Abstracts

Introduction

The prevalence of and mortality from tuberculosis have declined worldwide, and most countries are likely to achieve the goal of reducing tuberculosis prevalence and mortality by 50% by 2015 in comparison with 1990 rates.^{( 1}1. World Health Organization. The global plan to stop TB 2011-2015: transforming the fight towards elimination of tuberculosis - reprinted with changes, 2011. Geneva: WHO; 2011. ⁾ The World Health Organization has launched a new challenge: the elimination of tuberculosis by 2050.^{( 1}1. World Health Organization. The global plan to stop TB 2011-2015: transforming the fight towards elimination of tuberculosis - reprinted with changes, 2011. Geneva: WHO; 2011. ⁾ Tuberculosis is the second leading cause of death from an infectious disease, accounting for approximately 1.3 million deaths worldwide in 2012; this demonstrates the severity of tuberculosis, especially in the 22 countries that collectively account for 80% of the disease burden.^{( 2}2. World Health Organization. Global tuberculosis report 2013. Geneva: WHO; 2013. ⁾

Brazil ranks 16th among the countries with the highest number of cases, the tuberculosis incidence and mortality rates in 2012 being 36.1/100,000 population and 2.4/100,000 population, respectively.^{( 2}2. World Health Organization. Global tuberculosis report 2013. Geneva: WHO; 2013. ⁾ Brazil has faced major challenges to achieving the ambitious goal of eliminating the disease. For example, in the city of São Luís, which is one of the priority cities for tuberculosis control in the country,^{( 3}3. Ministério da Saúde. Secretaria de Vigilância em Saúde. Programa Nacional de Controle da Tuberculose. Manual de recomendações para o controle da tuberculose no Brasil. Brasília: Ministério da Saúde; 2011. ⁾ the tuberculosis incidence and mortality rates in 2012 were 53.1/100,000 population and 3.9/100,000 population, respectively,^{( 4}4. Ministério da Saúde. Sala de Apoio à Gestão Estratégica. [homepage on the Internet]. Brasília: SAGE/SUS [cited 2014 Jun 02]. Available from: http://189.28.128.178/sage/
http://189.28.128.178/sage/... ⁾ indicating limited access to diagnosis and health care, as well as poor treatment adherence.^{( 3}3. Ministério da Saúde. Secretaria de Vigilância em Saúde. Programa Nacional de Controle da Tuberculose. Manual de recomendações para o controle da tuberculose no Brasil. Brasília: Ministério da Saúde; 2011. ⁾

Given that pulmonary tuberculosis (PTB) is transmissible, tuberculosis control measures should prioritize PTB over other clinical presentations. In addition, PTB is a preventable cause of death, meaning that deaths from PTB can be prevented by appropriate health promotion, protection, and recovery measures implemented in local health care systems.^{( 5}5. Malta DC, França E, Abreu DX, Oliveira H, Monteiro RA, Sardinha LMV, et al. Atualização da lista de causas de mortes evitáveis (5 a 74 anos de idade) por intervenções do Sistema Único de Saúde do Brasil. Epidemiol Serv Saúde. 2011;20(3):409-12. http://dx.doi.org/10.5123/S1679-49742011000300016
https://doi.org/10.5123/S1679-4974201100... ⁾

Studies of deaths from tuberculosis currently represent an important method for understanding the difficulties that health care systems face in controlling the disease and for identifying the most vulnerable groups. However, a literature review has shown that few studies have examined the spatial distribution of deaths from tuberculosis in the country.^{( 6}6. Mota FF, Vieira-da-Silva LM, Paim JS, Costa Mda C. Distribuição espacial da mortalidade por tuberculose em Salvador, Bahia, Brasil. Cad. Saude Publica. 2003;19(4); 915-22. http://dx.doi.org/10.1590/S0102-311X2003000400014
https://doi.org/10.1590/S0102-311X200300... ⁾

Although access to health care has improved in Brazil, an optimal level of equity has yet to be achieved, inequality resulting in health outcomes that are not always fair or acceptable,^{( 7}7. Organização Pan-Americana da Saúde. Health in the Americas: 2012 Edition. Regional overview and country profiles. Washington: OPAS; 2012. ⁾ e.g., deaths from PTB. Therefore, the present study is evidently relevant, given that it can aid in strengthening health systems and services so that the problem of tuberculosis can be faced.

