to increase to 29% by 2025 and 42% by 2050. The increase in absolute numbers i s even more dramatic. In 2000, about 27 million Brazilians were 50 or over. B y 2025, this i s projected to increase to 63 million, and to 96 million by 2050. The median age of the Brazilian population was 26 in 2000 and is projectedto rise to 36 by 2025, and44 by 2050. 'Datafrom the BrazilianInstituteof GeographyandStatistics(IBGE) 4 Figure 3.5 The rapid aging of Brazil's population: 2000 to 2050 Brazil: 2888 RLE FEMALE 80+ 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 220-24 5-29 15-19 10-14 5-9 0-4 10 8 6 4 2 0 0 2 4 6 8 10 Population 5 kg) in adult life. Voluntary weight reduction in overweight or obese individuals with impaired glucose tolerance (although screening for such individuals may not be cost- effective in many countries). Practicing an endurance activity at moderate or greater level of intensity (e.g. brisk walking) for one hour or more per day on most days per week. Ensuring that saturated fat intake does not exceed 10% of total energy and for high-risk groups, fat intake should be ~ 7 o%f total energy. Achieving adequate intakes of NSP through regular consumption of wholegrain cereals, legumes, fruits and vegetables. A minimum daily intake of 20 g is recommended. Three studies in diabetes prevention demonstrate the effectiveness of lifestyle changes and community interventions in the prevention o f Type 2 diabetes.47,56,57,58 me studies used combinations o f dietary treatment and/or an increase in physical activity for individuals at high risk of diabetes. Each of the three studies was able to demonstrate significant decreases in T y e 2 diabetes among high risk subjects (Le. those with impaired glucose tolerance). 8 34 CARMENStrategy 6.31 Many of these strategies are being supported through CARMEN (Conjunto de Accidnes para la Reduccidn Multifactorial de las Enfermedades No transmisibles), an international network to support the prevention of NCDs throughout LAC, organized by WHOPAHO. CARMEN focuses on supporting municipalities within its member states in three main strategies: integrated prevention, promotion of health equity, and demonstrative effect (the testing of interventions in a demonstrated area to determine effectiveness and acceptability, before expanding the intervention). Several Brazilian municipalities are participating in the CARMENproject. An evaluation of their successes and failures, and recommendations for their expansion to other Brazilian municipalities would be useful. SECONDARYPREVENTIONOF MAJOR NCD 6.32 There i s a large and growing body of literature on the effective treatment of chronic diseases. Here we will briefly review some of the major findings with application to Brazil for hypertension and diabetes only. These two diseases have been prioritized under the National Health Plan. They are very important precursors of cardiovascular disease and their control will contribute significantly to reductions in CVD mortality. 6.33 Many, but not all, cases of hypertension and diabetes can be prevented. Once these diseases occur it i s vital that they be controlled through a combination of low-cost medications and lifestyle changes that reduce exposure to risk factors. This review of secondary prevention of NCDs will focus on diabetes and hypertension for several reasons. These two diseases are precursors for more serious problems. Adequate blood pressure and blood sugar control leads to a longer and better quality life, and prevents many serious sequelae such as heart attacks, heart failure, strokes, and kidney failure, and in the case of diabetes good control prevents amputations and blindness. When these complications occur, they usually leave a person significantly disabled, requiring more costly medications and treatments, and more frequent hospitalizations placing a burden on the family and on the health care system. The Brazilian MOH has highlighted the importance of the management and control of these two diseases as one of its six major national health goals. In line with these goals, the Family Health Program, the principal vehicle for primary health care delivery provided b y SUS includes as one of eight priorities the control o f diabetes and hypertension among people under its care. The interventions discussed here will be primarily individual interventions. Nevertheless, it i s important to keep in mind that population-based interventions to decrease exposure to risk factors discussed in the section on primary prevention will also have a positive impact on the control o f diabetes and hypertension. Hypertension 6.34 As noted previously, hypertension is an important risk factor for cardiovascular disease and its treatment and control can prevent the occurrence of heart attacks, heart failure, coronary artery disease, and strokes. The treatment of hypertension i s one of the 35 most cost-effective interventions available.44 Effective interventions to lower BP include: BPscreening for adults 18 years and older D A S Hdiet -rich in fruits and vegetables and low infat Weight loss Physical activity Reduced salt intake Stress management Pharmaceutical treatment including diuretics, beta blockers, calcium antagonists, alpha blockers, and angiotensin lIantagonists. A variety of multifactorial interventions involving biofeedback, lifestyle changes and psychosocial interventions 6.35 The effectiveness of hypertension treatment to reduce heart attacks and strokes i s related to the individual's underlying risk o f having coronary artery disease or a stroke. Not all individuals with a given elevated blood pressure have the same risk. Identifying those with higher risks substantially increases the cost-effectiveness o f treatment. Such strategies have been explored in developed countries and the factors included in a `high risk' CVD profile are elevated blood sugar, waist circumference, and high cholesterol, in addition to hypertension. Patients with all four risk factors are at particularly high risk for C V D and it i s generally considered particularly cost-effective to control their blood pressures. It i s not clear how applicable these findings are to the Brazilian context. The utilization of this type o f risk score has not been studied in Brazil. Hypertensioncontrol in Brazil 6.36 There are multiple Brazilian studies assessing various aspects of the care o f people with hypertension. They tend to corroborate findings in developed countries. Some of the principal findings from this review include: The importance o f continuity o f care inblood pressure control59 Lack of adherence to treatment was noted as a major obstacle to blood pressure control6' Variable effects of education programs with some demonstrating an impact on blood pressure,61,6263 others on knowledge,64 and others showing little effect.65,66,67 Knowledge of risk factors and availability o f health services (among workers at the Bank of Brazil) did not guarantee treatment68 A study of "Centros de Saitde Escola" (Health Centers with Schools/Education) carried out over a period o f one year in 90/91 among 3793 adult users found that (i)there was muchlower thanexpected coverage ofhypertensionanddiabetes and (ii) patients seen in rapid care settings and found to have hy ertension did not return for follow-up care and often did not receive appointmentsr9 One cost-effectiveness study was identified related to the treatment of hypertension in Pelotas, Brazil (Costa et.al. 2002).70It found that treatment with 36 diuretics and beta-blockers was more cost-effective than treatment with ACE inhibitors and calcium channel blockers. Diabetes 6.37 Several trials have shown that lifestyle interventions can delay or prevent diabetes by 40-60% among people at risk.71In the U.S., a large randomized, controlled trial showed that the lifestyle intervention group (weight loss and moderate physical activity) had a significantly lower incidence of diabetes than a treatment and placebo group. 6.38 In a review of the cost-effectiveness of interventions for preventing and treating diabetes and its complications in developing countries, the following interventions were found to be cost-saving (cost/QALY): glycemic control in persons with H b A l c > 9%; blood pressure control in person with >95/160 mm Hg; foot care in person with high risk of ulcer; and preconception care for preventing birthdefects. Diabetes in Brazil 6.39 Education was noted to be an important factor in improving control of diabetes. Inone study planning and development of interventionsrelated to nutrition education was based on a biopsychosocial evaluation of women with type 2 diabetes that showed an impact on weight reduction, and glucose and cholesterol control. 72 Inanother study education provided by a multidisciplinary team was shown to be protective in preventing people with diabetes from developing complications requiring amputation (Gamba, 2001).70While these studies are limited they corroborate the importance noted in developed countries of assuring knowledge required in self-care, as well as adherence to treatment that i s necessary for improved outcomes in chronic diseases. Cardiovascular Disease (CVD) and Strokes 6.40 As the main risk factors for CVD and strokes are the same (Le. hypertension, smoking, overweight and hypercholesterolemia), prevention initiatives will be considered together. Reduction of exposure to risk factors such as smoking and overweight promote both primary prevention o f CVDs and also secondary prevention by reducing the likelihood of complications. 6.41 A review of effectives strategies for CVD in the United States73found that: A key strategy for reducing risk factors is to educate the public and health care practitioners about the importance of prevention. People can reduce their risk for C V D by controlling high blood pressure and highblood cholesterol levels. A class of drugs called statins can reduce deaths from heart disease by reducing cholesterol levels, and medications that reduce blood pressure levels can reduce the risk for heart disease, strokes, and other coronary events. Patients who take beta blockers within days or weeks of a heart attack have a greater chance of surviving the heart attack. 37 0 People should be educated about the signs and symptoms of heart attacks and stroke and the importance of calling for help ("91 1") quickly. Research indicates that nearly 70% of deaths from heart disease occur before a person can be admitted to a hospital, and about 48% of stroke victims die before emergency medical personnel arrive. Other important ways that people can reduce their risk for heart disease and stroke are to avoid smoking, adopt healthier diets, and increase their physical activity. Cost-effectiveness of Interventions Average Cost intervention Effectiveness (international$) Increasingtobacco taxes to the highest regionaltax rate (75%) $19 Legislationto decrease salt content of processedfoods, plus appropriate labeling and enforcement $127 Legislationand health education to reducecholesterol $135 Healtheducation through mass mediato reducecholesterol $136 Voluntarycooperation for food manufacturerswith government to decrease salt in processedfoods, plus appropriate labeling $244 Informationdissemination $436 Hypertension-loweringdrug treatment and education on lifestyle modificationincluding dietary advice, delivered by physicians to $811 individuals with systolic blood pressure > 140 mmHg Clean indoor air law enforcement $972 Cholesterol-loweringdrug treatment (statins)and education on lifestylemodification including dietary advice, delivery by physiciansto individuals whose serum cholesterol concentration $1.326 exceeds 220 mg/dl Nicotinereplacementtherapy $3.083 * For more information, see the World Health Report2002 and the WHO website on cost-effectiveness analysis, or www.who.int/whosis/cea 38 Issues Related to Chronic Care 6.42 While evidence on the effectiveness of interventions for the control of specific NCDs i s important, it is becoming increasingly clear that health systems will need to shift the way they operate if they are to address the needs of chronic patients. Most health systems are organized to provide care for acute illnesses that resolve quickly. Several publications including a recent WHO re ort summarize some of the steps necessary for the shift to better chronic care toO C C U ~ . ~ ' ~ ~ A major issue is adherence to treatment. The effectiveness of any treatment i s greatly diminished by low adherence rates. It is estimated that overall adherence to long-term therapies for chronic illnesses i s approximately 50% in developed countries, and it tends to be much lower in developing countries.76 Adherence is not just a matter of individual compliance. Interventions that are poorly designed (Le. those that do not take into account the social, cultural, and economic realities o f the patient, the capacity o f the health care system to support the intervention, the characteristics of the disease) have little chance of succeeding. The support of family members and community organizations also plays a key role in maintaining adherence. (WHO, 2003b). 