Prevalence of high risk for cardiovascular disease among the Brazilian adult population, according to different risk calculators: a comparative study

Deborah Carvalho Malta Pedro Cisalpino Pinheiro Renato Teixeira Azeredo Filipe Malta Santos Antonio Luiz Pinho Ribeiro Luisa Campos Caldeira Brant About the authors

Resumo

O estudo visa comparar a proporção de indivíduos classificados como portadores de alto risco cardiovascular (RCV) na população adulta brasileira, segundo seis diferentes calculadoras de risco, visando analisar a concordância entre as medidas. Estudo transversal, no qual foram utilizados dados laboratoriais da Pesquisa Nacional de Saúde (PNS). As prevalências do RCV em 10 anos para a população entre 45 e 64 anos foram: Escore de risco global (ERG) da Sociedade Geral de Cardiologia (SBC):38,1%, “American College of Cardiology” e “American Heart Association” ACC/AHA, 44,1%, “Framingham Heart Study”/ERG 19,4%, SCORE da “European Society of Cardiology”, 14,6, Organização Mundial da Saúde/Sociedade Internacional de hipertensão (OMS/ISH) e Lim et al. As calculadoras de RCV apresentaram baixa concordância para identificar os indivíduos de alto risco e alta concordância dos de risco baixo/moderado, exceto pela ACC/AHA. O emprego de diferentes calculadoras resultou em diferentes populações elegíveis para iniciar a terapia farmacológica para prevenção cardiovascular, o que pode implicar em percepções de risco inadequadas, baixo custo efetividade desse tratamento e dificuldade de implementação de políticas públicas.

Palavras-chave
Doenças cardiovasculares; Fatores de risco; Prevalência; Brasil

Abstract

This study compares the proportion of the Brazilian adult population classified as being at high risk of cardiovascular disease (CVD) based on six different CVD risk calculators in order to assess the agreement across different tools. A cross-sectional study was conducted using laboratory data from the National Health Survey (NHS). The prevalence rates of high 10-year risk of CVD among individuals aged between 45 and 64 years were as follows: Brazilian Society of Cardiology (BSC) global risk score (GRS) – 38.1%; American College of Cardiology/American Heart Association (ACC/AHA) score – 44.1%; Framingham Heart Study/GRS – 19.4%; European Society of Cardiology SCORE – 14.6%; World Health Organization/International Society of Hypertension (WHO/ISH) score – 3.1%; and Lim et al. – 2.5%. The CVD calculators showed poor agreement for the identification of high-risk individuals and a high level of agreement for the identification of low/moderate risk individuals, except for the ACC/AHA risk score. The findings show that the proportion of individuals classified as eligible for preventive drug therapy varies from tool to tool, which could lead to the misinterpretation of risk, poor cost-effectiveness of therapy and difficulty implementing public policies.

Key words
Cardiovascular diseases; Risk factors; Prevalence; Brazil

Introduction

Cardiovascular diseases (CVD) were responsible for approximately 18 million deaths in 2016, with around 80% of CVD deaths occurring in middle and low-income countries11 World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Genebra: WHO; 2018.. These diseases are associated with poor socioeconomic conditions (such as poverty and low income and education levels22 Souza MFM, Malta DC, França EB, Barreto ML. Transição da saúde e da doença no Brasil e nas Unidades Federadas durante os 30 anos do Sistema Único de Saúde. Cien Saude Colet. 2018; 23(6):1737-1750.,33 Harper S, Lynch J, Smith GD. Social determinants and the decline of cardiovascular diseases: understanding the links. Annu Rev Public Health. 2011; 32:39-69.), rapid urbanization, increased life expectancy11 World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Genebra: WHO; 2018.

2 Souza MFM, Malta DC, França EB, Barreto ML. Transição da saúde e da doença no Brasil e nas Unidades Federadas durante os 30 anos do Sistema Único de Saúde. Cien Saude Colet. 2018; 23(6):1737-1750.
-33 Harper S, Lynch J, Smith GD. Social determinants and the decline of cardiovascular diseases: understanding the links. Annu Rev Public Health. 2011; 32:39-69., behavioral risk factors (smoking, drinking, poor diet, sedentarism) and metabolic risk factors (obesity, high blood sugar, high blood pressure, hyperlipidemia)11 World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Genebra: WHO; 2018.

2 Souza MFM, Malta DC, França EB, Barreto ML. Transição da saúde e da doença no Brasil e nas Unidades Federadas durante os 30 anos do Sistema Único de Saúde. Cien Saude Colet. 2018; 23(6):1737-1750.

3 Harper S, Lynch J, Smith GD. Social determinants and the decline of cardiovascular diseases: understanding the links. Annu Rev Public Health. 2011; 32:39-69.
-44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016.. It is known that combined overlapping risk factors (RFs) result in an increased risk of CVD. The early detection of individuals at high risk of CVD and timely treatment is therefore a priority11 World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Genebra: WHO; 2018.,44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016..

The World Health Organization (WHO) recommends a number of population-wide policy interventions to encourage the primary prevention of CVD, including regulatory measures such as taxing tobacco, alcohol and ultra-processed foods55 World Health Organization (WHO). "Best Buys" Tackling NCDs: Best buys and other recommended interventions for the prevention and control of noncommunicable diseases. Genebra: WHO; 2017.,66 World Health Organization (WHO). Global action plan for the prevention and control of noncommunicable diseases 2013-2020. Genebra: WHO; 2013. and the creation of environments that facilitate healthy lifestyles and empower individuals and communities to make healthy choices66 World Health Organization (WHO). Global action plan for the prevention and control of noncommunicable diseases 2013-2020. Genebra: WHO; 2013.. Also within the context of primary prevention – defined in this case as prevention prior to a cardiovascular event – the WHO recommends the identification of high-risk individuals using risk scores or calculators to estimate the combined risk of CVD44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016..

