SciELO - Scientific Electronic Library Online

 
vol.115 issue5Neck Circumference and 10-Year Cardiovascular Risk at the Baseline of the ELSA-Brasil Study: Difference by SexEvolving Outcome of Acute Myocardial Infarctions in Five Brazilian Geographic Regions Over Two Decades author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Arquivos Brasileiros de Cardiologia

Print version ISSN 0066-782XOn-line version ISSN 1678-4170

Arq. Bras. Cardiol. vol.115 no.5 São Paulo Nov. 2020  Epub Dec 04, 2020

https://doi.org/10.36660/abc.20190438 

Original Article

Mortality Due to Acute Myocardial Infarction in Brazil from 1996 to 2016: 21 Years of Disparities in Brazilian Regions

Letícia de Castro Martins Ferreira1 
http://orcid.org/0000-0002-2916-4477

Mário Círio Nogueira1 
http://orcid.org/0000-0001-9688-4557

Marilia Sá Carvalho2 
http://orcid.org/0000-0002-9566-0284

Maria Teresa Bustamante Teixeira1 
http://orcid.org/0000-0003-0727-4170

1Universidade Federal de Juiz de Fora - Faculdade de Medicina, Juiz de Fora, MG – Brazil

2Fundação Oswaldo Cruz Programa de Computação Científica – PROCC, Rio de Janeiro, RJ – Brazil


Abstract

Background:

Acute myocardial infarction (AMI), the leading cause of death in Brazil, has presented regional disparities in mortality rate time trends in recent years. Previous time trend studies did not correct for cause-of-death garbage codes, which may have skewed the estimates.

Objective:

To analyze regional and gender-based inequalities in the AMI mortality trend in Brazil from 1996-2016.

Methods:

A 21-year time series study (1996-2016). Data are from the Mortality Information System and population estimates from the Brazilian Institute of Geography and Statistics. Corrections of deaths due to ill-defined causes of death, garbage codes, and underreporting were made. The time series broken down by major geographic regions, gender, capital cities, and other municipalities was analyzed using the linear regression technique segmented by Jointpoint. Statistical significance level was set at 5%.

Results:

In the period, mortality decreased more sharply in women (−2.2%; 95% CI: −2.5; −1.9) than in men (−1.7%; 95% CI: - 1.9; −1.4) and more in the capital cities (−3.8%; 95% CI: - 4.3; −3.3) than in other municipalities (−1.5%; 95% CI: - 1.8; −1.3). Regional inequalities were observed, with an increase for men living in other municipalities of the North (3.3; 95% CI: 1.3; 5.4) and Northeast (1.3%; 95% CI: 1.0; 1.6). Statistical significance level was set at 5%. Mortality rates after corrections showed a significant difference in relation to the estimates without corrections, mainly due to the redistribution of garbage codes.

Conclusions:

Although AMI-related mortality has decreased in Brazil in recent years, this trend is uneven by region and gender. Correcting the numbers of deaths is essential to obtaining more reliable estimates.

Keywords: Myocardial Infarction; Epidemiology; Mortality; Time Series Studies; Myocardial Ischemia

Resumo

Fundamento:

O infarto agudo do miocárdio (IAM), principal causa de morte no Brasil, apresenta disparidades regionais nas tendências temporais das taxas de mortalidade dos últimos anos. Estudos anteriores de tendências temporais não fizeram correção para os códigos-lixo de causas de mortalidade, o que pode ter enviesado as estimativas.

Objetivo:

Analisar as desigualdades regionais e por sexo na tendência de mortalidade por IAM no Brasil no período de 1996 a 2016.

Métodos:

Estudo de séries temporais de 21 anos (1996 a 2016). Os dados são do Sistema de Informações sobre Mortalidade (SIM) e das estimativas populacionais do Instituto Brasileiro de Geografia e Estatística (IBGE). Foram feitas correções de óbitos por causas mal definidas, códigos-lixo e sub-registro. As séries temporais desagregadas por grandes regiões, sexo, capitais e interior foram analisadas utilizando a técnica de regressão linear segmentada por Jointpoint. O nível de significância estatística adotado foi de 5%.

Resultados:

No período, a mortalidade diminuiu mais acentuadamente no sexo feminino (–2,2%; IC 95%: –2,5; –1,9) do que no masculino (–1,7%; IC 95%: –1,9; –1,4), e mais nas capitais (–3,8%; IC 95%: –4,3; –3,3) do que no interior (–1,5%; IC 95%: –1,8; –1,3). Foram verificadas desigualdades regionais com aumento para homens residentes no interior do Norte (3,3; IC 95%: 1,3; 5,4) e Nordeste (1,3%; IC 95%: 1,0; 1,6). O nível de significância estatística adotado foi de 5%. As taxas de mortalidade após correções, principalmente pela redistribuição dos códigos-lixo, apresentaram expressiva diferença em relação às estimativas sem correções.