The objectives of the present study were to characterize, on the basis of sociodemographic and operational variables, deaths from PTB in the city of São Luís in the 2008-2012 period; to determine whether there were differences between the PTB deaths that were confirmed by autopsy and those that were not in terms of the study variables; and to determine the geographic areas of the city in which PTB deaths were most common.

Methods

This was an exploratory ecological study conducted in the city of São Luís (Figure 1), which is located in northern Maranhão. The metropolitan area of São Luís comprises the municipalities of Paço do Lumiar, Raposa, and São José de Ribamar, among others. Specifically, the city of São Luís is located on the western portion of São Luís Island, between S02°28'21" W44°07'49" and S02°39'34" W44°20'59"; it has an area of 834.780 km² and a population of 1,014,837 inhabitants.^{( 8}8. Instituto Brasileiro de Geografia e Estatística [homepage on the Internet]. Brasília: IBGE [cited 2013 Dez 27]. Censo Demográfico 2010. Available from: http://cidades.ibge.gov.br/xtras/perfil.php?lang=&codmun=211130&search=maranhao|sao-luis
http://cidades.ibge.gov.br/xtras/perfil.... ⁾

The study population consisted of individuals whose underlying cause of death was PTB (International Classification of Diseases, 10th revision-ICD-10-codes A15.0-A15.3 and A16.0-A16.2) and who resided in the urban area of São Luís between 2008 and 2012.

Data were collected in July of 2013 from the Sistema de Informação sobre Mortalidade (SIM, Brazilian National Mortality Database) of the São Luís Municipal Department of Health Office for Epidemiological and Health Surveillance. Individuals who died in São Luís but resided elsewhere were excluded, as were those who died from clinical presentations other than PTB.

The variables of interest were obtained from death certificates and included sociodemographic characteristics (including age, gender, skin color/ethnicity, marital status, level of education, and occupation) and operational variables (including place of death, medical care prior to death, autopsy, underlying cause of death, and physician completing the death certificate).

After having analyzed the consistency of the data collected, we converted the data to STATISTICA, version 10.0 (StatSoft Inc., Tulsa, OK, USA), and the variables were regrouped and analyzed. Regarding the variable age, the individuals who died from PTB in São Luís were categorized on the basis of the median age, being therefore classified by age above or below the median.

Using the same program, we subsequently performed a bivariate analysis cross-referencing the independent variables (sociodemographic and operational variables) with the dependent variable "death confirmed by autopsy" (yes, no). After that, we used the chi-square test for analysis of proportions. The probability of a type I error was set at 5%. At that point, death certificates that were incomplete or on which the cause of death was recorded as unknown were excluded from the study.

In order to determine the areas in which deaths from PTB were most common, we used geocoding (with the freeware TerraView, version 4.2.2, developed by the Brazilian National Institute for Space Research), by standardizing the addresses of the individuals residing in the urban area of São Luís and comparing the addresses with a street segment digital map (StreetBase^(r); Imagem, São José dos Campos, Brazil) in World Geodetic System 1984 latitude and longitude, available in shapefile format.

We subsequently performed kernel density estimation, which consists of an exploratory interpolation generating a density surface for visual identification of "hotspots"; the points within a region of influence are counted and then weighted by the distance between each point and the location of interest.^{( 9}9. Camara G, Carvalho MS. Análise espacial de eventos. In: Druck S, Carvalho MS, Câmara G, Monteiro AV, editors. Análise espacial de dados geográficos. Brasília: EMBRAPA; 2004. ⁾

A radius of 1,000 m being taken into consideration, a thematic map of the distribution of PTB deaths by home address was generated in ArcGIS software, version 10.2 (Esri, Redlands, CA, USA).

The present study was approved by the Research Ethics Committee of the University of São Paulo at Ribeirão Preto School of Nursing (Ruling no. 259,935, May 8, 2013).

Results

We identified 221 deaths from tuberculosis. Of those, 193 were related to PTB. Of those, 190 (98.44%) were deaths of individuals whose death certificates listed PTB as the underlying cause of death, without mention of bacteriological or histological confirmation (ICD-10 16.2); 1 (0.52%) was the death of an individual whose death certificate read "pulmonary tuberculosis, no bacteriological or histological examination having been performed" (ICD-10 A16.1); 1 (0.52%) was the death of an individual whose death certificate read "pulmonary tuberculosis confirmed by histological examination" (ICD-10 A15.2); and 1 (0.52%) was the death of an individual whose death certificate read "pulmonary tuberculosis confirmed by sputum smear microscopy, with or without culture" (ICD-10 A15.0).

The median age of the individuals who died in São Luís was 52 years, the youngest being 16 years old and the oldest being 93 years old.