6.43 Other steps necessary to improve care for chronic conditions include integrated health services (integrated across primary and specialty as well as hospital care), adequate follow-up and continuity of care, emphasis on enabling self-management, utilization of ancillary personnel to educate and support people with chronic conditions, community support services, standardized treatment protocols, and consistent financing. Brazil's Arterial Hypertension and DiabetesMellitus Care Reform Plan 6.44 The chronic nature o f these conditions and their enormous impact on the morbidity and mortality of the Brazilian population poses a challenge for the Unified Health System (SUS). The SUS must create an environment that protects the population from risk factors for those diseases, offer early detection, and ensure systematic treatment and monitoring of individuals identified as suffering from these diseases. In response, the federal government coordinated the elaboration of an Arterial Hypertension and Diabetes Mellitus Care Reform Plan that was approved in 2001. Its major objective was to organize and expand high-quality, solution-oriented care for people suffering from arterial hypertension (AH) and diabetes mellitus (DM) within the public health services network. The plan also includes objectives related to primary prevention through risk factor reduction. To reinforce the plan and ensure its implementation Congress introduced and passed various bills including one that sets forth guidelines for the prevention and comprehensive health care of people with diabetes within SUS and ensures protection o f the rights of those who suffer from this disease. In 2004 a Working Group representing experts from the various levels of SUS, academia, patient associations and NGOs was established to update recommendations for DM and AH and comprehensive care. 6.45 The Care Reform Plan consists of four implementation states including i ) multiplier training o f professionals in the primary care network, ii)media campaign and identification of suspected cases o f AH and DM, iii)confirmation of diagnosis, and iv) 39 registration and referral of confirmed cases to primary care facilities. Duringthe National Campaign for DM detection in 2001 blood glucose tests were conducted on nearly 21 million individuals. A campaign in 2002 screened 12 million people for AH and 36% of these were elevated and referred to health care facilities. The essential drug list contains medications recommended in a standard of care manual for AH and DM. 6.46 Under the current administration the Care Reform Plan was updated and includes: i)ongoing on-site education in primary care techniques; ii)incentives for expanding the information system registry used to refer people to PHCs; iii)monitoring and assessment indicators to improve oversight of AH and DM-related treatment activities; iv) clinical protocols and training; v) policies that promote healthier lifestyles, well-being, and prevent the risks associated with illness and the related complications; vi) combining primary care approaches with moderately and highly complex procedures in order to reduce the risks of complications. However, it i s not clear how fully the plan has been implemented, nor whether the quality is adequate. Campaigns to screen and detect diabetes and hypertension continue to be a major NCD activity. Limitations 6.47 The vast majority of studies reported in the literature test efficacy rather than effectiveness. The above review tries, wherever possible, to locate more comprehensive reviews in order to formulate conclusions on the types of interventions that are likely to be most effective. 6.48 A review of the literature from L A C of health promotion identified 131 studies published in the international and Latin American literature. However, in two-thirds the type o f evaluation carried out was not described.77 And most o f the papers with well- carried out methodology were related to maternal and child or reproductive health. The relative lack of published conducted in L A C in the field of NCDs and risk factors means that the review relies on studies from developed countries for evidence o f interventions potentially effective in LAC. However, many factors should be considered before attempting to generalize interventions designed inNorth America and Europe to the LAC context. They include socio-cultural factors that are relevant in deciding how best to enable and promote behavior change, socioeconomic factors, and other economic and infrastructure factors (e.g. public healthlprimary care/educational/municipal infrastructure). Studies related to the treatment of NCDs among thepoor 6.49 N o studies were identified that specifically assessed the impact o f interventions to reduce exposure to risk factors among the poor, nor were studies identified that assessedthe impact of treatment modalities, associated support services, or the impact of education or communication interventions. 40 NCD/ RISKFACTORPREVENTIONPOLICIESANDACTIVITIES INBRAZIL 6.50 Prevention of NCDs through interventions that decrease exposure to risk factors i s carried out in the context of disease prevention or more broadly in the context of health promotion.'" Health promotion goes beyond the prevention of NCDs and other illnesses and has "a state of complete physical, mental and social well-being" as a goal. In this section we will review policies, activities and responsibilities of different departments within the M O H related NCD prevention. National level 6.51 Health promotion activities have been carried out by the Ministry o f Health since the late 1980s. A unit specifically dedicated to Health Promotion was first established in 1998, under the Health Policy Secretariat. After the elections of 2002, the M O H was restructured and the technical areas related to health promotion were first moved to the Executive Secretariat, and in 2005 to the Health Surveillance Secretariat. 6.52 The Health Surveillance Secretariat (SVS) was created in June 2003 and absorbed what was previously the Epidemiology Center (CENEPI). It has a Department of Situation Analysis (DHSA) that i s responsible for collecting and analyzing information about N C D mortality, morbidity and risk factors. The DHSA provides epidemiological information related to NCDs and risk factors. It also monitors information related to service delivery including preventive services. This area i s just beginning to be developed in Brazil. CENEPI carried out the first Brazilian Risk Factor and NCD Survey in 16 capital cities and the results were recently published (2004).* The challenge now i s for the M O H to develop a coherent plan to address its findings. 6.53 The DHSA includes a Coordinating Unit for NCDs and Injuries, and a Coordinating Unit for Information and Epidemiological Analysis. The CU for NCDs and Injuries will carry out surveillance related to these health outcomes. Health promotion activities in the M O H will be coordinated out of this CU. One of its first actions was to finalize a National Health Promotion Policy (NHPP) which emphasizes N C D and injury prevention (and includes some communicable disease health promotion activities e.g. dengue, malaria). The policy will be launched in late 2005 and includes objectives, actions and specific activities, many related to the primary prevention o f NCDs. The new health promotion section in the NCD CU should play a key role in implementing, monitoring and evaluating the interventions outlined in the NPHP. 6.54 The Health Care Secretariat (SAS) has activities related to health promotion and NCD prevention in primary care settings. The Department o f Primary Health Care (DPHC) has included health promotion has one of the 8 major priorities of the Family i vHealth promotion was defined at the first International Conference on Health Promotion meeting in Ottawa in 1986 as "the process of enabling people to increase control over, and to improve, their health". Health promotion actions are multi-sectoral and include i)building healthy public policy, ii)creating supportive environments, iii)strengthening community actions, iv) developing personal skills, and v) reorienting health services. (Ottawa Charter for Health Promotion) 41 Health Program. As noted earlier, while not generally as effective as population-based health promotion activities, there are effective interventions that are recommended at the individual/clinical level. However, the FHP has focused more on screening and identifying people with NCDs, particularly diabetes and hypertension. Health promotion activities within the FHP are not well-developed. 6.55 The National Cancer Institute (INCA) i s also part of the SAS and i s responsible for cancer prevention and detection activities. The National Tobacco Control Program established in 1987 was developed under its leadership. This program i s considered one of the more successful national tobacco control programs.78 It i s responsible for strong anti-tobacco legislation and a nationwide, decentralized program. While a rigorous evaluation o f the impact on smoking prevalence o f this program i s not available, the recent RFS suggests that prevalence is decreasing. 6.56 Other Secretariats with the MOH with some responsibility for health promotion include the Secretariat for the Management o f Work and Education in Health which trains health workers in health promotion and NCD prevention and treatment. It is also responsible for popular education in health an important area for health promotion. The Secretariat for Science and Technology i s responsible for evaluating the implementation and impact o f health policies including health promotion through research and investigations. 6.57 In response to the health transition, the MOH has developed a variety of intersectoral health promotion policies. These include: 0 National tobacco and anti-drug policy National and state policies on cancer control 0 National policy for the reduction of morbi-mortality from accidents and violence 0 National and state policies related to occupational health National food and nutrition policy 0 Food security policy However, most of these policies have not yet been operationalized in concrete national, state or municipal health plans. NCDand Risk Factor Surveillance 6.58 NCD surveillance in Brazil includes mortality and hospitalization rates for specific diseases. It also includes a Cancer registry in 19 cities. However, risk factor surveillance has only recently begun in Brazil. The first step was publication in 2004 o f results from the NCD Risk Factor Survey - carried out in 15 capital cities and the DF. This survey is expected to be repeated approximately every five years. In addition, plans are currently in the works for a school-based adolescent risk factor survey (also in capital cities) to monitor the exposure to and uptake o f risk factors among youth. These surveys will provide a base for N C D risk factor surveillance. However, they are limited because they do not represent the entire population, particularly people living in rural areas. 42 Extending surveillance to the entire population i s an important challenge for Brazil, given its geographic size, many demands, and limited resources. Likewise, extending the surveillance system in time to one that provides continuous (instead of periodic) information will also be a challenge. Such information i s invaluable for monitoring the effectiveness of interventions. It i s also important as a source of information on which to base timely public health decisions. Waiting five years to find out whether interventions are having the desired impact does not permit efficient or effective allocation of resources to address these public health problems. MonitoringState and Municipal Performance on NCDPrevention and Control 6.59 As part of the decentralization process the MOH has developed a system in which bipartite (nationalhtate) and tripartite (national/state/municipal) `pactos' or agreements are reached on a set of indicators for monitoring performance in improving health outcomes. Health budget disbursements are tied to achievement of these indictors which are generally agreed to Table 6.3 Health PerformanceIndicators Relatedto NCDsfor State of Goib for areas considered to be priorities (e.g. infant mortality). 