Conclusões:

Embora a mortalidade por IAM apresente redução no Brasil nos últimos anos, essa tendência é desigual segundo região e sexo. Desse modo, as correções dos números de óbitos são essenciais para estimativas mais fidedignas.

Palavras-chave: Infarto do Miocárdio/mortalidade; Epidemiologia; Mortalidade; Estudos de Séries Temporais; Isquemia Miocárdica

Introduction

In recent decades, cardiovascular diseases (CVD), and specifically ischemic heart disease (IHD) have become the primary causes of death worldwide, although age-standardized mortality rates have dropped.1

When conducting mortality studies, one must to pay attention to the quality of death records which, in Brazil, differs between geographic regions and between municipalities, with it being better in the capital cities. Some indirect indicators of standard data quality are the proportion of deaths from ill-defined causes of death, use of garbage codes, underreporting, and ignored age and gender. They reflect difficulties in diagnosing the diseases that caused death, accessing health services, filling out the death certificate, and/or entering data into the system.2 One way to solve this problem and properly estimate mortality rates is to make corrections that will allow greater comparability between regions over time.1,3,4

This study aims to analyze regional and gender inequalities in the AMI-related mortality trend in Brazil from 1996-2016, correcting deaths from ill-defined causes, garbage codes, and underreporting.

Methods

Time series (21 years: 1996 to 2016) of AMI-related mortality in the capital cities and cities and towns in the countryside (other municipalities) of large Brazilian regions were analyzed. Annual data on AMI-related deaths (code I21. ICD 10) were obtained from the Mortality Information System (SIM in Portuguese) on the DATASUS website - Department of Informatics of the Unified Health System (http://datasus.saude.gov.br) and population estimates from the Brazilian Institute of Geography and Statistics. As publicly available secondary data, the study was exempted from approval by a research ethics committee in accordance with CONEP Resolution 510 of April 7, 2016. The SIM has national coverage and, in recent years, it has greatly improved the quality of its database between Brazilian regions and municipalities. To make a more meaningful comparison between regions and over the time period covered by this study, we corrected the numbers obtained from the SIM regarding ill-defined causes of death, use of garbage codes, and underreporting using procedures adopted in other studies.1,3,5

Figure 1 outlines the procedures used to correct AMI-related deaths for ill-defined causes, use of garbage codes, and underreporting.

Figure 1 Procedures for the correction of the number of deaths regarding ill-defined causes, use of garbage codes, and subregisters. 

The ill-defined causes of death are those in which the cause of death has not been established and have therefore been classified in codes R00-R99 from Chapter XVIII of ICD-10: “Unclassified Symptoms, Signs, and Abnormal Clinical and Laboratory Findings Not Elsewhere Classified”.2,5,6 The redistribution of ill-defined causes of death (Correction 1) was conducted in the following manner: for each year and region, correction factors (CF1) were calculated using equation (1) for each gender and for each five-year age range. To calculate the redistribution of deaths, the number of deaths was multiplied by the CF1.3,7

(FC1)= Total deathsdeaths by external causes (total deathsdeaths by external causes) deaths by illdefined causes

Deaths with garbage codes are those in which ICD 10 codes are used that are nonspecific and do not precisely identify the underlying cause of death.5 The following garbage codes are used in cardiology: I50, I46, I47.0, I47.1, I47.2, I47.9, I48, I49.0, I49.9, I51.4, I51.5, I51.6, I51.9 and I70.0. To proceed with Correction 2 for garbage codes, deaths with cardiologic garbage codes were added to deaths recorded as being caused by AMI in the following proportion: 70% of deaths per 150 in people between 30-60 years of age and 80% for people over 80 years of age, along with other causes 75% (30-60 years of age) and 60% (older than 60 years of age).5

To correct for underreported death (Correction 3), meaning deaths that were neither recorded in the Civil Registry System nor in the SIM, the correction factors estimated for Brazil, region, and states,7 which are readily available in DATASUS, were used.6 Correction 3 was made by multiplying the underreporting correction factor for deaths in other municipalities. This correction was not carried out in the state capitals, as studies have shown that death records in these cities are reliable.2,3,5 With respect to the periods from 1996 to 1999 and from 2014 to 2016, for which correction factors were not available, values for the next closest years were used.