Table 1 shows the sociodemographic and operational characteristics of the individuals who died from PTB in São Luis, the results being presented in decreasing order of frequency. Most of the individuals who died were male (n = 142; 73.60%), Mulatto (n = 133; 68.91%), and single (n = 102; 53.13%). A greater proportion of the individuals had had 9 years of schooling (n = 64; 33.16%), and most were housekeepers (n = 79; 40.93%). With regard to the operational variables, most of the deaths occurred in a hospital (n = 143; 74.08%), and most of the individuals received medical care prior to death (n = 162; 83.94%, including those requiring hospitalization). Most of the deaths were certified by a substitute physician (n = 70; 36.27%). Most of the individuals who died did not undergo an autopsy (n = 108; 55.95%). Of the death certificates, 31 (16.05%) had no information as to whether an autopsy had been performed.

Table 2 shows the distribution of PTB deaths (an autopsy having been performed or not) according to the sociodemographic and operational characteristics. The variable skin color/ethnicity was found to be significantly associated with an autopsy having been performed (p = 0.003). An autopsy having been performed was also found to be significantly associated with death outside the hospital, e.g., at home or in the street (p = 0.001). Likewise, the variable "not having received medical care prior to death" was found to be significantly associated with an autopsy having been performed (p = 0.001).

An autopsy having been performed was found to be significantly associated with death certified by the Mortality Surveillance System (p = 0.001). Most of the deaths after which an autopsy was not performed were certified by an attending physician or by a substitute physician (p = 0.001).

During the study period, 183 deaths from PTB (95%) were geocoded. Of the cases that were not geocoded, 4 (2%) had incomplete addresses in the SIM and 2 (1%) lived in the rural area of São Luís.

Figure 2 shows a thematic map of the distribution of PTB deaths in the urban area of São Luís by home address. The map highlights areas in which mortality rates were highest (deaths per km²⁾. These areas are represented by darker shades, denoting a heterogeneous spatial distribution of PTB deaths in São Luís during the study period.

The hotspots for PTB deaths are concentrated in the districts of Anjo da Guarda, Liberdade, João de Deus, Bequimão, Cidade Operária, Coroadinho, Monte Castelo, and Centro (city center), the density ranging from 2.53 deaths/km² to 8.12 deaths/km².

Discussion

Although PTB is preventable, curable, and easily diagnosed-and although the Brazilian Unified Health Care System guarantees universal access to tuberculosis treatment, thus facilitating (at least to a certain extent) access to health services-4,500 people in Brazil die from the disease.^{( 10}10. de Oliveira GP1, Pinheiro RS, Coeli CM, Barreira D, Codenotti SB. Mortality information system for identifying underreported cases of tuberculosis in Brazil. Rev Bras Epidemiol. 2012;15(3):468-77. ⁾ The present study focused on deaths from PTB, which, as evidenced by the literature, is highly lethal and the main transmissible form of tuberculosis.^{( 11}11. Larroque MM. Mortalidade por tuberculose em municípios prioritários do estado de Mato Grosso do Sul [dissertation]. Campo Grande: Universidade Federal do Mato Grosso do Sul; 2011. ⁾

The number of deaths from PTB in São Luís was relatively high in comparison with the number of deaths from PTB in other capitals of northeastern Brazil and the country as a whole.^{( 4}4. Ministério da Saúde. Sala de Apoio à Gestão Estratégica. [homepage on the Internet]. Brasília: SAGE/SUS [cited 2014 Jun 02]. Available from: http://189.28.128.178/sage/
http://189.28.128.178/sage/... ⁾ We found a significant number of deaths whose underlying cause was recorded as ICD-10 code 16.2, which is consistent with studies conducted in the city of Campo Grande^{( 12}12. Espindola LCD. Estudo da mortalidade por tuberculose em Campo Grande - MS, 2001 a 2008 [dissertation]. Rio de Janeiro: Escola Nacional de Saúde Pública; 2010. ⁾ and in the state of Rio de Janeiro, Brazil.^{( 13}13. Selig L, Belo M, Cunha AJ, Teixeira EG, Brito R, Luna AL, et al. Óbitos atribuídos à tuberculose no Estado do Rio de Janeiro. J Pneumol. 2004;30(4):417-24. ⁾