The identification of high-risk individuals permits the adoption of specific preventive measures (counseling and drug therapy), including the prescription of statins44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016. or drug therapy in the prehypertension stage77 Whelton PK, Carey RM, Aronow WS, Casey Junior DE, Collins KJ, Himmelfarb CD, DePalma SM, Gidding S, Jamerson KA, Jones DW, MacLaughlin EJ, Muntner P, Ovbiagele B, Smith Junior SC, Spencer CC, Stafford RS, Taler ST, Thomas RJ, Williams KA, Williamson JD, Wright Junior JT. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension. 2018; 71(6):1269-1324.,88 Malachias MVB, Souza WKSB, Plavnik FL, Rodrigues CIS, Brandão AA, Neves MFT, Bortolotto LA, Franco RJS, Poli-de-Figueiredo CE; Sociedade Brasileira de Cardiologia. 7ª Diretriz Brasileira de Hipertensão Arterial. Arquivos Brasileiros de Cardiologia. 2016; 107(Supl. 3):1-82.. Both these interventions are aimed at preventing death and non-fatal adverse cardiovascular events, particularly coronary artery disease (CAD) and strokes, the two leading causes of death in Brazil11 World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Genebra: WHO; 2018.,99 Taylor F, Huffman MD, Macedo AF, Moore THM, Burke M, Smith GD, Ward K, Ebrahim S. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2013; (1):CD004816.. CVD risk calculators have therefore become important tools for supporting public health actions, particularly in primary health care services, and informing decision-making about counseling and treatment44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016.. However, selecting which calculator to use in Brazil remains a topic of debate, as an equation derived from national cohort studies representing the country’s specific population characteristics (racial composition, socioeconomic and geographic conditions, specific laboratory reference values, etc.) does not yet exist, meaning that risk estimations can often be inaccurate.

Current CVD risk calculators differ according to the characteristics of the population from which they were derived (sex, age group, race, etc.) and the presence or absence of population-specific risk prevention measures, which vary over time and depending on local health policies1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80.. Although calculators tend to include similar RFs, the CVD risk weightings assigned to individual factors and the 10-year CVD outcomes and how they are adjudicated can vary from tool to tool1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80.. For example, some calculators only estimate the risk of CVD death, while others include various non-fatal cardiovascular events. In addition, each calculator adopts its own threshold for high risk of CVD, taking into account the above characteristics and the risk authors consider acceptable for the indication of statins based on the medicine’s country-specific benefit-risk ratio1111 WHO CVD Risk Chart Working Group. World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health. 2019; 7(10):E1332-E1345..

The objective of this study was therefore to compare the proportion of the Brazilian population at high risk of CVD estimated using different risk calculators in order to assess the agreement across different tools. The implications of the findings for preventive interventions in high-risk individuals and policy planning in Brazil are then discussed.

Methods

We conducted a cross-sectional study using data from the National Health Survey (NHS), a nationwide study undertaken by the Brazilian Institute of Geography and Statistics (IBGE, acronym in Portuguese) and Ministry of Health in 2013 and completed with a laboratory subsample in 2014 and 20151212 Szwarcwald CL, Malta DC, Souza Júnior PRB, Almeida WS, Damacena GN, Pereira CA, Rosenfeld LG. Exames laboratoriais da Pesquisa Nacional de Saúde: metodologia de amostragem, coleta e análise dos dados. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190004..

We used a subsample consisting of 25% of the census tracts selected for the 2013 NHS using the same stratified sampling design as the survey, applying probability inversely proportional to the difficulty of data collection1111 WHO CVD Risk Chart Working Group. World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health. 2019; 7(10):E1332-E1345.. The following three-stage cluster sampling design was used: Stage 1 – selection of primary sampling units (census tracts or composition of tracts); Stage 2 – random selection of a fixed number (10 to 14) of permanent private households from each census tract; Stage 3 – random selection of one person aged 18 years and over living in each household from a list of eligible participants drawn up at the time of the interview1111 WHO CVD Risk Chart Working Group. World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health. 2019; 7(10):E1332-E1345.. Based on the percentage of the NHS census tracts used to undertake the laboratory tests and a non-response rate of 20%, the expected number of individuals with laboratory data was approximately 12,0001111 WHO CVD Risk Chart Working Group. World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health. 2019; 7(10):E1332-E1345.. Biochemical tests were performed on 8,952 individuals in Brazil as a whole. Further details about the selection process1212 Szwarcwald CL, Malta DC, Souza Júnior PRB, Almeida WS, Damacena GN, Pereira CA, Rosenfeld LG. Exames laboratoriais da Pesquisa Nacional de Saúde: metodologia de amostragem, coleta e análise dos dados. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190004.,1313 Malta DC, Duncan BB, Schmidt MI, Machado ÍE, Silva AG, Bernal RTI, Pereira CA, Damacena GN, Stopa SR, Rosenfeld LG, Szwarcwald CL. Prevalência de diabetes mellitus determinada pela hemoglobina glicada na população adulta brasileira, Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl 2):E190006., aspects related to specimen collection1212 Szwarcwald CL, Malta DC, Souza Júnior PRB, Almeida WS, Damacena GN, Pereira CA, Rosenfeld LG. Exames laboratoriais da Pesquisa Nacional de Saúde: metodologia de amostragem, coleta e análise dos dados. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190004.,1313 Malta DC, Duncan BB, Schmidt MI, Machado ÍE, Silva AG, Bernal RTI, Pereira CA, Damacena GN, Stopa SR, Rosenfeld LG, Szwarcwald CL. Prevalência de diabetes mellitus determinada pela hemoglobina glicada na população adulta brasileira, Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl 2):E190006., blood pressure measurement1616 Malta DC, Santos NB, Perillo RD, Szwarcwald CL. Prevalence of high blood pressure measured in the Brazilian population, National Health Survey, 2013. São Paulo Med J. 2016; 134(2):163-170., assessment of smoking and prior CVD1717 Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa Nacional de Saúde 2013: percepção do estado de saúde, estilos de vida e doenças crônicas: Brasil, Grandes Regiões e Unidades da Federação. Rio de Janeiro: IBGE; 2014.,1818 Gonçalves RPF, Haikal DS, Freitas MIF, Machado IE, Malta DC. Diagnóstico médico autorreferido de doença cardíaca e fatores de risco associados: Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190016., and thresholds for estimating the proportion of the population with diabetes1313 Malta DC, Duncan BB, Schmidt MI, Machado ÍE, Silva AG, Bernal RTI, Pereira CA, Damacena GN, Stopa SR, Rosenfeld LG, Szwarcwald CL. Prevalência de diabetes mellitus determinada pela hemoglobina glicada na população adulta brasileira, Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl 2):E190006. and above-normal cholesterol levels1414 Malta DC, Szwarcwald CL, Machado IE, Pereira CA, Figueiredo AW, Sá ACMGN, Velasquez-Melendez G, Santos FM, Souza-Júnior PRB, Stopa SR, Rosenfeld LG. Prevalência de colesterol total e frações alterados na população adulta brasileira: Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl. 2):1-13.,1515 American medical association. Executive summary of the Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001; 285(19):2486-2497. can be found in previous studies.