Corrections due to ignored gender and age were not conducted in this study, since both categories presented reliable numbers during the studied period.4

Mortality rates were calculated and standardized by five-year age groups for adults (20 years of age and over) using the new standard world population as a reference.8 The standard world population was proposed by the WHO as a way to compare mortality rates between populations with different age groups. Estimates were prepared for each five-year period from 1950 to 2025 based on population censuses and other demographic sources, then adjusted for enumeration errors. From there, an average age structure of the world population was constructed. The specific mortality rates by age groups were applied to the respective population contingents of the standard population. Consequently, the number of deaths expected to occur in each age group was obtained, provided the population studied had the same age composition as the standard population. The standardized mortality rate was obtained by dividing the total number of deaths expected for the standard population. which can then be compared to other populations, and any differences found are not due to variations in the age structure.8,9

The time trend analysis of the corrected mortality rates as standardized by region, capital cities and other municipalities, and gender was performed by segmented linear regression using Joinpoint version 4.6.0.0,10,11 a method used in other AMI time trend studies.12,13 Models were adjusted with change points in the temporal evolution of the death rates (joint points) ranging from zero (trend represented by a single line segment) to three. Annual percentage changes (APC) were calculated for the period under analysis. Statistical significance level was set at 5%.

Results

A comparison of AMI-related mortality rates in all regions of Brazil, with and without corrections for ill-defined causes of death (Correction 1), garbage codes (Correction 2), and underreporting (Correction 3) by female and male is shown in Tables 1 and 2. respectively. Larger increases were noted after correction for garbage codes, reaching a 92% increase in 1996 among women living in other municipalities of the Central West. The proportional difference in mortality rates with and without corrections, between 1996 and 2016, showed little discrepancy between estimates in the capital cities. However, relative to other municipalities, an important disparity could be seen between them (Table 2). Figure 2 shows that not only does the magnitude of mortality rates increase after corrections, but the time trend slope also changes and percentages of correction are higher at the beginning of the period than at the end. Table 3 shows an increase in the frequency of garbage codes in Brazilian regions.

Table 1 Comparison of AMI-related mortality rates in Brazilian regions with and without corrections for ill-defined causes of death, garbage codes, and underreporting in women 

Region Location Year Standardized rates % Changes
Not corrected Correction 1 Correction 2 Correction 3 % Change 1 % Change 2 % Change 3 Total % change
Brazil Capital cities 1996 79.1 82.9 127.1 127.1 5 53 0 61
2016 42.7 43.7 65.2 65.2 2 49 0 53
Dif% −46 −47 −49 −49
Other municipalities 1996 52.9 64.3 105.4 116.5 21 64 11 120
2016 53.6 56.2 85.6 89.4 5 52 4 67
Dif% 1 −12 −19 −23
North Capital cities 1996 50.1 57.6 98.7 98.7 15 71 0 97
2016 38.1 40.9 62.7 62.7 7 53 0 64
Dif% −24 −29 −36 −36
Other municipalities 1996 22.5 37.6 57.7 83.4 67 54 45 271
2016 51.4 55.4 83.1 97.7 8 50 18 90
Dif% 129 47 44 17
Northeast Capital cities 1996 56.0 59.1 105.5 105.5 6 79 0 88
2016 42.5 43.8 63.8 63.8 3 46 0 50
Dif% −24 −26 −40 −40
Other municipalities 1996 25.4 47.5 68.4 87.9 87 44 28 246
2016 60.2 64.6 95.7 104.2 7 48 9 73
Dif% 137 36 40 19
Central West Capital cities 1996 52.4 55.2 99.5 99.5 5 80 0 90
2016 38.5 38.8 54.1 54.1 1 39 0 41
Dif% −27 −30 −46 −46
Other municipalities 1996 46.6 54.8 105.2 120.4 18 92 14 158
2016 54.4 55.6 83.1 90.0 2 49 8 66
Dif% 17 2 −21 −25
Southeast Capital cities 1996 91.7 95.7 140.1 140.1 4 46 0 53
2016 46.4 47.2 72.5 72.5 2 54 0 56
Dif% −49 −51 −48 −48
Other municipalities 1996 66.4 74.0 128.0 134.0 11 73 5 102
2016 47.9 50.4 79.7 81.8 5 58 3 71
Dif% −28 −32 −38 −39
South Capital cities 1996 91.9 92.2 130.3 130.3 0 41 0 42
2016 32.3 32.8 44.1 44.1 2 34 0 37
Dif% −65 −64 −66 −66
Other municipalities 1996 78.4 86.7 136.4 141.9 11 57 4 81
2016 44.1 45.4 74.2 76.9 3 63 4 74
Dif% −44 −48 −46 −46

Correction 1: AMI-related mortality rates corrected for ill-defined causes of death. Correction 2: AMI-related mortality rates corrected for garbage codes. Correction 3: AMI-related mortality rates corrected for underreporting.