This result indicates a major challenge for health care systems, i.e., a critical issue to be overcome; there is a possibility of false-positive results among the deaths whose cause was certified as PTB, given that most of the death certificates had no mention of bacteriological or histological confirmation. This might be due to the fact that the information was not recorded^{( 10}10. de Oliveira GP1, Pinheiro RS, Coeli CM, Barreira D, Codenotti SB. Mortality information system for identifying underreported cases of tuberculosis in Brazil. Rev Bras Epidemiol. 2012;15(3):468-77. ⁾ or to the fact that sputum smear microscopy is not prioritized in the hospital setting.^{( 14}14. Cecilio HP, Molena-Fernandes CA, Mathias TA, Marcon SS. Perfil das internações e óbitos hospitalares por tuberculose. Acta Paul Enferm. 2013;26(3):250-55. http://dx.doi.org/10.1590/S0103-21002013000300008
https://doi.org/10.1590/S0103-2100201300... ⁾ This may have been a source of bias in the present study.

With regard to the sociodemographic characteristics of the individuals who died from PTB in São Luís, the results of the present study are consistent with those of other studies,^{( 13}13. Selig L, Belo M, Cunha AJ, Teixeira EG, Brito R, Luna AL, et al. Óbitos atribuídos à tuberculose no Estado do Rio de Janeiro. J Pneumol. 2004;30(4):417-24.

14. Cecilio HP, Molena-Fernandes CA, Mathias TA, Marcon SS. Perfil das internações e óbitos hospitalares por tuberculose. Acta Paul Enferm. 2013;26(3):250-55. http://dx.doi.org/10.1590/S0103-21002013000300008
https://doi.org/10.1590/S0103-2100201300...

15. Oliveira GP, Torrens AW, Bartholomay P, Barreira D. Tuberculosis in Brazil: last tem years analysis - 2001-2010. Braz J Infect Dis. 2013;17(2):218-33. http://dx.doi.org/10.1016/j.bjid.2013.01.005
https://doi.org/10.1016/j.bjid.2013.01.0...

16. Augusto CJ, Carvalho WS, Gonçalves AD, Ceccato MG, Miranda SS. Characteristics of tuberculosis in the state of Minas Gerais, Brazil: 2002-2009. J Bras Pneumol. 2013;39(3):357-64. http://dx.doi.org/10.1590/S1806-37132013000300013
https://doi.org/10.1590/S1806-3713201300...

17. Moreira CM, Zandonade E, Reynaldo D, Maciel EL. Tuberculosis-related mortality in the state of Espírito Santo, Brazil, 1985-2004. J Bras Pneumol. 2008;34(8):601-6.

18. Lindoso AA, Waldman EA, Komatsu NK, Figueiredo SM, Taniguchi M, Rodrigues LC. Profile of tuberculosis patients progressing to death, city of São Paulo, Brazil, 2002. Rev Saude Publica 2008;42(5):805-12. http://dx.doi.org/10.1590/S0034-89102008000500004
https://doi.org/10.1590/S0034-8910200800...

19. Hino P, da Costa-Junior ML, Sassaki CM, Oliveira MF, Villa TC, dos Santos CB. Time series of tuberculosis mortality in Brazil (1980-2001). Rev Lat Am Enfermagem. 2007;15(5):936-41. http://dx.doi.org/10.1590/S0104-11692007000500009
https://doi.org/10.1590/S0104-1169200700...

20. Kayigamba FR, Bakker MI, Mugisha V, De Naeyer L, Gasana M, Cobelens F, et al. Adherence to tuberculosis treatment, sputum smear conversion and mortality: a retrospective cohort study in 48 Rwandan clinics. PLoS One. 2013;8(9):e73501. http://dx.doi.org/10.1371/journal.pone.0073501
https://doi.org/10.1371/journal.pone.007...

21. Alavi-Naini R, Moghtaderi A, Metanat M, Mohammadi M, Zabetian M. Factors associated with mortality in tuberculosis patients. J Res Med Sci. 2013;18(1):52-5.