Individuals were classified as being at high or low/moderate risk of CVD (hereafter called low risk) using the following six calculators/scores: 1) The calculator recommended by the Brazilian Society of Cardiology (BSC), based on the Framingham Heart Study-derived calculator plus other criteria detailed below; 2) The pooled cohort equation, which was introduced by the American College of Cardiology and American Heart Association (ACC/AHA) in 20132020 Goff Junior DC, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, Greenland P, Lackland DT, Levy D, O’Donnell CJ, Robinson JG, Schwartz JS, Shero ST, Smith Junior SC, Sorlie P, Stone NJ, Wilson PWF. American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63(25):2935-2959. and uses data from various cohort studies in the United States to derive and validate new sex and age-specific equations; 3) The global risk score derived from the 2008 Framingham Heart Study (GRS-FHS)2121 D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008; 117(6):743-753.; 4) The SCORE calculator, proposed by the European Society of Cardiology. Derived from various European cohort studies, SCORE proposes two estimation equations for coronary heart disease and non-coronary cardiovascular disease calculated for high-risk (Eastern) and low-risk (Western) regions of Europe. For the purposes of the present study, we cautiously chose the equation for high-risk regions, because we did not know the risk of the Brazilian population2222 Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, De Backer G, De Bacquer D, Ducimetière P, Jousilahti P, Keil U, Njølstad I, Oganov R G, Thomsen T, Tunstall-Pedoe H, Tverdal A, Wedel H, Whincup P, Wilhelmsen L, Graham IM, SCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003; 24(11):987-1003.; 5) Country-specific risk charts developed by Lim et al.2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062. using simulation models2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062.; 6) The calculator proposed by the WHO/International Society of Hypertension (WHO/ISH) for 14 WHO epidemiological sub-regions. For the purposes of this study, we selected the Americas sub-region B, which includes Brazil2424 World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions. Genebra: WHO; 2007. (Table 1). These calculators were chosen because they are the most commonly used tools in clinical practice or are specific to the region in which Brazil is located.

Table 1
Characteristics of the cardiovascular disease risk calculators assessed by the study -age group, variables used, 10-year CVD outcomes and high-risk thresholds.

It is important to note that age groups, variables, CVD outcomes, and thresholds for high-risk of CVD differ from calculator to calculator, as shown in Table 1.

With regard to the calculator recommended by the BSC1919 Précoma DB, Oliveira GMM, Simão AS, Dutra OP, Coelho OR, Izar MCO, Póvoa RMS, Giuliano ICB, Alencar Filho AC, Machado CV, Scherr C, Fonseca FAH, Santos Filho RD, Carvalho T, Avezum Jr. A, Esporcatte R, Nascimento BR, Brasil DP, Soares GP, Villela PB, Ferreira RM, Martins WA, Sposito AC, Halpern B, Saraiva JFK, Carvalho LSF, Tambascia MA, Coelho-Filho OR, Bertolami A, Correa Filho H, Xavier HT, Faria-Neto JR, Bertolami MC, Giraldez VZR, Brandão AA, Feitosa ADM, Amodeo C, Souza DSM, Barbosa ECD, Malachias MVB, Souza WKSB, Costa FAA, Rivera IR, Pellanda LC, Silva MAM, Achutti AC, Langowiski AR, Lantieri CJB, Scholz JR, Ismael SMC, Ayoub JCA, Scala LCN, Neves MF, Jardim PCBV, Fuchs SCPC, Jardim TSV, Moriguchi EH, Schneider JC, Assad MHA, Kaiser SE, Lottenberg AM, Magnoni CD, Miname MH, Lara RS, Herdy AH, Araújo CGS, Milani M, Silva MMF, Stein R, Lucchese FA, Nobre F, Griz HB, Magalhães LBNC, Borba MHE, Pontes MRN, Mourilhe-Rocha R. Atualização da Diretriz de Prevenção Cardiovascular da Sociedade Brasileira de Cardiologia - 2019. Arq Bras Cardiol. 2019; 113(4):787-891., individuals are classified based on an global risk score (GRS)2121 D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008; 117(6):743-753. (high-risk threshold of > 20% for men and > 10% for women) or the presence of other variables: individuals with chronic kidney disease (glomerular filtration rate of <60ml/min and not on dialysis); individuals with LDL cholesterol of ≥ 190mg/dL; individuals with LDL cholesterol of ≥ 70 to < 190mg/dL, together with at least one other risk factor (men aged ≥ 48 years and women aged ≥ 54 years, time since diagnosis of diabetes ≥ 10 years, smoker, high blood pressure – systolic or diastolic pressure ≥ 140 and ≥ 90mmHg, respectively); and individuals with metabolic syndrome based on International Diabetes Federation criteria2828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476. (triglycerides were replaced by total cholesterol due to the lack data on the former)1212 Szwarcwald CL, Malta DC, Souza Júnior PRB, Almeida WS, Damacena GN, Pereira CA, Rosenfeld LG. Exames laboratoriais da Pesquisa Nacional de Saúde: metodologia de amostragem, coleta e análise dos dados. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190004.,1414 Malta DC, Szwarcwald CL, Machado IE, Pereira CA, Figueiredo AW, Sá ACMGN, Velasquez-Melendez G, Santos FM, Souza-Júnior PRB, Stopa SR, Rosenfeld LG. Prevalência de colesterol total e frações alterados na população adulta brasileira: Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl. 2):1-13.. The BSC definition of high risk of CVD also includes individuals with subclinical atherosclerosis, abdominal aortic aneurysm, a family history (first degree relative) of early-onset CVD, and albuminuria1919 Précoma DB, Oliveira GMM, Simão AS, Dutra OP, Coelho OR, Izar MCO, Póvoa RMS, Giuliano ICB, Alencar Filho AC, Machado CV, Scherr C, Fonseca FAH, Santos Filho RD, Carvalho T, Avezum Jr. A, Esporcatte R, Nascimento BR, Brasil DP, Soares GP, Villela PB, Ferreira RM, Martins WA, Sposito AC, Halpern B, Saraiva JFK, Carvalho LSF, Tambascia MA, Coelho-Filho OR, Bertolami A, Correa Filho H, Xavier HT, Faria-Neto JR, Bertolami MC, Giraldez VZR, Brandão AA, Feitosa ADM, Amodeo C, Souza DSM, Barbosa ECD, Malachias MVB, Souza WKSB, Costa FAA, Rivera IR, Pellanda LC, Silva MAM, Achutti AC, Langowiski AR, Lantieri CJB, Scholz JR, Ismael SMC, Ayoub JCA, Scala LCN, Neves MF, Jardim PCBV, Fuchs SCPC, Jardim TSV, Moriguchi EH, Schneider JC, Assad MHA, Kaiser SE, Lottenberg AM, Magnoni CD, Miname MH, Lara RS, Herdy AH, Araújo CGS, Milani M, Silva MMF, Stein R, Lucchese FA, Nobre F, Griz HB, Magalhães LBNC, Borba MHE, Pontes MRN, Mourilhe-Rocha R. Atualização da Diretriz de Prevenção Cardiovascular da Sociedade Brasileira de Cardiologia - 2019. Arq Bras Cardiol. 2019; 113(4):787-891.. These problems were not included in the analysis due to lack of information on the conditions in the PNS database.