Table 2 Comparison of AMI-related mortality rates in Brazilian regions with and without corrections for ill-defined causes of death, garbage codes, and underreporting in men 

Region Location Year Standardized rates % Changes
Not corrected Correction 1 Correction 2 Correction 3 % Changes 1 % Changes 2 % Changes3 Total % changes
Brazil Capital cities 1996 145.4 153.0 205.1 205.1 5 34 0 41
2016 86.0 88.9 117.5 117.5 3 32 0 37
Dif% −41 −42 −43 −43
Other municipalities 1996 86.5 105.2 150.2 167.3 22 43 11 93
2016 89.4 95.8 131.7 138.4 7 37 5 55
Dif% 3 −9 −12 −17
North Capital 1996 88.6 105.5 154.5 154.5 19 46 0 74
2016 85.3 92.7 123.3 123.3 9 33 0 44
Dif% −4 −12 −20 −20
Other municipalities 1996 37.5 62.8 85.9 122.8 68 37 43 228
2016 88.3 96.7 131.2 153.6 10 36 17 74
Dif% 136 54 53 25
Northeast Capital 1996 101.2 107.1 160.6 160.6 6 50 0 59
2016 84.9 88.0 119.3 119.3 4 36 0 41
Dif% −16 −18 −26 −26
Other municipalities 1996 39.9 71.8 95.8 123.9 80 33 29 210
2016 104.2 113.7 152.6 166.4 9 34 9 60
Dif% 161 58 59 34
Central West Capital 1996 79.0 84.6 136.6 136.6 7 61 0 73
2016 79.5 80.9 99.9 99.9 2 23 0 26
Dif% 1 −4 −27 −27
Other municipalities 1996 82.9 99.8 154.1 172.8 20 54 12 109
2016 95.8 99.7 134.5 146.0 4 35 8 52
Dif% 16 0 −13 −16
Southeast Capital 1996 174.7 182.5 235.9 235.9 4 29 0 35
2016 92.9 95.8 128.0 128.0 3 34 0 38
Dif% −47 −48 −46 −46
Other municipalities 1996 118.5 134.1 195.1 206.1 13 46 6 74
2016 88.5 94.7 129.9 133.2 7 37 3 50
Dif% −25 −29 −33 −35
South Capital 1996 164.6 165.6 208.9 208.9 1 26 0 27
2016 67.1 68.5 82.3 82.3 2 20 0 23
Dif% −59 −59 −61 −61
Other municipalities 1996 134.4 149.4 203.7 204.8 11 36 1 52
2016 83.4 86.7 120.7 121.3 4 39 0 45
Dif% −38 −42 −41 −41

Correction 1: AMI-related mortality rates corrected for ill-defined causes of death. Correction 2: AMI-related mortality rates corrected for garbage codes. Correction 3: AMI-related mortality rates corrected for underreporting.

Figure 2 Time trends in mortality rates from acute myocardial infarction before (dashed line) and after corrections for ill-defined causes of death, underreporting, and garbage codes (continuous line) in Brazil, regions, capital cities, and other municipalities by gender from 1996 to 2016. 

Table 3 Frequency of deaths classified with garbage codes for AMI by year/gender/region in Brazil from 1996 to 2016 

North Northeast Central West Southeast South Total
Year M F M F M F M F M F M F General
1996 883 813 4865 5117 1670 1501 14198 15996 4429 5140 26045 28567 54612
1997 911 797 5056 5175 1816 1696 13429 15240 4181 4881 25393 27789 53182
1998 972 898 5547 5540 1758 1594 13342 15350 4507 5221 26126 28603 54729
1999 1043 863 5379 5367 1827 1492 12632 14569 4098 4691 24979 26982 51961
2000 1046 821 5402 5431 1674 1511 12138 13859 4157 4837 24417 26459 50876
2001 1194 919 5677 5678 1770 1566 11931 13603 3926 4563 24498 26329 50827
2002 1103 900 5783 6109 1963 1668 11649 13821 3915 4784 24413 27282 51695
2003 1184 1002 5871 6141 1974 1623 12100 14002 4047 4703 25176 27471 52647
2004 1183 935 6484 6644 1967 1700 12798 14552 4196 4864 26628 28695 55323
2005 1248 1051 6915 7299 2092 1649 12642 14131 4019 4712 26916 28842 55758
2006 1329 1008 8210 8463 2104 1870 13354 15142 4009 4939 29006 31422 60428
2007 1420 1111 8469 8733 2129 1830 13486 15326 4316 5161 29820 32161 61981
2008 1454 1205 8541 8888 2125 1849 13784 15670 4379 5207 30283 32819 63102
2009 1464 1182 8538 9017 2107 1868 13588 15727 4464 5223 30161 33017 63178
2010 1551 1237 8300 8616 2179 1965 14372 16825 4464 5394 30866 34037 64903
2011 1587 1405 8784 9292 2128 1907 14522 17280 4649 5834 31670 35718 67388
2012 1611 1369 8583 9038 2155 2072 14340 16909 4514 5302 31203 34690 65893
2013 1699 1428 8923 9437 2199 1994 14720 16875 4873 5556 32414 35290 67704
2014 1737 1458 8609 9238 2197 2085 14589 17028 4709 5564 31841 35373 67214
2015 1843 1506 9006 9925 2191 2045 15143 18112 4810 5615 32993 37203 70196
2016 1866 1586 9200 9829 2013 1770 16496 18841 5260 6063 34835 38089 72924

F: Female; M: Male.