22. Álvarez JL, Kunst AE, Leinsalu M, Bopp M, Strand BH, Menvielle G, et al. Educational inequalities in tuberculosis mortality in sixteen European populations. Int J Tuberc Lung Dis. 2011;15(11):1461-7. http://dx.doi.org/10.5588/ijtld.10.0252
https://doi.org/10.5588/ijtld.10.0252... ^{- 23}23. Jung RS, Bennion JR, Sorvillo F, Bellomy A. Trends in tuberculosis mortality in the United States, 1990-2006: a population-based case-control study. Public Health Rep. 2010;125(3):389-97 ⁾ in which the proportion of deaths was found to be higher among males, a finding that corroborates reports that PTB is less common in females than in males. This might be due to the fact that males participate in the workforce more than do females and the fact that males use health services less than do females, as well as to the fact that the prevalence of HIV infection, alcoholism, and drug abuse is higher in males.^{( 24}24. Holmes CB, Hausler H, Nunn P. A review of sex differences in the epidemiology of tuberculosis. Int J Tuberc Lung Dis. 2008;2(2):96-104. ^{, 25}25. Façanha MC. Evolução da mortalidade por tuberculose em Fortaleza (CE), entre 1980 e 2001. J Bras Pneumol. 2006;32(6):553-8. http://dx.doi.org/10.1590/S1806-37132006000600013
https://doi.org/10.1590/S1806-3713200600... ⁾ However, these variables were not investigated in our study.

The present study focused exclusively on deaths whose underlying cause was reported as being PTB, which is a preventable cause of death.^{( 5}5. Malta DC, França E, Abreu DX, Oliveira H, Monteiro RA, Sardinha LMV, et al. Atualização da lista de causas de mortes evitáveis (5 a 74 anos de idade) por intervenções do Sistema Único de Saúde do Brasil. Epidemiol Serv Saúde. 2011;20(3):409-12. http://dx.doi.org/10.5123/S1679-49742011000300016
https://doi.org/10.5123/S1679-4974201100... ⁾ Therefore, deaths associated with but not caused by tuberculosis were not included.

With regard to age and marital status, deaths from PTB were found to be more common among individuals who were 52 years of age or younger and among those who were single, a finding that is consistent with those of studies conducted in the Brazilian states of Mato Grosso do Sul,^{( 12}12. Espindola LCD. Estudo da mortalidade por tuberculose em Campo Grande - MS, 2001 a 2008 [dissertation]. Rio de Janeiro: Escola Nacional de Saúde Pública; 2010. ⁾ Minas Gerais,^{( 16}16. Augusto CJ, Carvalho WS, Gonçalves AD, Ceccato MG, Miranda SS. Characteristics of tuberculosis in the state of Minas Gerais, Brazil: 2002-2009. J Bras Pneumol. 2013;39(3):357-64. http://dx.doi.org/10.1590/S1806-37132013000300013
https://doi.org/10.1590/S1806-3713201300... ⁾ Ceará,^{( 25}25. Façanha MC. Evolução da mortalidade por tuberculose em Fortaleza (CE), entre 1980 e 2001. J Bras Pneumol. 2006;32(6):553-8. http://dx.doi.org/10.1590/S1806-37132006000600013
https://doi.org/10.1590/S1806-3713200600... ⁾ and Rio de Janeiro,^{( 13}13. Selig L, Belo M, Cunha AJ, Teixeira EG, Brito R, Luna AL, et al. Óbitos atribuídos à tuberculose no Estado do Rio de Janeiro. J Pneumol. 2004;30(4):417-24. ⁾ as well as in Africa.^{( 20}20. Kayigamba FR, Bakker MI, Mugisha V, De Naeyer L, Gasana M, Cobelens F, et al. Adherence to tuberculosis treatment, sputum smear conversion and mortality: a retrospective cohort study in 48 Rwandan clinics. PLoS One. 2013;8(9):e73501. http://dx.doi.org/10.1371/journal.pone.0073501
https://doi.org/10.1371/journal.pone.007... ⁾

With regard to skin color/ethnicity, the results of the present study are consistent with those of a study conducted in the city of Campo Grande^{( 12}12. Espindola LCD. Estudo da mortalidade por tuberculose em Campo Grande - MS, 2001 a 2008 [dissertation]. Rio de Janeiro: Escola Nacional de Saúde Pública; 2010. ⁾ and with those of a study conducted in other priority cities for tuberculosis control in the state of Mato Grosso do Sul,^{( 11}11. Larroque MM. Mortalidade por tuberculose em municípios prioritários do estado de Mato Grosso do Sul [dissertation]. Campo Grande: Universidade Federal do Mato Grosso do Sul; 2011. ⁾ in which most of the deaths occurred in individuals who were Mulatto,^{( 3}3. Ministério da Saúde. Secretaria de Vigilância em Saúde. Programa Nacional de Controle da Tuberculose. Manual de recomendações para o controle da tuberculose no Brasil. Brasília: Ministério da Saúde; 2011. ^{, 15}15. Oliveira GP, Torrens AW, Bartholomay P, Barreira D. Tuberculosis in Brazil: last tem years analysis - 2001-2010. Braz J Infect Dis. 2013;17(2):218-33. http://dx.doi.org/10.1016/j.bjid.2013.01.005
https://doi.org/10.1016/j.bjid.2013.01.0... ⁾ but inconsistent with those of a study conducted in the city of São Paulo, Brazil,^{( 18}18. Lindoso AA, Waldman EA, Komatsu NK, Figueiredo SM, Taniguchi M, Rodrigues LC. Profile of tuberculosis patients progressing to death, city of São Paulo, Brazil, 2002. Rev Saude Publica 2008;42(5):805-12. http://dx.doi.org/10.1590/S0034-89102008000500004
https://doi.org/10.1590/S0034-8910200800... ⁾ in which most of the deaths occurred in individuals who were White.