1. Average annual number of doctor visits per inhabitant Among these are indicators (basic specialties) related to NCDs. The NCD 2. Average monthly number of home visits per family indicators for the State of 3. Percent of the populationcovered by the PSF GoiAs (Table 6.3) are typical o f 4. Proportion of hospitalizations for ketoacidosis and other states. While there are diabetic coma multiple indicators related to 5. Proportion of hospitalizations for diabetes NCDs, there are no indicators related to decreasing exposure 6. Ratio of number of PAP smears in women 25-59 / total to risk factors, primary number women 25-59 prevention or health promotion. 7. Hospitalization rate for strokes in people aged 30-59 years All of the indicators are related to individual medical care, 8. Hospitalization rate for ketoacidosis and diabetic coma hospitalization and disease 9. Hospitalization rate for strokes rates. The emphasis i s on the 10. Hospitalization rate for congestive heart failure management of diseases and 11.Female mortality rate for cervical cancer decreasing morbidity / mortality, and there i s no 12. Female mortality rate for breast cancer monitoring o f the factors 13. Mortality rate for strokes related to the primary prevention of these diseases. 6.60 The inclusion of indicators related to health promotion and prevention o f risk factors i s important to assure that states are performing well not only in the delivery o f individual health services, but also in the delivery of population-and community-based programs that as already noted are generally more effective and less costly in reducing the burden of NCDs. 43 7. FINANCIALAND ECONOMICIMPACT OF EXPANDINGKEYNCDPREVENTIONACTIVITIES - FOUREXAMPLES Introduction 7.1 There i s little information in Brazil about the financial and economic impact of its growing non-communicable disease burden and the costs and effectiveness of alternative courses of policy action. This chapter presents analysis to start to fill this critical gap. It first projects, given current levels of prevention, care, and treatment, the consequences in terms of future burden of disease, financial and economic costs (the baseline). It then estimates the incremental costs and impact of an expanded non- communicable disease prevention program using four interventions as examples. The costs of the expanded response are then compared with the baseline in order to identify net cost savings. Then cost-effectiveness and benefit-cost ratios are calculated for the expanded response. Interventions Included for the Expanded NCDPrevention Program 7.2 Although there are a large number of non-communicable diseases, risk factors and possible prevention activities, for reasons of practicality and data availability, the analysis examines only three risk factors: physical inactivity, arterial hypertension, and smoking (Figure 4.1). These three are key risk factors that determine, to a large extent, the incidence and prevalence o f five non-communicable diseases: ischemic heart disease, cerebrovascular disease, diabetes mellitus, chronic obstructive pulmonary disease, and cancer of the trachea, bronchi and lungs. The expanded response i s made up of four primary prevention activities of proven effectiveness that address these three risk factors. These are: a comprehensive community campaign to promote physical activity (using Agitu S6o Puulo as an example), the treatment of arterial hypertension with first-line drugs, a tax increase for tobacco prompting a 10% increase in cigarette prices, and medical counseling of smokers. These interventions are described in Table 7.1. For this analysis, all four interventions are presented individually, and then three of them- comprehensive community campaign, treatment of arterial hypertension, and a tobacco tax increase--are presented as a "package". As these three interventions address different risk factors, their costs, and impact can simply be added without the risk of double counting. InTables 7.4 through 7.7, this "package" i s shown in addition to the individual interventions. The tobacco tax increase was selected for the "package" over medical counseling of smokers because, as will become clear in this analysis, it is much more cost-effective than medical counseling of smokers. Methodology This section summarizes key methodological issues. Annex 2 present more detailed information on the methods and data used in this economic analysis. 44 Time horizon and measurement of impact 7.3 The analysis estimates the costs of scaling up preventive interventions and the changes in noncommunicable disease incidence over the period 2005 to 2009. However, non-communicable diseases, once they are manifest, cause years of lost life or disability over a time horizon of many years. Therefore, the analysis estimates the long term benefits (DALYs averted, savings in treatment costs, and productivity losses averted) that stemfrom the changes indisease incidence from 2005 to 2009. 7.4 Benefits of an expanded response were expressed in DALYs avoided, financial, and economic cost savings relative to the baseline scenario. For this analysis, financial costs are the cost of treating patients with non-communicable diseases. Economic costs are the productivity losses due to non-communicable diseases. Disease burden was calculated in disability-adjusted-life years. Financial and economic costs were expressed in 2000 US$. Costs and consequences of scaling up preventive interventions were calculated as variations from the baseline scenario. Figure 7.1 Modeling the impact of preventive interventions - Interventions, risk factors and non-communicablediseases llitus 7.5 Inthe baseline scenario, future burden of disease was calculated building on the findings of Global Burden of Disease Study (GBDS).79 The GBDS provides country- specific projections of disease burden in disability adjusted life years (DALYs) based on historical data prior to 1996. The GBDS estimates of future burden of disease were adjusted by more recent information on country-specific burden of disease provided by the Projeto de Carga de Doenp do Brad (PCDB).80 7.6 The analysis of future economic costs in the baseline scenario hinged also on the information provided by the GBDS and PCDB. For each non-communicable disease, the average number of years of life lost and years of inability to work was estimated by 45 age and sex. The conversion of years of life lost and years of inability to work into monetary terms followed the human capital approach, assuming productivity losses per year of life lost and per year of inability to work as per capita GDP. Information on current GDP per capita was drawn from the World Development Indicator Database and adjusted for future years based on information about current real GDP growth from the same source.81 7.7 Financial costs were based on information about the average length of life with a certain disease retrieved from the GBDS and PCDB and average annual treatment costs for individual diseases.82 Costs of scaling uppreventiveinterventions 7.8 If not otherwise indicated, cost data include the incremental cost accruing to agencies responsible for implementing or scaling up the intervention (Table 7.1). Cost estimates do not include costs to patients, such as time or travel costs. Unit cost data Y were drawn from the literature and project reports. When Brazilian data were expressed e 7.1 Intervention design, scope of coverage increase, unit costs and lag time Intervention Intervention design, scope of coverage increase, unit costs andlag time between implementation and impacts Compre- Design: Agita S6o Pado is a comprehensive community campaign to promote physical hensive activity. It targets the entire population o f S5o Paulo State with special emphasis on school community children and workers. It employs multiple strategies, including mass media campaigns, campaign to large promotional events, exercise and walking classes and groups, worksite health promote promotion programs, stage-basedbehavior change, school events, physical education, and physical environmental and policy changesa3 activity Scope of coverage increase: The analysis estimates the cost and benefits of covering an additional 25% of the Brazilian population. Unitcosts: The low unitcost estimate of US$0.004 per person covered reflectsthe financial costs of the program.82 The high unit cost estimate of US$0.06 per population includes voluntary contributions of organizations and individuals.82 Lag time: none Provision of Design: treatment of uncomplicated arterial hypertension with first-line drugs through the anti- Brazilian Family Health Program (Prugrumada Sau'deda Familia, PSF). The PSF hypertensive includes regular home visits of enrollees by `enhanced health teams'. treatment Scope of coverage increase: 25% of all cases of arterial hypertension. Unitcosts: The low unitcost estimate of US$0.72 per person per month reflects drug costs, assuming provision can be piggy-backed on the PSF at no The high unit cost of US$2.69 reflects the costs of drugs and service.85 Lag time: none Tobacco tax Design: The intervention includes lobbying and advocacy activities sufficient to sponsor increase and implement a tobacco tax increase resulting in a 10% increase of cigarette prices. Costs: The monthly costs for lobbying and advocacy were estimated as US$ 5,050.86It was assumed that the activities mustbe sustained over a period of 18 months to achieve the intended result. Lag time: 18 months Medical Design: The intervention includes the one-time counseling of smokers by a physician. It counseling excludes the provision of nicotine-substitution therapy. of smokers Scope of coverage increase: The analysis estimates the costs and benefits o f providing counseling services to 25% of all smokers. Unitcosts: The costs of an individual counseling session were estimated as US$ 11.40!' Lag time: none 46 in Reais, data were inflated or deflated to the base year 2000 and converted to U.S. dollars using the average conversion rate in 2000. When data originated from studies outside Brazil, data were adjusted for purchasing power parity and converted into Reais, inflated or deflated to the base year 2000 and converted into U.S. dollars. Future costs were discounted at an annual rate o f 3%. Estimated effectiveness of preventive interventions Table 7.2 The estimated effcctivenessof preventiveinterventions publications from Brazil.'* Risk factor Intervention Effectiveness in Table 7.2 presents the reducing risk factor (likely scenario) estimates of effectiveness Arterial Treatment with 40% derived from this review Hypertension antihypertensive drugs and used in the 'likely' Physical Inactivity Agita Slio Paulo 3.2% scenario. In the cases of Smoking Tobacco Tax increase 5% arterial hypertension and Medical counseling 3% smoking these reflect rather Allowingfor uncertainty: sensitivity analyses 7.10 Likely, high and low estimates were made for the costs and impact of scaling up preventive interventions. For costs, the variation across scenarios reflects different assumptions about unit costs (Table 7.1). For impact, the variation reflects different assumptions about the effectiveness of interventions on the prevalence of risk factors. Assumptions are primarily based on information drawn from the literature. For net cost savings, cost-effectiveness ratios and benefit-cost ratios, the analysis includes likely, worst and best case scenarios. The likely case scenario used likely cost and impact estimates in the nominator and denominator. The worst case scenario compares highcost estimates with low impact estimates and, conversely, the best case scenario compares low cost estimates with high impact scenarios. Results The Baseline scenario 7.11 Under current levels of prevention, care and treatment, the future disease burden attributable to the five secondary diseases and the three risk factors was estimated as 16.9 million DALYs over the long run from the disease incidence from 2005-2009 (Table 7.3). (For comparison, the total burden of non-communicable diseases totaled approximately 24 million DALYs in 1998.) Around 44% o f the future disease burden was attributable to a single disease--ischemic heart disease. And about 56% of the disease burdenwas attributable to a single risk factor--arterial hypertension. 