In general, trends in corrected AMI-related mortality rates are declining. Capital cities, with higher mortality rates at the beginning of the period, showed a more pronounced decline and, consequently, lower rates in recent years. Mortality rates in men were higher than those in women throughout the analyzed period, with both falling. Only mortality rates of men living in other municipalities of the North and Northeast showed an upward trend. At the beginning of the series, mortality rates were higher in the Southeast and South and, due to the more pronounced decline in these regions, their values began to be lower at the end of the period than in the North and Northeast (Figure 3). A percentage difference of −43.6% was observed between 1996 and 2016 in Brazil, with it being higher in the South (−85.1%). The Northeast and North, which in 1996 had the lowest rates, began to reflect the highest rates in 2016 (Table 4).

Figure 3 Time series after mortality corrections for acute myocardial infarction in Brazil, regions, capital cities, and other municipalities by gender from 1996 to 2016. 

Table 4 AMI-related mortality rates * standardized by the new world population by Brazilian region in 1996 and 2016 

Region 1996 2016 % Difference
Brazil 149.86 104.35 −43.6
North 107.05 112.48 4.8
Northeast 107.42 121.30 11.4
Central West 135.32 100.47 −34.7
Southeast 172.93 102.92 −68.0
South 168.74 91.14 −85.1

*per 100,000 inhabitants.

The segmented regression analysis indicated a decline in mortality rates in all regions except the Northeast (Table 5). The South (APC = −3.4%; 95% CI: - 3.8; −3.0) and Southeast (APC = −3.3%; 95% CI: - 3.9; −2.7) showed the highest percentage of decrease and the North, the lowest (APC = −0.8%; 95% CI: - 1.3; −0.2).

Table 5 Analysis of the segmented regression of the AMI-related mortality trend by gender, capital cities, and other municipalities of the Brazilian regions, 1996-2016 

Region Municipality Gender Trend 1 Trend 2 Trend 3 Trend 4
Period APC IC95% Period APC IC95% Period APC IC 95% Period APC IC95%
Brazil Capital cities F 1996 to 2010 −4.1 −4.7; −3.6 2010 to 2016 −1.1 −3.2; + 1.0
M 1996 to 2010 −3.5 −4.0; −3.1 2010 to 2016 −1.2 −3.0; +0.6
Other municipalities F 1996 to 2016 −1.4 −1.7; −1.1
M 1996 to 2016 −1 −1.3; −0.8
All Both 1996 to 2016 −1.9 −1.7; −2.2
North Capital cities F 1996 to 2006 −5.1 −6.6; −3.6 2006 to 2016 0.9 −0.7; 2.5
M 1996 to 1999 0.6 −5.6; 7.2 1999 to 2002 −10.8 −23.0; 3.4 2002 to 2016 0.5 0.0; 1.0
Other municipalities F 1996 to 2016 0.20 −0.3; 0.7
M 1996 to 2005 1.3 0.3; 2.3 2005 to 2010 −1.8 −5.5; 2.0 2010 to 2016 3.3 1.3;5.4
All Both 1996 to 2010 −0.8 −1.3; −0.2 2010 to 2016 2.4 0.2; 4.7
Northeast Capital cities F 1996 to 2000 −5.3 −9.1; −1.2 2000 to 2016 −2.0 −2.5; −1.6
M 1996 to 2016 −1.6 −2.0; −1.3
Other municipalities F 1996 to 2002 0.60 −1.0; +2.2 2002 to 2006 5.20 0.2;10.5 2006 to 2010 −3.0 −7.6; +1.8 2010 to 2016 0.5 −1.1; 2.1
M 1996 to 2016 1.30 1.0; 1.6
All Both 1996 to 2003 0.0 −1.2; 1.3 2003 to 2006 5.2 −4.0; 15.3 2006 to 2010 −2.8 −7.2; 1.8 2010 to 2016 1.0 −0.6; 2.6
Central West Capital cities F 1996 to 2016 −2.8 −3.3; −2.2
M 1996 to 2016 −1.7 −2.1; −1.2
Other municipalities F 1996 to 2016 −1.8 −2.2; −1.3
M 1996 a 2016 −1.0 −1.4; −0.6
All Both 1996 to 2016 −1.7 −2.1; −1.2
Southeast Capital cities F 1996 to 2010 −4 −5.0; −3.8 2010 to 2016 −0.6 −2.8;1.6
M 1996 to 2001 −5.7 −8.1; −3.2 2001 to 2016 −2.9 −2.5; −3.3
Other municipalities F 1996 to 2001 −5.2 −7.5 - 2.8 2001 to 2005 0.10 −5.5; 6.2 2005 to 2008 −6.4 −17.5 6.1 2008 to 2016 −0.9 0.2; −2.1
M 1996 to 2016 −2.3 −2.6; −1.9
All Both 1996 to 2009 −3.3 −3.9; −2.7 2009 to 2016 −1.3 −2.9; 0.4
South Capital cities F 1996 to 2016 −5.8 −6.2; −5.4
M 1996 to 2016 −5.2 −5.5; −4.8
Other municipalities F 1996 to 2016 −3.4 −3.8; −3.0
M 1996 to 2016 −2.9 −3.3; −2.5
All Both 1996 to 2016 −3.4 −3.8; −3.0