With regard to statistical analysis, participants with known CVD were excluded from the analysis because the calculators used in this study also exclude these cases. To permit comparisons across the different tools, an additional analysis was performed of participants aged between 45 and 65 years, the age group covered by all the calculators (n = 2,791). The risk estimation tools with continuous results (SCORE, ACC/AHA) or scores (GRS) were analyzed using a binary classification, where individuals above and below the threshold were assigned a value of 1 for high risk or 0 for low risk, respectively. The tools developed by Lim et al.2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062., the WHO/ISH2424 World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions. Genebra: WHO; 2007. and BSC1919 Précoma DB, Oliveira GMM, Simão AS, Dutra OP, Coelho OR, Izar MCO, Póvoa RMS, Giuliano ICB, Alencar Filho AC, Machado CV, Scherr C, Fonseca FAH, Santos Filho RD, Carvalho T, Avezum Jr. A, Esporcatte R, Nascimento BR, Brasil DP, Soares GP, Villela PB, Ferreira RM, Martins WA, Sposito AC, Halpern B, Saraiva JFK, Carvalho LSF, Tambascia MA, Coelho-Filho OR, Bertolami A, Correa Filho H, Xavier HT, Faria-Neto JR, Bertolami MC, Giraldez VZR, Brandão AA, Feitosa ADM, Amodeo C, Souza DSM, Barbosa ECD, Malachias MVB, Souza WKSB, Costa FAA, Rivera IR, Pellanda LC, Silva MAM, Achutti AC, Langowiski AR, Lantieri CJB, Scholz JR, Ismael SMC, Ayoub JCA, Scala LCN, Neves MF, Jardim PCBV, Fuchs SCPC, Jardim TSV, Moriguchi EH, Schneider JC, Assad MHA, Kaiser SE, Lottenberg AM, Magnoni CD, Miname MH, Lara RS, Herdy AH, Araújo CGS, Milani M, Silva MMF, Stein R, Lucchese FA, Nobre F, Griz HB, Magalhães LBNC, Borba MHE, Pontes MRN, Mourilhe-Rocha R. Atualização da Diretriz de Prevenção Cardiovascular da Sociedade Brasileira de Cardiologia - 2019. Arq Bras Cardiol. 2019; 113(4):787-891. calculate risks dichotomously. Agreement across the calculators was assessed by comparing the CVD risk classifications (prevalence of high and low risk) using the percent of pairwise agreement and the BSC calculator1919 Précoma DB, Oliveira GMM, Simão AS, Dutra OP, Coelho OR, Izar MCO, Póvoa RMS, Giuliano ICB, Alencar Filho AC, Machado CV, Scherr C, Fonseca FAH, Santos Filho RD, Carvalho T, Avezum Jr. A, Esporcatte R, Nascimento BR, Brasil DP, Soares GP, Villela PB, Ferreira RM, Martins WA, Sposito AC, Halpern B, Saraiva JFK, Carvalho LSF, Tambascia MA, Coelho-Filho OR, Bertolami A, Correa Filho H, Xavier HT, Faria-Neto JR, Bertolami MC, Giraldez VZR, Brandão AA, Feitosa ADM, Amodeo C, Souza DSM, Barbosa ECD, Malachias MVB, Souza WKSB, Costa FAA, Rivera IR, Pellanda LC, Silva MAM, Achutti AC, Langowiski AR, Lantieri CJB, Scholz JR, Ismael SMC, Ayoub JCA, Scala LCN, Neves MF, Jardim PCBV, Fuchs SCPC, Jardim TSV, Moriguchi EH, Schneider JC, Assad MHA, Kaiser SE, Lottenberg AM, Magnoni CD, Miname MH, Lara RS, Herdy AH, Araújo CGS, Milani M, Silva MMF, Stein R, Lucchese FA, Nobre F, Griz HB, Magalhães LBNC, Borba MHE, Pontes MRN, Mourilhe-Rocha R. Atualização da Diretriz de Prevenção Cardiovascular da Sociedade Brasileira de Cardiologia - 2019. Arq Bras Cardiol. 2019; 113(4):787-891. as a reference. The BSC calculator was used as a reference because it is the tool recommended by national guidelines. The percent agreement measures the proportion of individuals at high or low risk of CVD based on the calculator in question and on the BSC calculator.

The NHS was approved by the National Research Ethics Committee1717 Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa Nacional de Saúde 2013: percepção do estado de saúde, estilos de vida e doenças crônicas: Brasil, Grandes Regiões e Unidades da Federação. Rio de Janeiro: IBGE; 2014..

Results

Figure 1 shows the prevalence of high 10-year risk of CVD for the population aged between 45 and 64 years using the threshold suggested by each calculator. The calculator that showed the highest prevalence rate was the ACC/AHA risk score (44.1%; 95%CI, 41.7-46,5), followed by BSC (38.1%; 95%CI, 35.8-40.4), GRS (19.4%; 95%CI, 17.5-21.4), SCORE (14.6%; 95%CI, 12.9-16.4), WHO/ISH2424 World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions. Genebra: WHO; 2007. (3.1%; 95%CI, 2.4-4), and Lim et al.2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062. (2.5%; 95%CI, 1.8-3.3).

Figure 1
Proportion of the population aged between 45 and 64 years at high-risk of cardiovascular disease (CVD) based on the different CVD risk calculators, Brazil. National Health Survey - 2013, 2014-2015.

Figure 2 shows the prevalence of high risk of CVD using the same thresholds, but this time with the different age groups covered by each measure. The calculator that showed the highest prevalence rate was once again the ACC/AHA risk score (40-79 years – 39.4%; 95%CI, 37.6-41.3), followed by BSC (30-74 years – 28.8%; 95%CI, 27.4-30.2), GRS (30-74 years – 14.7%; 95%CI 13.6-15.9) and SCORE (45-64 years – 14.6%, 95%CI 12.9-16.4); with Lim et al. and WHO/ISH once again showing the lowest prevalence rates.

Figure 2
Proportion of the population at high-risk of cardiovascular disease (CVD) by the age groups and thresholds adopted by the different CVD risk calculators, Brazil. National Health Survey - 2013, 2014-2015.