7.12 The estimated financial costs of treating patients with either one or a combination of the five diseases were US$ 34 billion. The total financial and economic 47 costs were US$ 72 billion. The financial and economic costs of the baseline scenario are huge-about 10% of GDP in 2003. Smoking as a risk factor accounted for 35% of the financial costs. The risk factor arterial hypertension accounted for 31% of financial costs. More than 75% of the financial costs would be spent on the treatment o f ischemic heart disease. Ischemic heart disease also accounted for the majority of combined financial and economic costs (55%). Abbreviations: IHD: Ischaemic Heart Disease; CVD: Cerebro-vascular disease; DM: Diabetes mellitus; COPD: Chronic obstructive pulmonary disease. Note: Inthe case of physical inactivity, C V D reflects ischaemic stroke only; Cancer includes cancer of the trachea, bronchi and lung. Costs of scaling uppreventive interventions 7.13 Costs of scaling up preventive interventions reflect net present values over the time period 2005 and 2009 discounted at a rate of 3% and expressed in US$2000 (Table 7.4). The most expensive intervention by far i s the anti-hypertensive treatment, at $967 million. Medical counseling of smokers by physicians was also costly, at $474 million. Agitu Siio Paulo and the tobacco tax increases were the least costly, at $33 million and 0.9 million, respectively. Table 7.4 The costs of scaling uz)selected Dreventive interventions iUS$. 20001 Low cost High cost Intervention scenario scenario Likely scenario A Comprehensivecommunity campaign to promotephysical activity (25% of population) $4,186,000 $62,784,000 $33,485,000 B IIAntihypertensive treatment (25% coverage) II$158,515,000 II $1,776,849,000 11 $967,682,000 C Tobacco tax increases (10% increase in prices) $90,000 D Medical counselingof 25% of smokers $473,844,000 Total for Package A+B+C $1,001,257,000 48 Consequencesof scaling uppreventive interventions 7.14 Table 7.5 presents the long run benefits (averted disease burden, financial, and economic costs) stemming from the reduction in disease incidence from 2005 to 2009 due to the preventive interventions carried out over those same five years. Each intervention has a likely, low and high impact scenario, reflecting different assumptions about the impact of the intervention on the reduction of the risk factor. Financial and economic data are presented as net present values discounted at a rate of 3%. Table 7.5 Scaling up selected preventive interventions Reductions in (1)burden of disease; (2) financial costs; and (3) financial and economic costs (1) Burden of disease [DALYs] (2) Financial costs [Million US$, 20001 (3) Financial and economic costs [Million US$, 20001 Note: Benefits reflect variations from the baseline scenario 7.15 Key findings: 0 Providing anti-hypertensive treatment 25% of hypertensives would have the biggest impact on both disease burden and costs: it would reduce the burden of non- communicable diseases by 646,000 DALYs, would reduce treatment costs by US$ 482 million and avoid financial and economic losses of US$ 2,153 million. 49 e Scaling up comprehensive community campaigns to promote physical activity would generate the second largest impact on both disease burden and costs. It would result in a reduction o f the non-communicable disease burden by approximately 135,000 DALYs. It would reduce treatment costs for non-communicable diseases by US$ 348 million and avoid financial and economic losses o f US$622 million. e A tax increase resulting in a 10% increase in cigarette prices would reduce the burden of non-communicable diseases by 63,700 DALYs, save US$ 171 million in treatment costs, and avoid financial and economic losses of US$306 million. e Medical counseling of 25% of smokers would decrease the burden of non- communicable diseases by 48,000 DALYs-about a quarter fewer DALYs than the increase in cigarette prices, above. Savings in treatment costs would amount to US$ 32 million and US$57 million of financial and economic losses would be avoided. e The first three interventions as a package would reduce the burden of non- communicable diseases by 845,000 DALYs-equivalent to a 5% decrease of the 16.9 million DALY disease burden estimated for the baseline. It would generate savings o f financial costs of roughly US$ 1 billion, equivalent to 3% o f the costs of treating non- communicable diseases. Avoided financial and economic losses would total US$ 3,082 million, or 4.3% o f the expected economic costs o f the baseline scenario. Potential net cost savings of scaling uppreventive interventions 7.16 The agencies' costs of scaling up preventive interventions were compared with the expected treatment in order to identify potential net cost savings for three different scenarios (Table 7.6). The scaling up comprehensive community campaigns for physical activity like Agita Sdo Paulo and the increase in cigarette prices promised net cost savings (savings in treatment costs outweigh the costs of implementing the intervention). In the likely cost scenario, Agita Sdo Paulo yielded net cost savings of US$ 310 million and the increase in tobacco taxes yielded savings o f US$ 170 million (note this does not take into account additional tax revenue from tobacco tax increases). Note: Inthe likely case reventive interventions scenario, the likely cost scenario for scaling up i s compared with the likely impact scenario for financial cost savings. The worst-case scenario reflected the high cost scenario for scaling up and the low impact scenario for financial cost savings. Conversely, the best-case scenario compared the low cost scenario for scaling up and the high impact scenario for financial cost savings. Inthe absence of low and high cost estimates, cost estimates o f the likely scenario substituted for missing data. 50 For the anti-hypertensive treatment, net cost savings only occurred in the best case scenario. For medical counseling of smokers, the costs of scaling up the intervention outweighed savings in treatment costs in all three scenarios. D Medicalcounselingof smokers $10,697 $9,360 $9,984 0.1 0.1 0.1 Total for PackageA+B+ C $1,432 $984 $1,184 2.5 3.8 3.1 Benefit-cost ratiosfor the scaling up of preventive interventions 7.18 Benefit-cost ratios are estimated as the ratio of savings of economic costs over the costs of scaling up the interventions. Benefit-cost ratios were greater or equal to one for all interventions and in all scenarios except in the case of medical counseling for smokers (Table 7.7). In the latter case, benefit-cost ratios consistently fell in the range o f 0.1. The increase in tobacco taxes produced across all scenarios benefit-cost ratios above 1,000. Agitu Sa"oPuulo generated a ratio of roughly 20 in the likely case scenario. In the same scenario, the ratio was approximately 3 for the package of interventions and 2 for the provision of anti-hypertensive treatment. DISCUSSION CONCLUSIONS AND 7.19 Continuing with the status quo would be a costly proposition. For the baseline scenario, the model predicted, for the period 2005 and 2009, the future disease burden at 16.9 million DALYs for the three risk factors and five related diseases. For comparison, the total burden of disease was 24 million DALYs in 1998. Most notably, roughly 50% o f the future disease burden was attributable to a single risk factor--arterial hypertension. 51 The predicted future disease burden due to physical inactivity, arterial hypertension and smoking would cause treatment costs o f US$ 34 billion and additional U S $ 38 billion of productivity losses. Expected treatment costs and productivity losses amounted to 10% of GDP in 2003. 7.20 Interventions exist that, if introduced or scaled up, would significantly reduce the burden o f disease. The evaluation included four of them: scaling upAgita Siio Paulo and anti-hypertensive treatment, introducing higher tobacco taxes and providing medical counseling to smokers. The first three would reduce the future burden of non- communicable diseases by roughly 850,000 DALYs or 5% of the future disease burden attributable to the three risk factors. While this reduction seems minor in relative terms, the financial implications are significant. More than US$ 1 billion of treatment costs and US$ 2 billion of productivity losses would be avoided. Furthermore, it i s important to note that the model used a conservative approach to scaling up with rather small increases incoverage of few interventions. 7.21 The model predicted that the scaling up of comprehensive community campaigns similar to Agita S6o Paulo, the increase in cigarette prices and, under favorable assumptions, the scaling up o f anti-hypertensive treatment would create net savings. In other words, these interventions would generate savings in treatment costs that would outweigh the costs of scaling up or implementing the intervention. In the likely scenario, these savings would amount to US$ 310 million for comprehensive community campaigns and US$ 170 million of the increase in tobacco taxes. For the same scenario, the combination of all three interventions would be cost neutral. 7.22 Investing in the tested interventions promises good value for money. Interventions with costs per DALY o f less than twice the GDP are generally considered highly cost-effective.88The cost per DALY for Agita Siio Paulo, the treatment with anti- hypertensive drugs and an increase in tobacco taxes were estimated as ranging between U S 1.40 and US$ 1,500, thus, falling clearly below this threshold. Similarly, benefit-cost ratios for these interventions seem highly attractive with quotients o f 2.2, 18.6 and >1,000. The cost-effectiveness of the package of all three interventions was estimated as approximately US$ 1,200, the corresponding benefit cost ratio as 3.1. 7.23 The option of scaling up medical counseling of smokers proved economically less attractive. However, the assumptions regarding the effectiveness o f the intervention were conservative and cheaper delivery models might be feasible. 7.24 Overall, the study showed that the scaling up of community campaigns to promote physical activity, the treatment o f uncomplicated hypertension with first-line drugs, and tobacco tax increases should be financially and economically attractive, and will help lessen Brazil's burden o f non-communicable diseases in the future. However, much more could be done beyond scaling up these three key interventions. A more comprehensive reform program would involve refocusing the health system much more fundamentally towards non-communicable disease prevention and control. The tools presented in this chapter could be used to explore the likely impact o f more fundamental reforms, as well. 52 8. CONCLUSIONSAND RECOMMENDATIONS 8.1 Brazil, along with nearly all of the countries of the region, i s undergoing a demographic and epidemiological transition that i s a consequence of urbanization, improvements in health care, changing lifestyles and globalization. In some region of Brazil the transition carries a double burden in that there are high rates of both communicable diseases and NCDs (Northeast) while the more developed regions (South and Southeast) have reduced the burden from communicable diseases but still have a high burden from NCDs. The double burden is also found among the poor who carry an unfair share of both communicable diseases and NCDs. 8.2 Emerging information from Brazil confirms what i s found in OECD countries and puts to rest the myth that NCDs are a problem of the rich and not the poor. Recent studies in Brazil reveal that many risk factors that cause NCDs such as heart disease, strokes, cancer and diabetes are more prevalent among the poor and less educated (a proxy for socioeconomic status). Among Brazilian women obesity i s increasingly linked to poverty and this trend i s worsening. RFS data show that less educated Brazilians are more likely to smoke and less likely to eat a healthy diet that includes fruits and vegetables. Less educated women are less likely to have PAP smears and mammograms. RFS data also suggests a higher prevalence of hypertension, diabetes and heart diseases among the less educated. While good data from Brazil on morbidity and mortality by income, social class, or education levels are not available, the presence of higher levels of these risk factors i s almost certainly associated with higher disease and disability levels. 8.3 Unlike communicable diseases, NCDs are chronic conditions that place a burden both on the individual--who must obtain treatment on a continual basis to control the disease and prevent, if possible, disability-and on the health care system. Where the primary health care system does not function properly, chronic diseases are more likely to progress to unnecessary disability, costly hospitalizations, and premature mortality. 