F: female; M: male; Both=Male+Female. All=entire region. APC: annual percentage changes. 95% CI: confidence interval. Statistical significance level: 5%.

A markedly different pattern was observed between the capital cities and other municipalities and between genders. There was a decrease in AMI-related mortality rates in all capitals, as well as in other municipalities of the Southeast, South, and Mid-West (Table 5). Conversely, the same rates increased in other municipalities of the North and Northeast, with the highest percentage of increase among women living in the Northeast between 2002 to 2006 (APC = 5.2%; 95% CI: 0.2%; 10.5%) (Table 5).

More significant declines in AMI-related mortality rates were observed among women living in the capital cities, except for the Southeast, where it was higher among men. The greatest decrease was observed among women living in the capital cities of the South from 1996 to 2016 (APC = −5.80%; 95% CI: - 6.2%; −5.4%). Smaller decreases were found among men living in other municipalities, except in the North (APC = 1.3%; 95% CI: 0.3; 2.3) and Northeast (APC = 1.3%; 95% CI: 1.0; 1.6), where an increase was observed (Table 5).

Discussion

Few nationally-based studies using data from the SIM system have attempted to estimate standardized mortality rates by means of the standard world population and with corrections for the proportion of deaths from ill-defined causes of death, the use of garbage codes, and underreporting.4 These corrections are used by studies worldwide.2,3,14 It was observed in this study that the number of ill-defined and underreported causes declined over the years studied, thus indicating improvements in data quality.3 This decrease occurred differently between regions and between capital cities and other municipalities. It was more recent in the North and Northeast and in other municipalities than in other regions and capital cities. In contrast, the use of garbage codes showed no signs of significant reduction, remaining very high in number in all regions.

The decline in trends in AMI-related mortality rates has been observed worldwide and in most Brazilian regions,1,3,1417 which is being studied for the first time in this study by capital cities and other municipalities. Time series studies of mortality from cardiovascular diseases developed in Brazil analyzed large regions and found differences in mortality from chronic non-communicable diseases (NCDs) in those regions.3,1619 This difference is partly justified by the increase in mortality rates among men living there.

Discrepancies found in the analysis by capital cities and other municipalities can be explained by demographic and epidemiological transitions, as well as the implementation of public health policies that occurred differently in these regions.16,20 Areas with greater socioeconomic development had earlier demographic and epidemiological transitions through urbanization, greater access to services, and the presence of an aging population. This led to a rise in chronic noncommunicable diseases and AMI-related mortality rates. Subsequently, they began to drop as new public policies were implemented.21 This transition occurred at different times in different cities and regions. The South and Southeast experienced it before the North and Northeast, even as capital cities preceded other municipalities.16,19 Capital cities generally offer more healthcare resources, better socioeconomic conditions, better health indicators, and better death records. Access to medium- and high-complexity services is also greater. Therefore, AMI-related mortality rates in the capital cities were lower than those in other municipalities, primarily in the final studied period. The time series in the 1990s revealed that mortality rates in other municipalities were lower in some regions and underwent an inversion in the middle of the period.14,16,17 The issue of underreporting also needs to be taken into account, along with the lack of access to health services for diagnosis and the proper completion of death certificates in other municipalities, which may explain the smaller number of cases recorded in this period20 and justifies, in part, the need for the corrections made.