Table 2 shows the prevalence of high risk of CVD together with the percent agreement between each tool and the BSC calculator. The calculator that showed the highest level of agreement with the BSC calculator for prevalence of high-risk of CVD was GRS (50.9%; 95%CI, 47.1-54.7), which is to be expected given that GRS is part of the BSC calculator’s estimation equation. The findings also show that 43.5% (95%CI, 39.7-47.3) of the high-risk individuals predicted by the BSC calculator were considered high-risk by the ACC/AHA risk score, compared to 29.4% (95%CI 26-33.1) for SCORE and less than 10% for WHO2424 World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions. Genebra: WHO; 2007. and Lim et al.2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062..

Table 2
Percent agreement across the six CVD risk calculators for the classification of individuals as high or low-risk using the calculator recommended by the Brazilian Cardiology Society as a reference, Brazil. National Health Survey 2014-2015.

With regard to percent agreement for the prevalence of low risk, the calculator that showed the lowest level of agreement was the ACC/AHA risk score. The findings show that 55.5% the low-risk individuals predicted by the BSC calculator were considered low-risk by the ACC/AHA risk score. GRS showed 100% agreement with the BSC calculator, which is to be expected considering that the latter uses the GRS, albeit with lower thresholds and including other categories in the definition of high-risk. The percent agreement between SCORE and the BSC calculator was 94.6% (95%CI, 92.8-95.9), while the level of agreement between WHO and Lim et al.2323 Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet. 2007; 370(9604):2054-2062. and the BSC calculator was over 99%.

Discussion

In general, the CVD risk calculators assessed by this study showed a low level of agreement with the BSC calculator for detecting high risk of CVD and a high level of agreement for identifying individuals at low risk of CVD, except the ACC/AHA risk score. The findings also show that the proportion of individuals classified as being at high risk of CVD by these commonly used calculators varied considerably, reaching up to 39% of the population aged between 45 and 65 years. This means that the proportion of individuals eligible for preventive drug therapy varies from tool to tool, which could lead to the misinterpretation of risk, poor cost-effectiveness of therapy and difficulty implementing public policies.

The low level of agreement for the identification of high-risk individuals found by the present study has been reported by previous studies. Using data from hypothetical patients and 24 calculators, Allan et al.2626 DeFilippis AR, Young R, Carrubba CJ, McEvoy JW, Budoff MJ, Blumenthal RS, Kronmal RA, McClelland RS, Nasir K, Blaha MJ. An analysis of calibration and discrimination among multiple cardiovascular risk scores in a modern multiethnic cohort. Ann Intern Med. 2015; 162(4):266-275. found poor agreement between pairs of tools (67%), highlighting the need to calibrate calculators to specific populations to ensure the effective implementation of clinical guidelines on preventive drug therapy2424 World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions. Genebra: WHO; 2007.. Risk scores have been developed or calibrated mainly for populations in the United States and Europe, with a lack of studies in low and middle-income countries, where socioeconomic factors such as access to health care and racial and cultural characteristics have a particularly strong impact on risk of CVD1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80..

Other studies draw attention to the overestimation of risk, especially by the ACC/AHA risk score. In a prospective study with 4,000 male patients in the United States, DeFilippis et al.2626 DeFilippis AR, Young R, Carrubba CJ, McEvoy JW, Budoff MJ, Blumenthal RS, Kronmal RA, McClelland RS, Nasir K, Blaha MJ. An analysis of calibration and discrimination among multiple cardiovascular risk scores in a modern multiethnic cohort. Ann Intern Med. 2015; 162(4):266-275. found that discordance between events predicted using the Framingham Risk Score2121 D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008; 117(6):743-753. and ACC/AHA risk score2020 Goff Junior DC, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, Greenland P, Lackland DT, Levy D, O’Donnell CJ, Robinson JG, Schwartz JS, Shero ST, Smith Junior SC, Sorlie P, Stone NJ, Wilson PWF. American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63(25):2935-2959. and observed events ranged from 37 to 154%. The authors suggest that these differences may be due mainly to: the use of old cohorts to derive the risk scores – with probable changes in population characteristics over time, improvements in therapy and the identification of new risk factors in recent years; difficulties in assessing certain risk factors, such as the number of cigarettes smoked and alcohol intake; or study limitations affecting the identification of events. In a study using a 1997-2001 to 2012 cohort of patients aged 55 years and older without previous CVD living in Rotterdam, Kavousi et al.2727 Kavousi M, Leening MJG, Nanchen D, Greenland P, Graham IM, Steyerberg EW, Ikram MA Stricker BH, Hofman A, Franco OH. Comparison of Application of the ACC/AHA Guidelines, Adult Treatment Panel III Guidelines, and European Society of Cardiology Guidelines for Cardiovascular Disease Prevention in a European Cohort. JAMA 2014; 311(14):1416-1423. observed an overestimation of risk, drawing attention to the fact that almost all men and 65% of women would be eligible for statins based on the ACC/AHA risk score.

Within the Latin American context, the findings of a study conducted in Honduras2828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476., a lower middle-income country, which calculated the risk of CVD using four different calculators showed that an elevated proportion of individuals were high-risk. The prevalence of high-risk men and women based on the ACC/AHA risk score, GRS and MESA Risk Score was 62.0% and 29.8%, 46.1% and 15%, and 70.6% and 17.7%, respectively2828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476.. The findings of a study1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80. undertaken with 2,183 individuals from different regions of Peru assessing agreement between seven CVD risk scores were similar to ours. Agreement between the scores was poor and the variation in proportion of high-risk individuals was high: 29% (16.9-31.0%) based on the ACC/AHA risk score to 0.6% (0.2-8.6%) using the WHO score. The authors of this study1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80. concluded that there is uncertainty as to the selection of an appropriate CVD risk calculator in Peru and other low and middle-income countries, which is corroborated by the findings of the present study.

As mentioned above, the low level of agreement across scores for the identification of high-risk individuals may be related to the different CVD outcomes included in each score and the underlying risk factors assigned to different populations. In this respect, data supporting these factors is more readily available in developed countries due to the larger number and frequency of longitudinal studies. Other factors may include changes in RFs and the behavior of the population towards these factors over time, highlighting the need for prevalence and follow-up studies using biochemical and anthropometric data to measure trends1010 Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol. 2015; 14(2):74-80..

