Mortality estimations for the indigenous population in Brazil based - - PDF document

1

Mortality estimations for the indigenous population in Brazil based on census data

Marden Barbosa de Campos Gabriel Mendes Borges Ricardo Ventura Santos Bernardo Lanza Queiroz Abstract This paper presents, for the first time, estimations of mortality in the indigenous population in Brazil based on data from the most recent national census (2010), along with a comparison with estimates for non-indigenous population. We use different methods to characterize indigenous households from the census data. The results point to striking differences, showing much higher mortality rates for the indigenous population in all age groups considered. keywords: Mortality, Indigenous populations, Demographic methods, Race/ethnicity

2 Introduction

Indigenous peoples are among the most marginalized social segments in different parts of the world. In Latin America, increasing attention has been paid to the health conditions of indigenous peoples, especially in the context of efforts to reduce socioeconomic disparities. Availability of health statistics is a key point in order to achieve these goals (Coimbra et al. 2013; McSweeney and Arps 2005; Montenegro and Stephens 2006, Coimbra and Santos 2004; Garnelo et al. 2003; Montenegro and Stephens, 2006; Santos et al. 2008). In Brazil, significant advances have been made in recent decades in order to revert the lack of data on indigenous peoples in national statistics. In the 1990s, the category “indigenous” was included as one of the categories in the question about color or race in the decennial national census, resulting in the production

• f important demographic data for formulation, implementation and evaluation of

public policies. An important innovation of the 2010 Brazilian National Census was the inclusion of questions on the occurrence of deaths in the households. Despite the wide recognition of major disparities across ethnic-racial groups in Brazil, analyses focusing on mortality differentials are relatively scarce, in particular those that take into consideration the indigenous segment of the population (Chor and Lima 2005; Matijasevich et al. 2007; Wong et al. 2014). In this scenario, the possibility of using census data to estimate indigenous mortality is of paramount importance. This paper aims to present, for the first time, estimations of mortality in the indigenous population in Brazil based on data from the most recent national census (2010), along with a comparison with estimates for the non-Indigenous population. Methods Death registration systems in Brazil, as in most developing countries, are incomplete, have several data quality limitations and lack detailed information for population subgroups, posing considerable challenge to mortality estimations in the country. Different methods have been proposed to overcome these limitations, for instance by adjusting death counts based on the relationships

3

between the distribution of deaths, the age structure and the rates of change in the populations or directly using questions included in surveys and censuses (Timæus et al., 2013). This paper adopts both approaches and uses information on household mortality recently incorporated into the 2010 Brazilian population census, which is further adjusted by a death distribution method. The 2010 census asked for all households whether someone living in that household had died in the previous 12 months. In the case there was a reported death in the household, it was asked the sex and age of the deceased. These data are analyzed through a simple two-step method, similar to the method previously used by Silva et al (2016). In the first step, evaluation methods are applied to the household mortality records in order to check the quality and to adjust the numbers of reported deaths, allowing for more accurate estimates

• f mortality levels. In this paper, we evaluate the quality of mortality data reported

in the census at the national level. A recent study estimated that the completeness of deaths reported in the census was 80% for females and 85% for males (Queiroz and Sawyer, 2012). In the second step of the method, household mortality records (including age and sex of the deceased) are linked to household information such as race classification of the members of the

• household. This simple method allows the estimation of age-specific mortality

rates by specific sub-population groups in Brazil, which in turn can be used to compute life tables. The two steps method and their caveats are detailed below. Queiroz and Sawyer (2012) showed that the quality of mortality data in the 2010 Census is satisfactory and can be used as a reliable source. A major advantage of census data is that the numerators and denominators derive from the same source. There are, however, important limitations, such as under- enumeration of deaths due to the dissolution of households following the death

• f a resident and the inexistence of survivors to report the death. In Brazil, this is

a particularly relevant issue, as there is a relatively high proportion of single- person households. According to Queiroz and Sawyer (2012), the observed mortality structure is of good quality, except for the elderly. Notwithstanding, these limitations might be partially corrected by the use of demographic methods.

4

Census data also allow the investigation of regional and socioeconomic differentials in mortality, since they provide much more reliable and detailed information on these issues than vital statistics sources. Coverage of Reported Deaths The demographic literature has developed several methods based upon equations of population dynamics to evaluate the coverage of reported deaths relative to populations. The death distribution methods (DDM) are commonly used to estimate adult mortality in a non-stable population. A stable population is

• ne in which the birth and death rates are unchanging over a long period of time.