It is interesting to highlight the turning point that these rates have undergone in all regions since 2000. A greater drop was observed in the Southeast, South, and Central West starting that year, whereas an increase in mortality rates was noted in other municipalities in the North and Northeast. This was a time when public policies in the healthcare area began to expand with increased funding, such as the National Primary Care Policy (PNAB in Portuguese) and the National Emergency Care Policy (PNAU in Portuguese). The Mobile Emergency Care Service (SAMU in Portuguese) was the first component of the PNAU to be implemented in the country in the early 2000s.22 Later came incentives for the implementation of Emergency Care Units (UPA).22 Concomitantly, primary care services wound up with an expanded structure through the implementation of the Family Health Strategy.21

Two movements led to a decline in mortality rates: one in relation to the prevention, control, and treatment of risk factors for IHD, with greater access to quality primary care, and another in the transportation, early diagnosis, and treatment of IHD through SAMU and emergency care units (UPA in Portuguese). However, in the North and Northeast, mortality rates rose in other municipalities. Historically, these have been the regions with the highest numbers of underreporting and the most difficulties in accessing healthcare services, especially in other municipalities.2,3 Federal financial incentives aimed at organizing primary care and urgent and emergency services provided them with much-needed expansion in healthcare services and, with that, improvements in diagnoses and records of the causes of deaths, which, when added to changes arising from an aging population, could explain why mortality rates have risen.22

IHD mortality rates in women were lower than in men and the reduction in female deaths was also greater, which is in line with data found in the literature.15,16 The cardiac protection promoted by female hormones (estrogen) may contribute to it. The presence of estrogen in the cardiovascular endothelium triggers the release of nitric oxide, leading to vasodilation; regulates prostaglandin production; and inhibits smooth muscle proliferation, factors related to AMI.23

The limitations of this study are inherent to the use of secondary data, although the quality of death records did improve during the analyzed period. Moreover, corrections were made that enhanced its validity. The factors associated with AMI-related mortality, such as obesity, smoking, and arterial hypertension, were not the object of this study.24 Permeating all of these factors are socioeconomic and cultural conditions that strongly influence the mortality rates identified in regional differences.15

Conclusions

The evolution of AMI-related mortality in Brazil from 1996 to 2016 showed a downward trend, characterized by important inequalities and disparities between genders, capital cities and other municipalities and regions. The importance of correcting causes of death (due to ill-defined causes, garbage codes, and underreporting) was emphasized to encourage the construction of more reliable indicators that would allow a proper assessment of mortality trends to be made.

Sources of Funding

There were no external funding sources for this study.

Study Association

This article is part of the thesis of Doctoral submitted by Letícia de Castro Martins Ferreira, from Universidade Federal de Juiz de Fora.

REFERENCES

1. GBD. Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2016 (GBD 2016) Results; 2017. [acesso em 15 maio. 2020]. Disponível em: http://ghdx.healthdata.org/gbd-results-tool. [ Links ]

2. Malta DC, Moura L, Prado RR, Escalante JC, Schmidt MI, Duncan BB. Chronic non-communicable disease mortality in Brazil and its regions, 2000-2011. Epidemiol Serv Saúde. 2014;23(4):599-608. [ Links ]

3. Schmidt MI, Duncan BB, Stevens A, Luft V, Iser BPM, Moura L, et al. Saúde Brasil 2010: uma análise da situação de saúde e de evidências selecionadas de impacto de ações de vigilância em saúde. Série G: estatística e informação em saúde. Brasília, DF: Ministério da Saúde; 2011. p. 117-35. [ Links ]

4. Soares DA, Gonçalves MJ. Cardiovascular mortality and impact of corrective techniques for dealing with underreported and ill-defined deaths. Rev Panam Salud Pública. 2012;32(3):199-206. [ Links ]

5. Gadelha AMJ, Leite IC, Valente JG, Schramm JMA, Portela MC, Campos MR. Relatório final do projeto estimativa da carga de doença do Brasil – 1998. Projeto Carga de Doença. Rio de Janeiro: FENSPTEC; 2002. (Tecnologias em Saúde para a Qualidade de Vida) [ Links ]

6. Ministério da Saúde. DATASUS. Sistema de Informação de Mortalidade; 2018. [acesso em 15 maio 2020]. Disponível em: http://datasus.saude.gov.brLinks ]

7. Szwarcwald CL, Morais-Neto OL, Frias PG, Souza Jr PRB, Escalante JJC, Lima RB, et al. Busca ativa de óbitos e nascimentos no Nordeste e na Amazônia Legal: estimação das coberturas do SIM e do SiNASC nos municípios brasileiros. In: Saúde Brasil 2010: uma análise da situação de saúde e de evidências selecionadas de impacto de ações de vigilância em saúde. Brasília, DF: Ministério da Saúde; 2011. p. 51-79. [ Links ]

8. Ahmad OB, Boschi-Pinto C, Lopez AD, Murray CJL, Lozano R, Inoue M. Age standardization of rates: a new WHO standard; 2001. p. 14. [ Links ]