8.4 Fortunately, most NCDs can be prevented. The evidence for interventions that are likely to have decrease exposure to risk factors and prevent NCDs i s reviewed in this document. Interventions can target the entire population or community, or they can target individuals. Population or community-based interventions generally have greater impact and are more cost-effective. These are interventions that promote health and contribute to the primary prevention of NCDs. 8.5 In cases where NCDs cannot be prevented, the disability and early mortality they cause can often be prevented. Population-based interventions can contribute to reductions in morbidity from NCDs but need to be accompanied b y interventions in primary care settings that provide effective, long-term treatment to people with NCDs. 8.6 Inthe case of Brazil, the cost of treating NCDs accounts for nearly half the cost of hospital admissions. Demand will only increase as the health transition continues to evolve. At the same time, N C D burden of disease rates for Brazil are much greater than those for Canada, Cuba and the USA suggesting that much of the burden i s avoidable. 53 8.7 An important question relates to the potential financial and economic impact of implementing effective health promotion/NCD prevention activities in Brazil. This study carried out the first such assessment looking at the impact of four specific interventions. The analysis predicted that the increase in cigarette taxes, the scaling up o f comprehensive community campaigns to promote physical activity similar to Agita S6o Paulo, and, under `favorable' assumptions, the scaling up of anti-hypertensive treatment would create net savings. In the `likely' scenario with more conservative assumptions, these savings would amount to US$ 310 million for comprehensive community physical activity campaigns and US$ 170 million of the increase in tobacco taxes, and the combination of all three interventions would be cost neutral. They would reduce the burden o f NCDs by 845,000 DALYs or approximately 5% of the total when compared to maintaining the status quo. 8.8 Primary prevention activities that are targeted at broad populations are generally more effective in preventing NCDs and less costly than those targeted at individuals. However, in Brazil until recently such activities have been scattered and unfocused. There i s no national plan for N C D / risk factor prevention or health promotion. N C D prevention activities tend to be run by vertical programs with limited coordination between them. INCA has led the tobacco control program for over ten years. Brazil has established many of the tobacco control activities considered cost-effective, including limitations on advertising and promotion, health education, warning labels, and bans on smoking in certain public places. (Tax increases - the most cost-effective - have been minor). Cigarette smugglingremains a major issue. Information to measure the tobacco control program's success i s not available. Attempts to scale up to a nation-wide level comprehensive community programs similar to the highly cost-effective Agita S6o Paulo program are just now being considered. Few other activities related to NCD prevention are being promoted at the national level. The recent transfer of responsibilities for the coordination of health promotion activities to the SVSDHSA is a hopeful sign. This department has responsibility for surveillance and control of NCDs and injuries, among others, and has more o f the technical capacity necessary to lead a program for NCD prevention. 8.9 Brazil has responded to the health transition with measures to improve secondary prevention. It has focused on increasing the coverage and quality o f primary care services via the Family Health Program (FHP). It has developed plans to improve care for those with diabetes and hypertension and has carried out national screening campaigns to identify people with these diseases. Coverage i s still reported to be very low - only approximately 10%.At the same time the FHP has included health promotion as one of its priorities. However, there are no guidelines for how this should be carried out. 8.10 The bipartite and tripartite `pactos' for performance measurement at the state and municipal levels have very few indicators related to NCDs and no indicators related to exposure to risk factors. While the `pactos' are in jeopardy o f being overloaded, the fact that there i s not one indicator related to NCD risk factors suggests that addressing them i s not seen as a priority. Changes in NCD levels (for example, heart disease from smoking) may lag several years behind the implementation of interventions. However, 54 changes in risk factors (for example, smoking prevalence) can occur much more quickly so can be useful as annual indicators of performance. 8.11 At this time an information system to assess the prevalence of risk factors and NCDs does not exist. It i s required in order to measure the intervention impact. The dearth of information both contributes to the lack of activity in this area (out of sight, out of mind) and also makes it impossible to know whether programs are having an impact (for example, tobacco control). The Brazil Risk Factor and NCD Survey carried out for the first time in 2002/3 provides a baseline in 16 cities and this i s a start, along with the World Health Survey. RECOMMENDATIONS 8.12 Table 8.1 at the end of this chapter provides a summary of recommendations on key strategies for the short and medium and their expected impact. Develop and implement health strategies to prevent NCDsby focusing on risk factors 8.13 Primary prevention of NCDs i s achieved by preventing exposure to the risk factors that cause these diseases. This i s generally more cost-effective than trying to provide good health care for the diseases after they occur. Primary prevention i s carried out in two ways: i) population-based activities such as regulation, education, communication, and interventions in communities, schools, and workplaces; and ii) information and activities directed at individuals by their health providers. Specific priority interventions to reduce smoking, promote dietary change, increase physical activity, and strengthen screening and treatment of hypertension and diabetes are detailed in Table 7.1, along with other policy recommendations from this paper, and in Chapter 5. For smoking, priority actions include increasing taxes, better controlling tobacco sales to minors, and strengthen measures to control tobacco smuggling. For diet, promoting healthy dietary choices through schools, community groups, and the media, legislation and/or encouraging voluntary action to reduce the use of partially hydrogenated fats and salt levels in processed foods, and continuing to strengthen the dietary quality of school lunches are key measures. For physical activity, scaling up the successful Agita Silo Paul0 to cover much more of Brazil, supporting physical activity in the schools, and encouraging environmental design that facilitates walking, bicycling, and greenways for exercising are key. Together these measures could make a huge change in reducing the projected burden from coronary heart disease, strokes, diabetes, and certain cancers in Brazil. 8.14 Population-based prevention requires effective leadership. The focus of NCDs in Brazil has been on improving treatment, with the important exception of smoking. How can the focus best shift to the prevention of exposure to risk factors and increased emphasis on primary prevention?A first step has been taken with the formation of a health promotion section in the SVSDHSA and its clear definition and recognition as thefocal group within the MOH at the national level is needed to provide leadership on development, implementation, and evaluation of cost-effective population-based NCD 55 / risk factor prevention activities. Many NCD/ risk factor prevention activities are multi- sectoral and this group will need to coordinate not only with state and municipal departments of health and other health agencies such as INCA, but other ministries, academia, and civil society all of whom are important in a multi-sectoral approach to NCD / risk factor prevention. The mandate of this group should be to develop strategies and provide leadership on population-based activities. The close links this group has to the developing Brazilian population-based risk factor information systems i s ideal since these systems will provide part of the evidence base for the development o f well-directed actions as well as information for the evaluation of their impact. 8.15 One of the first tasks of the group assigned to NCD prevention would be to develop a national health plan for NCD / risk factor prevention. The NHPP soon to be launched will form the basis for such a plan. Specific short and medium term goals will need to be set at the national and state levels to assess whether the plan i s being successfully implemented and having an impact. The elaboration o f a plan would benefit from the formation of a multisectoral national advocacy body consisting of civil society, NGOs, academia and others who have critical roles to play if such policies are to be effectively implemented and evaluated. Their involvement as key stakeholders and partners in N C D /risk factor prevention efforts i s important earlier rather than later. Furthermore, the implementation of effective regulatory policies for NCD / risk factor prevention may encounter obstacles from related industries and lobbies. A national advocacy body could be a counter-force to such pressures. 8.16 Strengthen capacity. The current capacity to address N C D / risk factor prevention and scale up primary prevention activities i s weak and will need to be strengthened within the MOH at each of its three levels (national, state and municipal), as well as in other sectors. Primary prevention involves communication and information for promoting changes in behavior. These require special skills in a relatively new area of social sciences including the behavioral sciences. Increasing regulatory capacity will also be important. Capacity building and the development o f training institutions that can partner with the government to assure the necessary skills for primary prevention require a long-term strategy and commitment for its implementation, and would be an important component o f the national N C D prevention plan. 8.17 Provide financing for development of primary prevention activities that are identified as potentially effective in Brazil and that address the present and emerging risk factors including, among other, smoking, physical inactivity, poor dietary intake and obesity. The economic analysis shows that some population-based interventions are actually cost-saving in the long term. However, to obtain these cost-saving benefits requires up front investment either by increasing the budget or reallocating budget. Risk factor prevention activities could, in large part, be self-financing if increases in taxes related to risk factor control such as for tobacco and alcohol are implemented and earmarked for health. 8.18 Improve FHP /primary careproviders capacity to deliverprimary prevention ' / health promotion activities to individuals under their care. While not generally as cost-effective as population-based interventions, primary care providers have an 56 important role to play in promoting health behaviors and preventing NCDs. InBrazil, one of eight priorities for the FHP i s health promotion. However, this area of work i s still in the early stages of development. Cost-effective interventions to assist individuals in reducing exposure to risk factors are available. An assessment of which of these interventions can be financed by the FHP i s necessary, included in an FHP health promotion strategy, and implemented. The FHP i s developing and implementing a performance management system using multiple indicators. Performance indicators for health promotion should be developed and included in the system. Improve surveillanceand monitoring of risk factors and prevention strategies 8.19 Improve NCD /risk factor surveillance. Information on the prevalence of NCD risk factors in Brazil is limited. At the present there is no systematic, ongoing system for obtaining risk factor information. Given the current and growing burden of NCDs the surveillance of risk factors i s as important as the surveillance of communicable diseases. Information i s required to assure that the prevention interventions that are implemented are having the desired impact. The World Health Survey provides a national baseline for several risk factors, and the RFS, with more extensive risk factor information, provides a baseline for the 16 capital cities in which it was carried out. The frequency and sampling of future risk factor surveys needs to be determined as part of the national NCD prevention plan. Ideally, a system providing continuous information on risk factor prevalence will eventually be implemented. Similar to communicable disease, N C D / risk factor surveillance information should be linked to actions for their prevention and control. Information obtained from N C D / risk factor surveillance should include information about socioeconomic status, education levels, ethnicity and other population characteristics that could help target appropriate interventions. 