With regard to the level of education and occupation, the findings of the present study are consistent with those of other studies,^{( 11}11. Larroque MM. Mortalidade por tuberculose em municípios prioritários do estado de Mato Grosso do Sul [dissertation]. Campo Grande: Universidade Federal do Mato Grosso do Sul; 2011. ^{, 12}12. Espindola LCD. Estudo da mortalidade por tuberculose em Campo Grande - MS, 2001 a 2008 [dissertation]. Rio de Janeiro: Escola Nacional de Saúde Pública; 2010. ^{, 16}16. Augusto CJ, Carvalho WS, Gonçalves AD, Ceccato MG, Miranda SS. Characteristics of tuberculosis in the state of Minas Gerais, Brazil: 2002-2009. J Bras Pneumol. 2013;39(3):357-64. http://dx.doi.org/10.1590/S1806-37132013000300013
https://doi.org/10.1590/S1806-3713201300... ^{, 18}18. Lindoso AA, Waldman EA, Komatsu NK, Figueiredo SM, Taniguchi M, Rodrigues LC. Profile of tuberculosis patients progressing to death, city of São Paulo, Brazil, 2002. Rev Saude Publica 2008;42(5):805-12. http://dx.doi.org/10.1590/S0034-89102008000500004
https://doi.org/10.1590/S0034-8910200800... ⁾ in which illiteracy or a low level of education and low-income jobs were reported as risk factors for PTB; all of the above are risk factors for PTB and are responsible for a higher incidence of the disease, as well as contributing to treatment nonadherence and death.^{( 10}10. de Oliveira GP1, Pinheiro RS, Coeli CM, Barreira D, Codenotti SB. Mortality information system for identifying underreported cases of tuberculosis in Brazil. Rev Bras Epidemiol. 2012;15(3):468-77. ⁾

With regard to operational characteristics, approximately 75% of the deaths analyzed in the present study occurred in a hospital. This is consistent with studies conducted in the Brazilian states of Rio de Janeiro^{( 13}13. Selig L, Belo M, Cunha AJ, Teixeira EG, Brito R, Luna AL, et al. Óbitos atribuídos à tuberculose no Estado do Rio de Janeiro. J Pneumol. 2004;30(4):417-24. ⁾ and Mato Grosso do Sul,^{( 11}11. Larroque MM. Mortalidade por tuberculose em municípios prioritários do estado de Mato Grosso do Sul [dissertation]. Campo Grande: Universidade Federal do Mato Grosso do Sul; 2011. ⁾ as well as with a study conducted in the city of São Paulo,^{( 18}18. Lindoso AA, Waldman EA, Komatsu NK, Figueiredo SM, Taniguchi M, Rodrigues LC. Profile of tuberculosis patients progressing to death, city of São Paulo, Brazil, 2002. Rev Saude Publica 2008;42(5):805-12. http://dx.doi.org/10.1590/S0034-89102008000500004
https://doi.org/10.1590/S0034-8910200800... ⁾ all of which showed rates higher than 80%. The death of patients hospitalized for tuberculosis suggests that the primary health care system faces difficulties in management, in providing access to diagnostic resources, in case management, and in referring patients to other health services.^{( 26}26. Selig L, Kritski AL, Cascão AM, Braga JU, Trajman A, de Carvalho RM. Proposal for tuberculosis death surveillance in information systems. Rev Saude Publica. 2010;44(6):1072-8. http://dx.doi.org/10.1590/S0034-89102010000600012
https://doi.org/10.1590/S0034-8910201000... ⁾ Another possible explanation is treatment nonadherence, which predicts the development of multidrug-resistant tuberculosis and worsening of the disease; treatment discontinuation can lead to hospitalization when individuals seek health services.^{( 20}20. Kayigamba FR, Bakker MI, Mugisha V, De Naeyer L, Gasana M, Cobelens F, et al. Adherence to tuberculosis treatment, sputum smear conversion and mortality: a retrospective cohort study in 48 Rwandan clinics. PLoS One. 2013;8(9):e73501. http://dx.doi.org/10.1371/journal.pone.0073501
https://doi.org/10.1371/journal.pone.007... ⁾