A meta-analysis highlighted the benefits of statins for primary prevention, showing that they reduce the risk of mortality from major cardiovascular events99 Taylor F, Huffman MD, Macedo AF, Moore THM, Burke M, Smith GD, Ward K, Ebrahim S. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2013; (1):CD004816.. These findings, combined with drug safety profiles and the reduced cost of these medicines has resulted in more permissive criteria for the indication of statins in some countries99 Taylor F, Huffman MD, Macedo AF, Moore THM, Burke M, Smith GD, Ward K, Ebrahim S. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2013; (1):CD004816.. However, there are still divergences regarding the definition of the threshold for high risk of CVD in each score. This definition considers the risk-benefit of statins99 Taylor F, Huffman MD, Macedo AF, Moore THM, Burke M, Smith GD, Ward K, Ebrahim S. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2013; (1):CD004816.. Traditionally, the high-risk threshold for the GRS has been set at >20%. A new score proposed by the ACC/AHA in 20132828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476. considers only fatal and non-fatal acute myocardial infarctions and strokes and reduces the risk threshold to >7.5%2828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476., thus increasing the proportion of the population classified as being at high risk of CVD. Some authors have criticized the reduction of the threshold, claiming that it results in overmedicalization and additional costs for certain countries, particularly those with limited resources2929 The Lancet Statins for millions more? Lancet. 2014; 383(9918):669.. A cost-effectiveness analysis conducted in Brazil3030 Ribeiro RA, Duncan BB, Ziegelmann PK, Stella SF, Vieira JLC, Restelatto LMF, Polanczyk CA. Cost-effectiveness of high, moderate and low-dose statins in the prevention of vascular events in the Brazilian public health system. Arq Bras Cardiol. 2015; 104(1):32-44. suggests that the use of moderate-dose statins is cost-effective for high-risk patients, considering the GRS and threshold of > 20%. However, it is not known whether the same is true for more permissive thresholds. In addition, consideration should be given to the availability of financial resources for the large-scale expansion of therapy.

Regardless of which calculator is used to calculate the risk of CVD, it is important to emphasize that CVD scores have limitations, in so far as they generally assess 10-year risk – meaning they may underestimate lifetime risk2828 Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature. 2020; 27:100476. – and do not include proximal risk factors such as socioeconomic conditions3131 Dalton JE, Perzynski AT, Zidar DA, Rothberg MB, Coulton CJ, Milinovich AT, Einstadter D, Karichu JK, Dawson NV. Accuracy of Cardiovascular Risk Prediction Varies by Neighborhood Socioeconomic Position: A Retrospective Cohort Study. Ann Intern Med. 2017; 167(7):456-464. – which are related to access to quality health care – and geographic location. Moreover, they do not consider modifying factors that increase the risk of CVD – which should be analyzed individually – such as family history of early-onset CVD3131 Dalton JE, Perzynski AT, Zidar DA, Rothberg MB, Coulton CJ, Milinovich AT, Einstadter D, Karichu JK, Dawson NV. Accuracy of Cardiovascular Risk Prediction Varies by Neighborhood Socioeconomic Position: A Retrospective Cohort Study. Ann Intern Med. 2017; 167(7):456-464., familial hypercholesterolemia3232 Sivapalaratnam S, Boekholdt SM, Trip MD, Sandhu MS, Luben R, Kastelein JJP, Wareham NJ, Khaw KT. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Heart. 2010; 96(24):1985-1989., chronic kidney disease, inflammatory diseases, and smoking3333 Jackson R, Lawes CMM, Bennett DA, Milne RJ, Rodgers A. Treatment with drugs to lower blood pressure and blood cholesterol based on an individual's absolute cardiovascular risk. Lancet. 2005; 365(9457):434-441.. Thus, for the proper application of these tools in clinical practice, it is important to carefully assess the real CVD risk and individualize treatment. Moreover, although the increase in risk of CVD due to RFs such as diabetes and high blood pressure may be gradual and can be reduced with treatment, some scores assess this risk dichotomously, while others do not include the treatment of these conditions as variables in the risk prediction model3333 Jackson R, Lawes CMM, Bennett DA, Milne RJ, Rodgers A. Treatment with drugs to lower blood pressure and blood cholesterol based on an individual's absolute cardiovascular risk. Lancet. 2005; 365(9457):434-441..

In the absence of a specific risk calculator for the Brazilian population, we will have to use one of the risk prediction tools cited above to calculate risk of CVD and assess eligibility for primary prevention, as recommended by the WHO44 World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling. Genebra: WHO; 2016.. However, when selecting the risk calculator it is important to be fully aware of the differences between the tools in terms of risk factors and CVD outcomes and choose the appropriate high-risk threshold. In addition, it is important to bear in mind that agreement across tools for the identification of high-risk individuals is poor, while the level of agreement for the assessment of low risk is high – except for the AHA/ACC score, which adopts more permissive criteria for the use of statins by reducing the high-risk threshold to >7.5%2121 D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008; 117(6):743-753.. In other words, with the majority of calculators, individuals identified as low-risk are very likely to be at low-risk; however, in the case of individuals identified as being at high risk, careful assessment is necessary before introducing drug therapy.

This study has some limitations. First, not all the equations used by the calculators were available and graph-based scores do not provide continuous risk assessments. Second, comparability may have been affected by differences in the definitions of predictors and CVD outcomes used in the scores assessed by this study, as shown in Table 1. However, the aim of this study was to compare the calculators precisely as they are, as recommended in the Brazilian guidelines – i.e. the tools applied in everyday clinical practice – in order to show just how important it is to understand the definitions used in the scores and each risk assessment tool’s limitations. Moreover, we believe that most health professionals that apply the scores in clinical practice are not necessarily aware of the technical details and modelling behind the “high 10-year risk of CVD” label, potentially resulting in the introduction of ineffective drug therapy for the prevention of CVD.

While doubts about the identification of individuals who should receive drug therapy to prevent CVD prevail in everyday clinical practice, non-pharmacological interventions – which encourage the adoption of healthy lifestyle habits such as stopping smoking, healthy eating, regular physical exercise and reducing alcohol intake – and population-wide interventions, regardless of individual baseline risk, should be implemented on a large scale to reduce CVD morbidity and mortality3434 Lloyd-Jones DM , Hong Y , Labarthe D , Mozaffarian D , Appel LJ , Horn LV , Greenlund K, Daniels S , Nichol D , Tomaselli GF , Arnett DK , Fonarow GC , Ho PM , Lauer MS , Masoudi FA , Robertson RM , Roger V , Schwamm LH, Sorlie P , Yancy CW, Rosamond WD, American Heart Association Strategic Planning Task Force and Statistics Committee. Defining and Setting National Goals for Cardiovascular Health Promotion and Disease Reduction. Circulation. 2010; 121(4):586-613.. With regard to specific preventive measures directed at high-risk populations, there is an urgent need to develop a Brazilian CVD risk calculator derived from data from national cohort studies or validate international calculators that correctly identify high-risk individuals to ensure effective treatment and inform policy planning. It is also important to advance efforts to define reference values for laboratory tests in Brazil based on nationwide studies taking into account ethnic diversity and local social and cultural characteristics, establishing recommended ranges as proposed by the NHS laboratory3535 Szwarcwald CL, Malta DC, Pereira CA, Figueiredo AW, Almeida WS, Machado IE, Bacal NS, Silva AG, Silva Júnior JB, Rosenfeld LG. Valores de referência para exames laboratoriais de colesterol, hemoglobina glicosilada e creatinina da população adulta brasileira. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190002..