8.20 Include risk factor prevention indicators in the bipartite and tripartite `pactos'. The fact that none of the state or municipal `pactos' include indicators to measure a reduction in exposure to risk factors reflects the lack of national and state policies and strategies for their control and the lack of information on their prevalence. NCD indicators currently included in the pactos are related to improvements in health care delivery at the primary care level. Inclusion or indicators to measure success in the more cost-effective area of primary prevention would be important incentives to implement and prioritize these activities. Improve secondary preventionof NCDs through better health care and screening 8.21 The focus of this paper i s on the primary prevention of NCDs by decreasing and eliminating exposure to risk factors. However, until this i s achieved the burden o f NCDs in the population will continue to grow. Premature mortality, disability, and the costs of hospitalizations can be prevented through better health care particularly at the primary care level. The Brazilian MOH recognized the growing importance of NCDs when it included the control of diabetes and hypertension as national health priorities. 8.22 Strengthen the FHP's capacity to respond to the NCD epidemic. Actions to improve the clinical management and control o f hypertension and diabetes are already 57 underway in Brazil. Cost-effective interventions are available, particularly for hypertension. However, the context of long-term chronic diseases requires shifts in the way health care i s provided. Issues that need to be addressed and enhance effectiveness of disease control are related to long-term adherence to treatment, continuity of care, integrated care and the ability of patients to self-manage their own disease. These in turn require more effective use of personnel (particularly non-physicians), training, and improvements in organizational and management aspects of primary care. 8.23 Furthermore, the prevention of premature mortality and disability is more likely with early detection of NCDs. This requires investments in screening for NCDs particularly hypertension, diabetes, and some cancers. Screening in turn depends on its uptake and investments will be needed to increase the demand for screening as well as for treatment by the population. Strengthen research on effective NCD/ risk factor prevention 8.24 There i s very little information specific to Brazil on the effectiveness of primary and secondary prevention interventions. Most of the information on NCD and risk factor prevention comes from developed countries with very different cultural contexts. Many of the interventions require changes in lifestyle and behaviors or are applied in settings like schools, communities and workplaces where cultural context i s important. Research i s urgently needed that provides evidence for the effectiveness and costs o f interventions for the primary and secondary prevention of NCDs in Brazil. 8.25 Increase capacity to carry out prevention effectiveness research. Behavioral research uses methodologies that are very different from those used for interventions implemented during the 20thcentury such as vaccines and drug treatments. Investments are required for both training research scientists in these areas as well as developing research institutions that can support prevention effectiveness studies. 8.26 Assess the effectiveness of prevention interventions among the poor. The poor carry a higher burden o f NCDs, however, many interventions such as those using mass media, information requiring skills like reading or interpretation, demand for screening, and taxes among others, often do not reach the poor. Research identifying interventions that are effective among the poor and other sub-populations (ethnicity, race, region) are urgently needed to assess and assure the equality of prevention activities. This i s true in developed countries as well as inBrazil. 58 1 % e, I a h .-i a 5 e, *0 3 e Y C 0 0 e, Y m a Pr '+ 0 x .3 Y 3 2 O 3 vl 6 2C C Y v3 e e, 2 C 0 D 8 C Y 4 3E0 9. ANNEX 1 Table Al. Lifeexpectancyat birth in LAC countries, 2001 Country 2001 Cuba 76.9 Chile 76.3 Costa Rica 76.1 Uruguay 75.0 Panama 74.9 Barbados 74.4 Mexico 74.2 Argentina 73.9 Dominica 73.8 Venezuela, Bolivarian Republic of 73.6 Jamaica 72.7 Bahamas 71.9 Saint Lucia 71.3 Antigua and Barbuda 71.O Saint Kitts and Nevis 70.9 Colombia 70.7 Paraguay 70.6 Ecuador 70.3 Saint Vincent and the Grenadines 70.0 Belize 70.0 Trinidad and Tobago 69.9 Nicaragua 69.5 El Salvador 69.5 Brazil 68.7 Peru 68.5 Suriname 67.4 Honduras 67.3 Grenada 67.2 Dominican Rewblic 67.0 Guatemala 66.2 Guyana 64.0 Bolivia 62.7 Haiti 50.0 Source: WHR2002.AnnexAll LAC 70.7 fable 1. 63 Table A2. Prevalence of selected riskfactors in American countries by mortality stratumyear 2000* Mortality stratum Risk factor" Prevalencecriteria AMRB AMRD MeanA MeanB MeanD Alcohol Proportionconsuming alcohol 67% 66% 62% 85% 37% 19% Blood pressure Mean systolic pressure (mmHg) 127 128 128 134 131 130 Childhood sexual abuse Proportionof adults with historyof abuse 15% 9% 15% 12% 17% 26% Cholesterol Meancholesterol (mmolll)"' 5.3 5.1 5.1 5.6 4.0 5.0 Indoor smoke from solid fuels Proportion using biofuel 1% 25% 53% 0% 32% 71% Iron deficiency Mean haemoglobin level (giDI) 13.7 13.1 13.1 13.7 12.8 12.3 Low fruit and vegetableintake Average intake per day (9) 290 190 340 430 320 317 Overweight Body mass index (kg/rn2) 26.9 26.0 26.0 25.0 24.4 21.2 Physicalinactivity Proportionwith no physical activity 20% 23% 23% 17% 17% 16% Underweight Proportion less than 2 SD weight for age 2% 5% 12% 3% 14% 34% Proportion not using modern Unplanned pregnancies contraception 33% 45% 68% 39% 62% 79% Unsafeinjection(s) exposing to Hepatitis Unsafe health care injections B each year 0% 0% 0% 0% 3% 9% Concentrationof particles less 10 micron Urban air pollution (IMm3) 13 15 20 13 24 25 Proportionvitamin A deficientwith night Vitamin A deficiency blindness 0% 9% 9% 0% 10% 17% Proportion not consuming US Zinc deficiency recommendedintake 6% 26% 68% 4% 20% 54% * factor - Alcohol, childhood sexual abuse, physical inactivity are ages 15+; Blood pressure, cholesterol, overweight, and fruit and Estimates are age standardized to the WHO reference population and the denominator is the population most relevant to the risk vegetables are ages 30+; iron, vit A, zinc, and underweight are for 4;and for unplanned pregnancies females 15-44 ....* Many risk factors were characterized at multiple levels -- here they are collapsed to show exposure or no exposure (or means). 1 mmol/ = 38.7 mg/dL 64 Table A3. Attributable mortality by specific risk factor and mortalitystratum, 2000 AMR A AMR B AMR D Childhoodand maternal undernutrition 0.2% 2.4% 8.8% Underweight 0.0% 1.0% 4.9% Iron, Vit A or zinc deficiency 0.2% 1.4% 4.0% Other diet-relatedrisks and physical inactivity 47.4% 38.9% 23.1% Blood pressure 13.3% 12.8% 7.8% Cholesterol 12.6% 6.5% 3.7% Overweight 9.8% 10.1% 6.5% Low fruit and vegetable intake 6.2% 5.4% 2.7% Physical inactivity 5.6% 4.1% 2.4% Sexual and reproductivehealth risks 0.6% 2.1% 6.3% Unsafesex 0.6% 1.9% 5.5% Lack of contraception 0.0% 0.2% 0.8% Addictive substances 24.0% 18.5% 6.9% Tobacco 23.3% 8.5% 1.2% Alcohol 0.2% 9.5% 5.5% Illicit drugs 0.6% 0.4% 0.2% Environmentalrisks 1.2% 3.8% 8.0% Unsafewater, sanitation and hygiene 0.0% 1.2% 4.5% Urban air pollution 1.0% 1.2% 1.0% Indoor smokefrom solid fuels 0.0% 0.6% 2.0% Lead exposure 0.1% 0.8% 0.6% Occupational risks 1.0% 1.4% 0.4% Risk factors for injury 0.1% 0.7% 0.4% Carcinogensand airborne particulates 0.9% 0.6% 0.0% Other selected risks to health (unsafe health injections and childhood sexual abuse) 0.1% 0.1% 0.4% ~ Total deaths (000) 2 778 2587 510 Attributable % of deaths by all risk factors 74.5% 67.0% 53.9% The combined effects of any group of risk factors in this table will often be less than the sum of their separate effects. 65 Table A4. Attributable DALYs by specific riskfactor and mortalitystratum, 2000 AMR A AMR B AMR D Childhood and maternal undernutrition 1.1% 3.0% 10.1% Underweight 0.1% 1.3% 5.5% Iron, Vit A, and zinc deficiency 1.0% 1.6% 4.6% Other diet-relatedrisks and physicalinactivity 24.7% 14.1% 7.6% Blood pressure 6.1% 4.1% 2.3% Cholesterol 5.4% 2.4% 1.2% Overweight 7.6% 4.3% 2.5% Low fruit and vegetable intake 3.0% 1.9% 0.8% Physical inactivity 2.8% 1.5% 0.8% Sexual and reproductivehealth risks 1.1% 2.2% 4.9% Unsafe sex 1.1% 2.2% 4.9% Lack of contraception 0.0% 0.5% 1.2% Addictive substances 23.9% 16.8% 7.7% Tobacco 13.4% 3.8% 0.4% Alcohol 7.9% 11.7°/o 5.7% Illicit drugs 2.6% 1.4% 1.6% Environmentalrisks 0.8% 4.8% 8.5% Unsafewater, sanitation and hygiene 0.1% 1.6% 4.5% Urban air pollution 0.4% 0.4% 0.3% Indoor smoke from solid fuels 0.0% 0.6% 2.0% Lead exposure 0.3% 2.1% 1.6% Occupational risks 1.2% 1.6% 0.8% Risk factors for injury 0.3% 1.0% 0.6% Carcinogens, particulates, noise, ergonomic stressors 0.9% 0.6% 0.1% Other selected risks to health 0.9% 0.4% 0.6% Childhood sexual abuse 0.9% 0.3% 0.4% Unsafe health care injections 0.0% 0.0% 0.2% Total attributable DALYs(000) 24 72534 399 6 961 Total DALYs(000) 45 99179 562 16 803 Attributable RF % of DALYs 53.8% 43.2% 41.4% Thecombined effects of any group of risk factors in this table will often be less than the sum of their separate effects. 66 10. ANNEX2 METHODSDATA AND FOR ECONOMIC ANALYSIS The impact of three major risk factors i s considered: smoking, physical inactivity, and arterial hypertension. In defining an appropriate methodological approach, certain issues have to be taken into account. The first i s the lack of data on present parameters of NCDs in Brazil. Numerous studies have been carried out on the prevalence of risk factors, however many have important limitations. Some studies have a very narrow context which is difficult to extrapolate to the population at large. Others do not adequately define risk factor thresholds as measured by surveys, for example the exact measure of physical activity below which an individual i s labeled as being inactive. The measurement of the risk factor i s sometimes inaccurate; in the case of the most dynamic of these, arterial blood pressure, inadequate environments for measurement can make subjects anxious and lead to overestimates of the prevalence of hypertension. Despite these types o f flaws and limitations, sufficient data exists on risk factors to be able to make reasonable estimates of population prevalence. A more difficult parameter to establish i s the quantitative relation between risk factors and specific diseases they cause. There is almost no data on this from Brazil, and studies with estimates from other countries have to be used. In doing so, it i s assumed that these quantitative relations are relatively robust and do not vary greatly across different geographic and socio-cultural settings. Similarly, there i s little Brazilian data on the effectiveness o f preventive interventions in reducing the prevalence of risk factors and diseases they cause. Systematically collected data on costs are almost non-existent. It is difficult to use evidence from other countries in Latin America and the Caribbean because this extreme lack of evidence for NCD interventions extends throughout the whole region.34 Therefore, evidence i s usually used from countries in other regions. For effectiveness, study results from other countries are taken as proxies; these can obviously vary significantly but they still provide a useful general estimate of intervention results. Similarly, quality data on costs i s also uncommon to find. For studies from other countries and from WHO databases, costs are adjusted usingInternational Dollars (Purchasing Power Parity); for Brazilian studies, unit costs are assumed to be the same across the country. The second issue i s how to best measure the true impact o f NCDs. It i s evident that diseases causing death at a younger age have a greater impact and this has to be adequately captured by the methodology utilized. Perhaps o f greater importance i s capturing and quantifying the impact of disability (non-fatal health outcomes). This i s harder to measure than absolute events such as mortality; furthermore, disability can have a devastating effect on the exacerbation and propagation o f