Analysis of the mortality rates for communicable diseases reflects the efficacy of prevention and control measures, as well as the quality of diagnosis and medical care, being of limited use when there is a high proportion of deaths without medical care or of deaths due to ill-defined causes.^{( 5}5. Malta DC, França E, Abreu DX, Oliveira H, Monteiro RA, Sardinha LMV, et al. Atualização da lista de causas de mortes evitáveis (5 a 74 anos de idade) por intervenções do Sistema Único de Saúde do Brasil. Epidemiol Serv Saúde. 2011;20(3):409-12. http://dx.doi.org/10.5123/S1679-49742011000300016
https://doi.org/10.5123/S1679-4974201100... ⁾ In the present study, approximately 15% of all deaths from PTB occurred without medical care, and, in most cases, no autopsy was performed. However, the records showed that 26% of the deaths were certified by the Mortality Surveillance System.

Although an autopsy allows the diagnosis of diseases that were not suspected or elucidated before death, it should be used judiciously.^{( 27}27. Joaquim AF, Carandina L, Defaveri J. Tuberculose em necropsias realizadas no Serviço de Anatomia Patológica da Faculdade de Medicina de Botucatu. J Bras Patol Med Lab. 2006;42(3):193-200. http://dx.doi.org/10.1590/S1676-24442006000300008
https://doi.org/10.1590/S1676-2444200600... ⁾ In the present study, we found a significant association between not having received medical care prior to death and an autopsy having been performed. We also found that an autopsy was not performed in only 3 (1.85%) of the individuals who died without having received medical attention prior to death.

The findings described above show the importance of a health care system that allows equity of access to health care, thus allowing tuberculosis cases to be diagnosed in a timely manner, especially in the primary health care setting, in order to prevent deaths. Autopsy confirmation of tuberculosis clearly shows the inadequacy of health care systems in reducing social inequalities in health.^{( 28}28. Curtis S. Health and Inequality: Geographical Perspectives. London: Sage Publications; 2009. ⁾

As is the case with other endemic diseases, tuberculosis is strongly influenced by the environment, and the disease has long been associated with a low socioeconomic status; therefore, there is a need to study the disease and combat it, its spatial distribution being taken into account.^{( 19}19. Hino P, da Costa-Junior ML, Sassaki CM, Oliveira MF, Villa TC, dos Santos CB. Time series of tuberculosis mortality in Brazil (1980-2001). Rev Lat Am Enfermagem. 2007;15(5):936-41. http://dx.doi.org/10.1590/S0104-11692007000500009
https://doi.org/10.1590/S0104-1169200700... ⁾

The spatial distribution of deaths from PTB (with emphasis on the heterogeneous distribution observed in the city of São Luís) should be considered the starting point of a process of investigation and surveillance that can lead to the identification of problem areas and failures of the health care system to provide health care for the target population.

Our dot density map shows the sites where PTB deaths per km² are most likely to occur, showing the spatial distribution of hotspots and revealing geographic inequalities related to events occurring in the city. This effectively contributed to the identification of geographic areas in which PTB deaths occurred and in which preventive/curative measures are therefore required in order to reorganize health services to meet the health needs of the population.

The areas in which the density of deaths per km² was highest coincided with areas in which housing conditions and home quality were classified as poor in a study conducted in São Luís.^{( 29}29. Pereira PR, Cutrim Jr V, Rodrigues ZM. Desigualdades intraurbanas dos indicadores de saneamento em São Luís - MA [monograph on the Internet]. São Luís: Universidade Federal do Maranhão; 2013 [cited 2014 Jun 02]. [Adobe Acrobat document, 10p.]. Available from: http://www.joinpp.ufma.br/jornadas/joinpp2013/JornadaEixo2013/anais-eixo13-questaourbanaegestaodascidades/desigualdadesintraurbanasdosindicadoresdesaneamento.pdf
http://www.joinpp.ufma.br/jornadas/joinp... ⁾ These areas include the following districts: Coroadinho; Ilhinha; Turu; Anjo da Guarda; Vila Nova; Vila Luizão; Vila Embratel; Liberdade; Sá Viana; Divineia; and parts of Cidade Operária. The aforementioned study^{( 29}29. Pereira PR, Cutrim Jr V, Rodrigues ZM. Desigualdades intraurbanas dos indicadores de saneamento em São Luís - MA [monograph on the Internet]. São Luís: Universidade Federal do Maranhão; 2013 [cited 2014 Jun 02]. [Adobe Acrobat document, 10p.]. Available from: http://www.joinpp.ufma.br/jornadas/joinpp2013/JornadaEixo2013/anais-eixo13-questaourbanaegestaodascidades/desigualdadesintraurbanasdosindicadoresdesaneamento.pdf
http://www.joinpp.ufma.br/jornadas/joinp... ⁾ showed that unplanned urban growth in São Luís resulted in subnormal areas, classified as lacking public services (most of which are essential).