Conclusion

This study analyzed the proportion of individuals from a representative subsample of Brazil’s National Health Survey classified as being at high-risk of CVD based on six different risk scores. Using the score recommended by the Brazilian Society of Cardiology as a reference, in general, the other CVD risk calculators assessed by this study showed low sensitivity and high specificity for the identification of high-risk individuals. In addition, these scores, currently the most commonly used CVD risk assessment tools, showed a wide variation in the proportion of individuals classified as being at high risk of CVD. In the absence of a specific risk calculator derived from national cohort studies or validated for use with the Brazilian population, the generalization of risk equations and definition of thresholds for drug therapy should be rediscussed for each context considering the cost-effectiveness of the recommendations.

Acknowledgments

We are grateful to the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, acronym in Portuguese) for awarding a research scholarship to DCM and ALPR.

The Health Surveillance Secretariat through funding from the Ministério da Saúde.

References

  • 1
    World Health Organization (WHO). Noncommunicable diseases country profiles 2018 Genebra: WHO; 2018.
  • 2
    Souza MFM, Malta DC, França EB, Barreto ML. Transição da saúde e da doença no Brasil e nas Unidades Federadas durante os 30 anos do Sistema Único de Saúde. Cien Saude Colet 2018; 23(6):1737-1750.
  • 3
    Harper S, Lynch J, Smith GD. Social determinants and the decline of cardiovascular diseases: understanding the links. Annu Rev Public Health 2011; 32:39-69.
  • 4
    World Health Organization (WHO). Global NCD target prevent heart attacks and strokes through drug therapy and counselling Genebra: WHO; 2016.
  • 5
    World Health Organization (WHO). "Best Buys" Tackling NCDs: Best buys and other recommended interventions for the prevention and control of noncommunicable diseases Genebra: WHO; 2017.
  • 6
    World Health Organization (WHO). Global action plan for the prevention and control of noncommunicable diseases 2013-2020 Genebra: WHO; 2013.
  • 7
    Whelton PK, Carey RM, Aronow WS, Casey Junior DE, Collins KJ, Himmelfarb CD, DePalma SM, Gidding S, Jamerson KA, Jones DW, MacLaughlin EJ, Muntner P, Ovbiagele B, Smith Junior SC, Spencer CC, Stafford RS, Taler ST, Thomas RJ, Williams KA, Williamson JD, Wright Junior JT. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71(6):1269-1324.
  • 8
    Malachias MVB, Souza WKSB, Plavnik FL, Rodrigues CIS, Brandão AA, Neves MFT, Bortolotto LA, Franco RJS, Poli-de-Figueiredo CE; Sociedade Brasileira de Cardiologia. 7ª Diretriz Brasileira de Hipertensão Arterial. Arquivos Brasileiros de Cardiologia 2016; 107(Supl. 3):1-82.
  • 9
    Taylor F, Huffman MD, Macedo AF, Moore THM, Burke M, Smith GD, Ward K, Ebrahim S. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev 2013; (1):CD004816.
  • 10
    Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, Pillay T, Gilman RH, Checkley W, Malaga G, Smeeth L, Bernabé-Ortiz A, Miranda JJ, Peru Migrant Study; CRONICAS Cohort Study Group. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Crit Pathw Cardiol 2015; 14(2):74-80.
  • 11
    WHO CVD Risk Chart Working Group. World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health 2019; 7(10):E1332-E1345.
  • 12
    Szwarcwald CL, Malta DC, Souza Júnior PRB, Almeida WS, Damacena GN, Pereira CA, Rosenfeld LG. Exames laboratoriais da Pesquisa Nacional de Saúde: metodologia de amostragem, coleta e análise dos dados. Rev Bras Epidemiol 2019; 22(Supl. 2):E190004.
  • 13
    Malta DC, Duncan BB, Schmidt MI, Machado ÍE, Silva AG, Bernal RTI, Pereira CA, Damacena GN, Stopa SR, Rosenfeld LG, Szwarcwald CL. Prevalência de diabetes mellitus determinada pela hemoglobina glicada na população adulta brasileira, Pesquisa Nacional de Saúde. Rev Bras Epidemiol 2019; 22(Supl 2):E190006.
  • 14
    Malta DC, Szwarcwald CL, Machado IE, Pereira CA, Figueiredo AW, Sá ACMGN, Velasquez-Melendez G, Santos FM, Souza-Júnior PRB, Stopa SR, Rosenfeld LG. Prevalência de colesterol total e frações alterados na população adulta brasileira: Pesquisa Nacional de Saúde. Rev Bras Epidemiol 2019; 22(Supl. 2):1-13.
  • 15
    American medical association. Executive summary of the Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001; 285(19):2486-2497.
  • 16
    Malta DC, Santos NB, Perillo RD, Szwarcwald CL. Prevalence of high blood pressure measured in the Brazilian population, National Health Survey, 2013. São Paulo Med J 2016; 134(2):163-170.
  • 17
    Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa Nacional de Saúde 2013: percepção do estado de saúde, estilos de vida e doenças crônicas: Brasil, Grandes Regiões e Unidades da Federação Rio de Janeiro: IBGE; 2014.
  • 18
    Gonçalves RPF, Haikal DS, Freitas MIF, Machado IE, Malta DC. Diagnóstico médico autorreferido de doença cardíaca e fatores de risco associados: Pesquisa Nacional de Saúde. Rev Bras Epidemiol. 2019; 22(Supl. 2):E190016.
  • 19