poverty. Whereas mortality inflicts a one-time loss on a family, ongoing disability may hamper the individual's productivity while causing a chronic drain of resources for health care. Therefore, the 67 methodology to be adopted also needs to adequately reflect the impact of different degrees and durations of disability. While calculation of productivity loss shows the effects of age at death and disability to some extent, it i s a relatively crude measure and i s not sufficient. In fact, economic measures o f illhealth are an important part of any analysis, but the overriding criteria should be impact on health. This is in part due to the intrinsic value of health to policymakers and the population. The third issue concerns the relative burden o f risk factors and diseases in the poor. If poverty i s associated with an increased impact of NCDs, then targeting these diseases may be an effective way of targeting the poor. This could also have important implications on comparative cost-effectiveness of prevention among different economic classes, an important consideration in project design. Inconsidering all these issues, simplifying assumptions were necessary. Although many results in this document are approximate, they are still useful policy tools. In most parts o f the world, the scale and sophistication of N C D control and prevention are still at a basic stage. Therefore, the most pressing need i s not to spend time and resources arriving at exact estimates, but to send out an urgent call for action with whatever evidence i s on hand. This i s what the document sets out to achieve. Methodological Approach - The conceptual approach i s based on the existence of risk factors, each of which leads to an increased occurrence o f specific secondary diseases, in turn bringing about negative health impact and financial and economic losses. For each risk factor, the initial parameter i s its prevalence in the population. Results of different studies are considered in a broad context to define the appropriate overall value for the population. Characteristics such as the time period, geographic location, and characteristics o f the study population are considered in arriving at this definition. Definitions and methods of measuring the risk factors are also compared as it i s difficult to compare studies with important measurement incompatibilities. Risk factor definitions most easily related to consequence disease incidence are usually given preference. Each risk factor increases the probability o f certain secondary diseases. The quantitative relation between risk factors and diseases i s derived from the literature, using the World Health Report 2002 as a broad base (WHO 2002). The baseline is the incidence" o f the secondary diseases in the population not exposed to the risk factor. Subsequently, two basic approaches can be used according to data availability: 1. the use of an absolute disease incidence in populations exposed and unexposed to the risk factor; 2. the use of the absolute disease incidence/prevalence in unexposed individuals along with a measure of increased risk in exposed individuals. The calculations using these two approaches would be thus: " Or prevalence as may be the case. All mentions of incidence assumethe possible alternative use of prevalence according to the nature of the disease in question. 68 1 - IA-riskfactor=IA-exp-IA-unexp 2. IA-riskfactor=IA-unexp * (RRA-exp- 1) 1A-fiskfactor = Incremental Incidenceof Disease A per 100,000 individuals secondary to risk factor exposure IA-exp=IncidenceofDiseaseAper100,000individualsexposedtoriskfactor IA-unexp=IncidenceofDiseaseAper100,000individualsunexposedtoriskfactor RRA-exp = Relative Risk of Incidence of Disease A in individuals exposed to risk factor compared to unexposed individuals Ina situation where only the incidence of a secondary disease inthe general population is known, together with the relative risk of exposure to the risk factor, and the prevalence of the risk factor in the general population, the incidence of the disease attributable to the risk factor can be calculated inthe following manner: IA-riskfactor IA-general *={Priskfactor-general*(RRA-exp -1)1/ { [Priskfactor-general *( M A - e x p- 111-k 11 IA-fiskfactor=IncrementalIncidenceofDiseaseAper100,000individualssecondarytorisk factor exposure IA-general= Incidence of Disease A per 100,000 individuals in the general population Pfiskfactor-general= % Prevalence of the risk factor in the general population RRA-exp = Relative Risk of Incidence of Disease A in individuals exposed to risk factor compared to unexposed individuals The incidence of secondary diseases are calculated drawing on data and estimates from the Brazilian as well as the Global Burden of Diseases as well as individual studies. In those cases where prevalence data and/or specific information on relative risk are not adequate from individual studies, heavy use i s made of the calculations and estimates made in the World Health Report 2002 of the burden of disease attributed to specific risk factors. Since different diseases have different impacts, it i s necessary to translate health impact into comparable units. This is done utilizing the Global Burden of Disease methodology. This approach allows the addition of impact from years of life lost due to premature mortality with impact from years of life lived with disabilities of diverse degree. For both, impact i s translated into a single indicator so the relative effect of mortality and disability can be compared or combined. It also makes it possible to compare, add, or subtract the impact of different diseases. The basic equation for this methodology is: DALYj=YLLj YLDj + DALYj =Disability-Adjusted Life Years due to Diseasej YLLj = Years of Life Lost due to premature mortality (based on life expectancy at age of death), due to Diseasej 69 YLDj = Years Lived with Disability (based on duration of disability and Disability Weight), due to Diseasej As mentioned above, disability is quantified by assigning a Disability Weight (DW) to capture the severity of the incapacity caused by the disease. The higher the DW, the more severe i s the incapacity. This number ranges from 0 (no disability) to 1 (total disability). For example, vitiligo on the face has a DW of 0.020, Angina 0.223, blindness 0.624, and quadriplegia 0.895. Death has a DW of 1 as it i s by definition a total loss of the year. Therefore, a year lived with a disability is treated as a partial loss of a year and a year lost to mortality is treatedas atotal loss of a year. The methodology also factors in a future discount rate, usually set at an annual 3%, and an age-weighting factor which gives differing values to years lived at different ages. For reference, the formulae for calculating the two components of DALYs for one individual are: YLLS = KCera/(r+~)2[e'"P"L+a'[-(r+P)(L+a)-1] - e-(r+p)a[-(r+p)a-l]] + (l-K)/r( l-e'L) YLDs = D KCera/(r+p)2[e-(rf { P"L'a'[-(r+~)(L+a)-l]-e-(r+P)a[-(r+P)a-l]]+( 1-K)/r(1-e- rL> 1 p = Parameter for age-weighting function (usually 0.04) K =Age-weighting modulation factor (usually 1) C = Constant (usually 0.1658) a = Age at death L=Standardexpectation of life at age a e = Natural logarithm r =Discount rate (usually 0.03) D=Disabilityweight Therefore, once the discount rate, age-weighting factor, and Disability Weights are defined," the total impact from a given disease as a result of a given risk factor is a function of the following parameters:"" Impactfiskfactor=f (total incidence due to risk factor, age at incidence, degree of disability, duration of disability, total mortality due to risk factor, age at mortality, life expectancy by age) ~~ ~ viInthis analysis, the values used inthe Global Burden of Disease Study are adopted. For all parameters of the function except total incidence and total mortality, the distribution of values in the population or the weighted average in the population may be used. 70 Once impact i s calculated in DALYs, the overall health impact caused by a risk factor i s simply the sum of the impact caused by the incremental incidence of all its secondary diseases (Figure C): n Impacttotal-riskfactor = 1j = lImpactj-riskfactor ImpaCttotal.riskfactorTotal impact caused by the risk factor = Impactj.,iskfactor=Impact causedby risk factor through increased incidence of Diseasej n =Total number of secondary diseases causedby the risk factor For subsequent calculations, the risk factors and the secondary diseases selected are as follows: 1.Arterial Hypertension 0 Ischemic Heart Disease 0 Cerebrovascular Disease 2. Physical Inactivity""'... Ischemic Heart Disease Cerebrovascular Disease Diabetes Mellitus 3. Tobacco Use 0 LungCancer 0 Chronic Obstructive Pulmonary Disease Ischemic Heart Disease The same basic concept for impact applies to the financial and economic losses caused b y a risk factor and its secondary diseases; the formulae are similar and will not be repeated here. The two main components of losses considered in the study are economic losses (lost productivity (the human capital approach) due to disability or death), and financial losses (health care costs). Productivity loss i s calculated as GDPkapita applied to the time lost due to disability or death. Disability i s calculated from the non-fatal complications considering those that are severe enough to prevent effective participation in the work force. The incidence and duration of these complications are taken from the Global Burden of Disease Study. For periods of productivity lost to both death and disability, a 3% annual future discount i s applied. The value of a continuous period with discounting i s calculated usingthe discrete formula for a continuous stream o f life: ""l The major part of impact due to Physical Inactivity is captured by the three selecteddiseases; however the full range of impact would ideally include certain neoplastic diseases and psychiatric conditions. 71 npresent =Presentvalue of a stream of life of n years duration value r =Futurediscount rate For each disease, health care costs are calculated beginning with the reimbursement amounts in the public sector. Next, other health care costs from the public sector are taken into account, including both ambulatory costs and resources transferred through other mechanisms, notably the PAB (Piso de Aten@o Bcisica) which i s the main source of resources for primary health care. Care is taken to separate expenditures on prevention and promotion from true curative health care. To this is added the private sector expenditure on health care. Very few details are available on how the private sector plans spend their resources; similarly, little data i s available on how public resources are spent in ambulatory and primary health care in general. Therefore, for each disease, the general assumption i s made that the proportion of overall resources used in public sector hospitalization for the disease can be extrapolated to the other categories o f expenditures. The sequence and approach o f calculations i s shown in great detail in the spreadsheet developed for this report. It is important to mention at this point that the discounting of health impact (DALYs) and of productivity loss i s carried out in two stages. The first stage involves the discounting of the period of time lost or affected by the disease back to the year o f incidence of the disease. In the second stage, the net value at the year o f incidence i s then discounted back to the year of reference (2004). It is also possible to separately discount each individual year of impact or productivity loss back to the year of reference; however, this procedure i s more difficult and would unnecessarily complicate the calculations. The approach for health care costs i s somewhat different. These are considered one year at a time and discounted to 2004 in one stage only. Once the health, financial, and economic impact o f risk factors are quantified, a few sample interventions are selected as illustrations. Their effectiveness in modifying the risk factors (be they primarily behavioral or physiological) is translated into DALYs averted and financial and economic losses (health care costs and productivity loss) prevented. Their cost i s contrasted to their effectiveness and benefits to derive their cost- effectiveness (in Brazilian Reals o f expenditure per DALY averted), and cost-benefit (measured in losses preventedper Brazilian Real of expenditures). 