9. Medronho RA, Bloch KV, Luiz RR, Werneck GL. Epidemiologia. São Paulo: Atheneu; 2009. p. 38-42. [ Links ]

10. Kim H, Fay M, Feuer E, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med. 2000;19(3):335-51. [ Links ]

11. National Institute of Health. Joinpoint Regression Program, Statistical Methodology and Applications Branch, Surveillance Research Program, National Cancer Institute. [Internet]. USA: National Cancer Institute; 2018. [acesso em 15 maio 2020]. Disponível em: https://surveillance.cancer.gov/help/joinpoint/tech-help/citation. [ Links ]

12. Liu J, Chen T, Yu Y, Han X, Zheng X. Eleven-year trends of acute myocardial infarction incidence, mortality and case-fatality rate in Shanghai: a 1.09 million population based surveillance study. Int J Clin Exp Med. 2016;9(6):11385-96. [ Links ]

13. Vujcic IS, Sipetic SB, Dubljanin ES, Vlajinac HD. Trends in mortality rates from coronary heart disease in Belgrade (Serbia) during the period 1990-2010: a joinpoint regression analysis. BMC Cardiovasc Disord. 2013 Dec 9;13:112. [ Links ]

14. Marinho F, Passos VMA, Malta DC, França EB, Abreu DMX, Araújo VEM, et al. Burden of disease in Brazil, 1990-2016: a systematic subnational analysis for the Global Burden of Disease Study 2016. Lancet. 2018;392(10149):760-75. [ Links ]

15. Mansur AP, Favarato D. Mortality due to cardiovascular diseases in women and men in the five brazilian regions, 1980-2012. Arq Bras Cardiol. 2016;107(2):137-46. [ Links ]

16. Brant LCC, Nascimento BR, Passos VMA, Duncan BB, Bensenor IJM, Malta DC, et al. Variations and particularities in cardiovascular disease mortality in Brazil and Brazilian states in 1990 and 2015: estimates from the Global Burden of Disease. Rev Bras Epidemiol. 2017;20(suppl 1):116-28. [ Links ]

17. Ribeiro ALP, Duncan BB, Brant LCC, Lotufo PA, Mill JG, Barreto SM. Cardiovascular health in Brazil: trends and perspectives. Circulation. 2016;133(4):422-33. [ Links ]

18. Baena CP, Chowdhury R, Schio NA, Schio NA, Sabbag Jr AE, Grarita-Souza LC, et al. Ischaemic heart disease deaths in Brazil: current trends, regional disparities and future projections. Heart. 2013;99(18):1359-64. [ Links ]

19. França EB, Passos VMA, Malta DC, Duncan BB, Ribeiro ALP, Guimarães MDC, et al. Cause-specific mortality for 249 causes in Brazil and states during 1990–2015: a systematic analysis for the global burden of disease study 2015. Popul Health Metrics. 2017;15(1):39. [ Links ]

20. Borges GM. Health transition in Brazil: regional variations and divergence/convergence in mortality. Cad Saúde Pública. 2017;33(8):e00080316. [ Links ]

21. Pinto LF, Giovanella L. The Family Health Strategy: expanding access and reducing hospitalizations due to ambulatory care sensitive conditions (ACSC). Ciênc Saúde Coletiva. 2018;23(6):1903-14. [ Links ]

22. O'Dwyer G, Konder MT, Reciputti LP, Macedo C, Lopes MGM. Implementation of the Mobile Emergency Medical Service in Brazil: action strategies and structural dimension. Cad Saúde Pública. 2017;33(7):e00043716. [ Links ]

23. Mehta LS, Beckie TM, DeVon HA, Grines CL, Krumholz HM, Johnsonn MN, et al. Acute myocardial infarction in women. Circulation. 2016;133(9):916-47. [ Links ]

24. Gaziano TA, Prabhakaran D, Gaziano JMT. Risk factors: Global Burden of Cardiovascular Disease. In: Braunwald's heart disease: a textbook of cardiovascular medicine. 11. ed. Canadá: Elsevier; 2019. p. 1-17. [ Links ]

Received: April 02, 2019; Revised: October 07, 2019; Accepted: October 29, 2019

Mailing Address: Letícia de Castro Martins Ferreira • Faculdade de Medicina da Universidade Federal de Juiz de Fora - Avenida Eugênio Nascimento, s/nO. Postal Code 36038-330, Dom Bosco, Juiz de Fora, MG - Brazil E-mail: leticiacmferreira@gmail.com

Author Contributions

Conception and design of the research, Acquisition of data, Analysis and interpretation of the data, Statistical analysis and Writing of the manuscript: Ferreira LCM, Nogueira MC, Carvalho MS, Teixeira MTB; Critical revision of the manuscript for intellectual content: Nogueira MC, Carvalho MS, Teixeira MTB.

Potential Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License