    Précoma DB, Oliveira GMM, Simão AS, Dutra OP, Coelho OR, Izar MCO, Póvoa RMS, Giuliano ICB, Alencar Filho AC, Machado CV, Scherr C, Fonseca FAH, Santos Filho RD, Carvalho T, Avezum Jr. A, Esporcatte R, Nascimento BR, Brasil DP, Soares GP, Villela PB, Ferreira RM, Martins WA, Sposito AC, Halpern B, Saraiva JFK, Carvalho LSF, Tambascia MA, Coelho-Filho OR, Bertolami A, Correa Filho H, Xavier HT, Faria-Neto JR, Bertolami MC, Giraldez VZR, Brandão AA, Feitosa ADM, Amodeo C, Souza DSM, Barbosa ECD, Malachias MVB, Souza WKSB, Costa FAA, Rivera IR, Pellanda LC, Silva MAM, Achutti AC, Langowiski AR, Lantieri CJB, Scholz JR, Ismael SMC, Ayoub JCA, Scala LCN, Neves MF, Jardim PCBV, Fuchs SCPC, Jardim TSV, Moriguchi EH, Schneider JC, Assad MHA, Kaiser SE, Lottenberg AM, Magnoni CD, Miname MH, Lara RS, Herdy AH, Araújo CGS, Milani M, Silva MMF, Stein R, Lucchese FA, Nobre F, Griz HB, Magalhães LBNC, Borba MHE, Pontes MRN, Mourilhe-Rocha R. Atualização da Diretriz de Prevenção Cardiovascular da Sociedade Brasileira de Cardiologia - 2019. Arq Bras Cardiol 2019; 113(4):787-891.
  • 20
    Goff Junior DC, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, Greenland P, Lackland DT, Levy D, O’Donnell CJ, Robinson JG, Schwartz JS, Shero ST, Smith Junior SC, Sorlie P, Stone NJ, Wilson PWF. American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63(25):2935-2959.
  • 21
    D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation 2008; 117(6):743-753.
  • 22
    Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, De Backer G, De Bacquer D, Ducimetière P, Jousilahti P, Keil U, Njølstad I, Oganov R G, Thomsen T, Tunstall-Pedoe H, Tverdal A, Wedel H, Whincup P, Wilhelmsen L, Graham IM, SCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003; 24(11):987-1003.
  • 23
    Lim SS, Gaziano TA, Gakidou E, Reddy KS, Farzadfar F, Lozano R, Rodgers A. Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs. Lancet 2007; 370(9604):2054-2062.
  • 24
    World Health Organization (WHO). World Health organization/International Society of Hypertension (WHO/ISH) risk prediction charts for 14 WHO epidemiological sub-regions Genebra: WHO; 2007.
  • 25
    Allan GM, Nouri F, Korownyk C, Kolber MR, Vandemeer B,McCormack J. Agreement Among Cardiovascular Disease Risk Calculators. Circulation 2013; 127:1948-1956.
  • 26
    DeFilippis AR, Young R, Carrubba CJ, McEvoy JW, Budoff MJ, Blumenthal RS, Kronmal RA, McClelland RS, Nasir K, Blaha MJ. An analysis of calibration and discrimination among multiple cardiovascular risk scores in a modern multiethnic cohort. Ann Intern Med 2015; 162(4):266-275.
  • 27
    Kavousi M, Leening MJG, Nanchen D, Greenland P, Graham IM, Steyerberg EW, Ikram MA Stricker BH, Hofman A, Franco OH. Comparison of Application of the ACC/AHA Guidelines, Adult Treatment Panel III Guidelines, and European Society of Cardiology Guidelines for Cardiovascular Disease Prevention in a European Cohort. JAMA 2014; 311(14):1416-1423.
  • 28
    Sanchez EEM, Urrutia SA, Rodriguez AA, Duarte G, Murillo A, Rivera R, Henriquez AAP, Sanchez DMM, Ordoñez E, Norwood DA, Dominguez LB, Dominguez RL, Torres K, Fajardo EMR, Godoy CA. Cardiovascular risk assessment in the resource limited setting of Western Honduras: An epidemiological perspective. IJC Heart & Vasculature 2020; 27:100476.
  • 29
    The Lancet Statins for millions more? Lancet 2014; 383(9918):669.
  • 30
    Ribeiro RA, Duncan BB, Ziegelmann PK, Stella SF, Vieira JLC, Restelatto LMF, Polanczyk CA. Cost-effectiveness of high, moderate and low-dose statins in the prevention of vascular events in the Brazilian public health system. Arq Bras Cardiol 2015; 104(1):32-44.
  • 31
    Dalton JE, Perzynski AT, Zidar DA, Rothberg MB, Coulton CJ, Milinovich AT, Einstadter D, Karichu JK, Dawson NV. Accuracy of Cardiovascular Risk Prediction Varies by Neighborhood Socioeconomic Position: A Retrospective Cohort Study. Ann Intern Med 2017; 167(7):456-464.
  • 32
    Sivapalaratnam S, Boekholdt SM, Trip MD, Sandhu MS, Luben R, Kastelein JJP, Wareham NJ, Khaw KT. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Heart 2010; 96(24):1985-1989.
  • 33
    Jackson R, Lawes CMM, Bennett DA, Milne RJ, Rodgers A. Treatment with drugs to lower blood pressure and blood cholesterol based on an individual's absolute cardiovascular risk. Lancet 2005; 365(9457):434-441.
  • 34
    Lloyd-Jones DM , Hong Y , Labarthe D , Mozaffarian D , Appel LJ , Horn LV , Greenlund K, Daniels S , Nichol D , Tomaselli GF , Arnett DK , Fonarow GC , Ho PM , Lauer MS , Masoudi FA , Robertson RM , Roger V , Schwamm LH, Sorlie P , Yancy CW, Rosamond WD, American Heart Association Strategic Planning Task Force and Statistics Committee. Defining and Setting National Goals for Cardiovascular Health Promotion and Disease Reduction. Circulation 2010; 121(4):586-613.
  • 35
    Szwarcwald CL, Malta DC, Pereira CA, Figueiredo AW, Almeida WS, Machado IE, Bacal NS, Silva AG, Silva Júnior JB, Rosenfeld LG. Valores de referência para exames laboratoriais de colesterol, hemoglobina glicosilada e creatinina da população adulta brasileira. Rev Bras Epidemiol 2019; 22(Supl. 2):E190002.

Publication Dates

  • Publication in this collection
    19 Apr 2021
  • Date of issue
    Apr 2021

History

  • Received
    18 Jan 2021
  • Accepted
    18 Jan 2021
  • Published
    20 Jan 2021
ABRASCO - Associação Brasileira de Saúde Coletiva Av. Brasil, 4036 - sala 700 Manguinhos, 21040-361 Rio de Janeiro RJ - Brazil, Tel.: +55 21 3882-9153 / 3882-9151 - Rio de Janeiro - RJ - Brazil
E-mail: cienciasaudecoletiva@fiocruz.br