72 11. REFERENCES ~~~ Chackiel J. Elenvejecimiento de la poblaci6n latinoamericana:ihacia una relaci6n de dependencia favorable? ECLACKELADE -Santiago de Chile, Series Poblacidn y Desarrollo No. 4, 2000. Ministtrio de Sa6de. Inqutritodomiciliar sobre comportamentosde risco e morbidadereferida de doenqas e agravos n8o transmissiveis: Brazil, 15 capitais e distrito federal 2002-2003. Rio de Janeiro:INCA,2004. WHO. World Health Survey results in Brazil. Presented in 2004. WHO. The World Health Report - 2002: ReducingRisks, Promoting Healthy Life. Geneva, 2002. Monteiro C, CondeW, and Popkin B. I s obesity replacing or addingto undernutrition? Evidence from different socialclasses in Brazil. Public HealthNutrition.2002;5(1A):105-112. PuskaP. et.al. Changes in prematuredeaths in Finland: Successfullong-term, prevention of cardiovascular diseases. Bull WHO. 1998; 76:419-25. CELADE. Boletin Demogrifico, 69, January 2002. Zavala de Cossio MA. The Demographictransition in Latin America and Europe. In GuzmanJM et a1(eds). The fertility transition in LatinAmerica. Oxford, ClarendonPress, 1996, pp.94-109. Leeder et.al. A race against time: the challengeof cardiovasculardisease in developing economies.Report published in 2004 by the Trusteesof Columbia University in the City of New York. lo EscolaNacional de Saiide Pdblica (ENSP)/ FIOCRUZ / FENSPTEC. Projeto de Cargade Doenqa do Brasil. Rio de Janeiro, October 2002. l1 PereiraJ. Review and update of the burden of disease in Latin America and the Caribbean.Background paper for the World Bank. 2003. l2 Rose G. The strategy of preventive medicine. Oxford: Oxford University Press; 1992 13 Rose G. Sick individuals and sick populations. International Journalof Epidemiology 1985: 4; 32-38. 14 Ezzati M,Lopez AD, RodgersA, Vander Hoorn S, Murray cjl, and the Comparative Risk Assessment Collaborating Group. Selected major risk factors and global and regional burden of disease. Lancet 2002; 360: 1347-60. l5 da Costa LMand Goldfarb S. Government leadershipin tobacco control: Brazil's experience.IN: Tobacco control policy - strategies, successes and setbacks. Eds: deBeyer J. And Brigden LW. World bank and International Development Research Center.Washington DC 2003 l6 Westphal M.et.al. (2004) Informaqaesbisicas sobre as doenqas ngo transmissiveise fatores de risco para elaboraqgo de anilises econBmicas - incluindo prevalkncia, duraqgo das doenqas, mortalidade e outras informaqaesbisicas. World Bank backgroundpaper. l7 Piccini Rx,Victora CG. Hipertensgoarterial sistCmicaem Area urbanano sul do Brasil: prevalkncia e fatores de risco. Rev sa6de p6blica. 1994;28(4):261-67. l8 Fuchs FD.Hipertensgoarterial sistkmica. Epidemiologia e fatores de risco. Arq bras cardio. 1994;65(3):443- 44. l9 Cordeiro R, LimaFilhoEC, Almeida IM. Pressgoarterial entre trabalhadoresde curtume. Rev saiide piiblica. 1998;32(5):467-76. 2o Lessa I. adulto brasileiro e as doenqas epidemioldgicas da modernidade: epidemiologia das doenqas da 0 modernidadedas doenqas crbnicas ngo transmissiveis.SBo Paulo: HUCITEC; 1998. Achutti A. Achutti VR. Fatores de risco para aterosclerose. Elementospara descriqlo da situaqgo no Rio Grande do Sul. Arq bras cardiol. 1994;63(5):427-31. 22 Franco LJ. Epidemiologia do diabetes mellitus. In: Lessa I, organizadora.0 adulto brasileiro e as doenqas da modernidade.SBo Paulo-Rio de Janeiro: HUC1TEC;ABRASO; 1998. 23 Stephenet al. Environment and Health inDeveloping Countries: An Analysis of Intra-urban Differentials usingExistingData. London School of Hygiene and Tropical Medicine, 1994. 24 Charlotte Schoenborn et al. "Body Weight Status of Adults: United States, 1997-98", Advance Data from Vital and Health Statistics, No. 330, September 6, 2002, CDC. 25 MonteiroC, Conde W, and Popkin B. I s obesity replacing or adding to undernutrition? Evidence from different socialclasses in Brazil. Public HealthNutrition.2002;5( 1A):105-112. 73 26Monteiro C, Conde W, and Popkin B. Independent effects of income and education on the risk of obesity in the Brazilian adult population. Jour Nutr. 2001;131(3):S881-6. 27 Monteiro C et al. Obesity and inequities in health in the developing world. IntJour Obesity. 2004;l-6. 28 Monteiro C et.al. Presentation to the 32nd Sessionof the Permanent U.N. Committee on Nutrion. Data available at: www.ibge.gov.br 29 Monteiro, et al. A descriptive epidemiology of leisure-time physical activity in Brazil, 1996-97. Pan Am Jour Pub Hlth. 2003;14(4). 30 Barker DJP. Inutero programming of chronic disease. Clin Science. 1998;95:115-128. 3` Sichieri et al. Short stature and hypertension in the city of Rio de Janeiro, Brazil. AJPH. 2000;89(6):845-850. 32 Chronic Disease Epidemiology 33 Willett WC. Balancing life-style and genomics research for disease prevention. Science. 2002;296:695-8. 34 Hoffman K. and Jackson S. A review of the evidence for the effectiveness and costs of interventions preventing the burden of non-communicable diseases: How can health systems respond? Background paper for the World Bank. 2003. 35 DeBeyer J. and Brigden LW. Overview IN Tobacco control policy: Strategies, successesand setback. World Bank. 2003. 36 Da Costa L and Goldfarb S. Government leadership in tobacco control. In: Tobacco country profiles: Stratgegies, successes, and setbacks. ED: de Beyer J and Waverley L.2003. World Bank and RITC. 37 CDC. Recommendations to increase physical activity in communities. Am J Prev Med. 2002; 22(4 Suppl):67- 72. 38 Hernandez et al. Association of obesity with physical activity, television programs and other forms of video viewing among children in Mexico City. Int Jour Obes Re1Metab Dis. 1999;23:845-54. 39 Huet.al. Physical activity and television watching inrelation to riskfor type 2 diabetes mellitus in men. Arch IntMed. 2001; 161:1542-48. 40 Matsudo AM, Matsudo V, Andrade DR et.al. Physical activity promotion: experiences and evaluation o f the Agita S6o Paulo Program using the ecological mobile model. Journal of Physical Activity and Health. 2004; 1: 81-87. 41 Matsudo SM, Matsudo VR, Araujo TL, Andrade DR, ,Andrade EL, Oliveira LC, Braggion GF. The Agita S6o Paulo Program as a model for using physical activity to promote health. Pan Am J Public Health 2003; 14 (4): 265-272. 42 Matsudo V, Matsudo SM, Andrade D, Araujo T, Andrade E, Oliveira LC, Braggion G, Ausenka M.The Agitu S6o Paulo model in promoting physical activity. Chapter 183, 870-880. InMedeiros-Net0 G, Halpern A, Bouchard C (eds). Progress in Obesity Research:9, John Libbey Eorotext, 2003. 43 Pratt et.al. (2004) Economic analysis of Agita Sao Paulo. World Bank background paper. 44 Gortmaker et.al. Reducing obesity via a school-based interdisciplinary intervention among youth: Planet health. Arch Ped Adolesc Med. 1999;153:409-18. 45 Coutinho et.al. What Brazil is doing to promote healthy diets and active lifestyles. Pub HlthNut. 200231(1A):263-7. 46 CDC. Recommendations to increase physical activity in communities. Am J Prev Med 2002 May; 22 (4 Suppl): 67-72. 47 Willet et.al. (Forthcoming) Prevention of chronic diseaseby diet and lifestyle changes. In: Jamison DT et.al. Disease control priorities in developing countries, 2ndedition, 2006; Oxford University Press, for the World Bank. 48 Uusitalo et al. Fall in total cholesterol concentration over five years in association with changes in fatty acid composition of cooking oil in Mauritius: Cross sectional survey. BMJ 1996; 313:1044-1046. 49 FarquharJW et.al. The community-basedmodel of lifestyle intervention trials. AmJourEpid. 1978;108:103- 111. 50 Puska P. et.al. The North Karelia Project: 20 year results and experiences. National Public Health Institute, Helsinki 1995. 51 FarquharJW. The Stanford Five-City projectq Effects of community wide education on cardiovascular diseaserisk factors. JAMA. 1978;264:359-365. 74 52 Luepker RV. et.al. Community education for cardiovascular disease prevention :risk factor changes in the Minnesota Heart Health Program. AJPH. 1994;Sep 84(9) :1383-93. 53 Carleton RA. et.al. The Pawtucket Heart Health Program: Community changes in cardiovascular risk factors and projected diseaserisk. Amer J Public Health. 1995;Jun 85(6) :777-85. 54 Harvey D. et.al. Building the case for the urevention of chronic disease.2002. Ottawa,ON:Health Canada. 55 WHO. Diet, nutrition and the prevention of chronic diseases:a report of ajoint WHOFAO expert. 2003. j6 Tuomilehto J. et.al. Prevention of type 2 diabetes mellitusby changes in lifestyle among subjects with impaired glucose tolerance. NEJM. 2001;344: 1343-1350. 57 Eriksson KF and Lindgarde F. Prevention of type 2 (non-insulin-dependent) diabetes mellitusby diet and physical exercise. The six year Malmo feasibility study. Diabetologia. 1991;34(12):891-898. 58 Pan XR et.al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care. 1997;20(4):537-544. j9 Piccini RX and Victora CG. How well i s hypertension managed in the community? A population-based survey in a Brazilian city. Cad saude publica. Out.-dez. 1997;13(4):595-600. 6o Castro VD de, Car MR. 0 cotidiano da vida de hipertensos? Mudanqas, restriq6es e reaq6es. Rev Exc Enf USP. 2000~34(2):145-53. 61 Car MR, Pierin AMG, Aquino VLA. Estudo sobre a influencia do processo educativo no controle da hipertensso arterial. Rev Esc EnfUSP. 1992;25(3):259-69. Souza ALL. et.al. Uma experisncia multiprofissional na abordagem ao paciente hipertenso. Arq Bras Cardiol 1992;59( 1):3 1-35. 63 Santos RD, Nascimento LO, MaranhBo RC. AvaliaqBo dos efeitos a curto prazo de programas de conscientizaqgo e de terapia com pravastatina em individuos de alto risco para doenqa cardiovascular provenientes de Consult6rios particulares. Arq Bras Cardiol. 1997;69(4):225-30. 64 Rassi 1998 (cited by Westphal, Estudo 1) 65 Formigli VLA. et.al. HipertenqBo arterial em adultos de um bairro de Salvador, Bahia. Rev Baiana de saude I "dbl. Dez. 1998 -jan. 1999. 1999;23:7-20. Freitas 0 de C. Et.al. Prevalelencia da hipertenqBo arterial sistemica na populaqgo urbana de Catanduva-SP. Arq Bras Cariol 2001;77(1):09-15. 67 Chaves EC. A palestra educativa na orientaqgo sobre o autocuidado do hipertenso. RevEsc EnfUSP. 1990;24(3):397-402. Chor D. Hipertensgo arterial entre funcionfirios de banco estatal no Rio de Janeiro. HBbitos de vida e tratamento. Arq Bras Cardiol. 1998;71(5):653-660. 69 Sala A. et.al. AvaliaqBo do processo de atendimento a pacientes portadores de doenqacr6nica-degenerativa em uma unidade bBsicade sadde. Rev Sadde Pdbl 1993;27(6):463-71. 70Westphal M.F.et.al. (2004) Evidhcias de efetividade e custos das intervenq6es de prevenqgo das doenqas n5o transmissiveis no Brasil. Background paper for World Bank. 71 Narayan KMV.et.al. (Forthcoming) Diabetes: The pandemic and potencial solutions. In: Jamison DT et.al. Disease control priorities in developing countries, 2"dedition, 2006; Oxford University Press, for the World Bank. 72 Motta DGda. A educaqBo participante no controle metab6lico e qualidade de vida de mulheres com diabetes mellitus tipo 2. Sgo Paulo, 1998 [Tese de Doutorado - Faculdade de Sadde Pdblica de USP]. 73 Source: httu:l/www.cdc.govlnccdphplpe factsheetslpe cvh.htm 74WHO Innovative care for chronic conditions: building blocks for action: Global Report. WHO/MNC/CCW02.01,2002 75 Epping-JordanJE et.al. Improving the quality of health care for chronic conditions. Qual Saf Health Care 2004; 13~299-305. 76WHO Adherence to long-term therapies: Policy for action, Meeting report, June 4-5,2001. WHO/MNC/CCWO1.02. 77 Salazar L,Velez JA, Ortiz Y. En Buscade evidencias de efectividad enpromocion de la salud en America Latina: Estado del arte de experiencias regionales. Paper commissioned by the IUHPE.,2003. 78 De Beyer J. and Brigden LW. Overview IN Tobacco control policy: Strategies, successes and setback. World Bank. 2003. 75 79Murray, C. and A. Lopez (1996). The globalburdenof disease: a comprehensive assessment of mortality and disability from diseases, iniuries, and risk factors in 1990and projected to 2020. Cambridge. EscolaNacionalde Salide Pfiblica(ENSP)/ FIOCRUZ/ FENSPTEC. Projeto de Carga de Doenqa do Brasil. Rio de Janeiro, October 2002. World Bank (2004). The World DevelopmentIndicators. Akhavan, D.(2004).Economic evaluation of non-communicabledisease control in Brazil.Washington, DC, Backgroundpaper for World Bank. 83Matsudo, S., V. Matsudo, et al. (2003). "TheAgita SBo PauloProgramas a modelfor usingphysicalactivity to promote health." PanAmericanJournal of Public Health14(4): 265-272. 84Management Sciences for Health. InternationalDrugPrice IndicatorGuide; 2001. 85FundaGBoGetulioVargas and EPOSHealthConsultants(2002). DeterminaGBoe avaliaqgodo custo do programade salideda famflia. 86 *'Agere CooperaqBo em Advocacy, Dr. Iradj RobertoEghrari,June 17, 2004. electronic communication. Comas FuentesA, Suirez Gutihez R, Ldpez GonzilezML,CuetoEspinar a Coste-efectividaddel consejo sanitario antitabaco realizado em AtencidnPrimaria.Gaceta Sanitaria 1998; 12:126-132. Commissionon Macroeconomics and Health(2001). Macroeconomicsand health: Investingin health for economic development. Geneva, WHO. 76