Curtis^{( 28}28. Curtis S. Health and Inequality: Geographical Perspectives. London: Sage Publications; 2009. ⁾ reported that health services tend to be ineffective and insufficient in areas in which housing and sanitation are poor, i.e., areas in which social inequality is high. Although we did not investigate the health care facilities in the aforementioned areas, we have reasons to assume that there is a relationship among the deaths from PTB, the areas in which those who died resided, and the health systems and services in those areas.

In Brazil, high social inequality in access to health resources, education, income distribution, sanitation, and other aspects related to the standard of living of the population contributes to differences across social strata regarding the risk of illness^{( 19}19. Hino P, da Costa-Junior ML, Sassaki CM, Oliveira MF, Villa TC, dos Santos CB. Time series of tuberculosis mortality in Brazil (1980-2001). Rev Lat Am Enfermagem. 2007;15(5):936-41. http://dx.doi.org/10.1590/S0104-11692007000500009
https://doi.org/10.1590/S0104-1169200700... ⁾ and, consequently, the risk of death.

As demonstrated in the present study, the identification of priority areas for tuberculosis control could promote public health policies aimed at reducing health inequalities, allowing the optimization of resources and teams for PTB control in the study setting, as well as informing decisions regarding the selection of strategies and specific interventions targeting the most vulnerable populations.

For diseases for which reporting is mandatory, such as PTB, the use of data available in health information systems allows the monitoring of the problem, aids in the identification of relevant issues, and encourages the search for new interventions for disease control.^{( 26}26. Selig L, Kritski AL, Cascão AM, Braga JU, Trajman A, de Carvalho RM. Proposal for tuberculosis death surveillance in information systems. Rev Saude Publica. 2010;44(6):1072-8. http://dx.doi.org/10.1590/S0034-89102010000600012
https://doi.org/10.1590/S0034-8910201000... ⁾ Therefore, the data collected from the SIM allowed us to observe the dynamics and behavior of PTB in São Luís during the study period.

As a source of information for the study of deaths in a given area, the SIM has limitations, including underreporting,^{( 10}10. de Oliveira GP1, Pinheiro RS, Coeli CM, Barreira D, Codenotti SB. Mortality information system for identifying underreported cases of tuberculosis in Brazil. Rev Bras Epidemiol. 2012;15(3):468-77. ^{, 12}12. Espindola LCD. Estudo da mortalidade por tuberculose em Campo Grande - MS, 2001 a 2008 [dissertation]. Rio de Janeiro: Escola Nacional de Saúde Pública; 2010. ⁾ which is one of the consequences of unequal access to health care. Another limitation is missing data on death certificates, complete records being important for health management and planning.

The parameters used in the present study, particularly those used in the definition of the radius for kernel density estimation, were chosen on the basis of empirical knowledge and constitute one of the limitations of the present study. Nevertheless, we sought to choose parameters that contributed to the understanding of the study subject.

Other limitations of the present study are due to its observational design, including interference from spurious variables or confounding factors. We should also take into consideration the ecological fallacy, which is the major limitation of ecological studies; therefore the results cannot be considered at the individual level. Finally, the use of secondary data constitutes yet another limitation of the present study; incomplete or missing data may have contributed to an information bias.

The results of the present study can guide health managers and professionals in identifying priority areas for investment in health care, in order to eliminate preventable deaths from tuberculosis. The present study provides an opportunity to reconsider major clinical practice and environmental issues, as well as allowing us to reflect on the efficacy of public policies in reducing health inequalities and providing social protection to the population.

Acknowledgments

We would like to thank the São Luís Municipal Department of Health Office for Epidemiological and Health Surveillance for allowing us to conduct the present study and for providing the data for analysis.

References

Publication Dates

History