The DDM methods compare the distribution of deaths by age with the age distribution of the living and provide age patterns of mortality in a defined reference period. There are two major approaches: the General Growth Balance Methods, and the Synthetic Extinct Generation methods. In this paper, we use the General Growth Balance method (GGB), particularly the Benneth and Horiuchi Adjusted variant. This method is derived from the basic demographic balancing equation, which expresses the identity that the growth rate of a population is equal to the difference between its entry rate and exit rate. Equation 1 shows the basic relations of the method (1) Where N’(x) is the number of persons who reach the exact age x in the period, N(x+) is the number of persons at exact age x and over, r(x+) is the population growth rate, k1 and k2 are the relative coverage of the enumerated population in two censuses, C is the degree of completeness of death records

• ver the period, D’(x+) is the observed number of deaths of people with x or more

years of age and t is the interval corresponding to the intercensal period This identity holds for open-ended age segments x+, and in a closed population the only entries are through birthdays at age x. The “birth” rate x+ minus the growth rate x+ thus provides a residual estimate of the death rate x+. If the residual estimate can be calculated from population data from two population censuses and compared to a direct estimate using the recorded

            

                      x N x D C k k k k t x r x N x N ' ln 1 ' '

2 1 2 1 2 1

5

deaths, the completeness of death recording relative to population recording can be estimated. The method has a few strong and key assumptions: the population is closed to migration; the completeness of recording of deaths is constant by age; the completeness of recording of the population is constant by age; and ages of the living and dead are reported without error. The death distribution method has also an important limitation. Since they compare a distribution of deaths to an intercensal population, they estimate intercensal completeness of mortality estimates, and not the completeness at the beginning or end of the intercensal period. This matter is of particular importance when a distribution of deaths comes from data from the latest census or when vital registration is available for recent years. The assumption of the GGB method that the population is closed to migration is also of importance to Brazil and its

• regions. The GGB method uses information on deaths and growth rates

cumulated above a series of ages x. If there is some age x above which net migration is negligible, the performance of the methods above that age will be

• unaffected. The intercept and slope of the GGB method were obtained by
• rthogonal regression to points for the ages 35+ to 75+.

The Bennett and Horiuchi method, SEG, uses specific growth rates by age for converting an age distribution of deaths into an age distribution of a

• population. Once the observed deaths from a given age x in a population is equal

to the population of age x, adjusted by the rate of population growth by age range, we have the deaths of a population of age x+ that provide an estimate of the population on that age x. The extent of death registration coverage is given by the ratio of deaths estimated by the population above the age x and the observed population above the age x. Equation 2 presents the mathematical formalization

• f the method:

(2) where N’(x) is the number of people who reach the exact age x in a population with growth rate r, and D(x) is the number of deaths at age x. In this case, the estimate of deaths under-registration in the period is given by the ratio between the estimated number of people aged x (N’(x)) and the observed number of people aged x (N(x)). The diagnostic plot in this case should present a straight

   

   



 



 x a r a r

e a D x N'

6

line across age, showing constant completeness. A change in the slope over the age groups indicates possible coverage problems of population censuses, problems in the age statement of the living and dead or variation in the quality of the registration / listing of deaths by age group Hill and colleagues (2009) suggest that the combination of GGB and SEG methods may be more robust than the application of the two methods separately. The adjusted method consists of applying GGB to obtain estimates of the change in the population enumeration (k1/k2), and use this ratio to adjust the coverage

• f both census, and then apply the SEG method using the adjusted population

for the coverage of mortality data. Methods Caveat There are a few caveats to this method that should be addressed and taken into account when analyzing the results. We assume the same adjustment factor for all households when using death distribution methods. One could argue that the level of enumeration completeness would be lower for groups in less developed areas and better in more developed areas, but we assume that all groups would report at the same level for methodological simplicity. Additionally, we used only one adjustment factor across all age groups, which assumes the same degree of completeness of death records for all ages. If the coverage of infant mortality records is worse than in the adult age, probably this difference is higher in the lower income groups. If that is the case, this would imply again that our method would be underestimating the mortality differentials across sub-population groups. Lastly, a third, and main, limitation of the method presented here is that we do not have information on the deceased. We only know the age and sex of the deceased and can linked it to the household characteristics. Overall, the limitations of the method would in the worst-case scenario generate more conservative estimates of mortality differentials in life expectancy by sub-national groups. Indigenous at the census

7

In the 1991 Census, the indigenous category was included as an answer

• ption for the color/ race question present in the sample questionnaire, which had

not occurred since the 1960 Census. It is worth mentioning that the question about color or race was included in the universe (sum of the basic questionnaires and the sample) of the 2010 Census, allowing the entire population to be classified according to this variable. This point is particularly relevant for indigenous peoples, since this category represents only a small proportion (less than 0.5%) of the Brazilian population. In respect to estimations of indigenous mortality in censuses, it is important to say that the color or race of the deceased individual was not investigated, and it is necessary to use a methodological strategy to assign a color/race to this individual, a process that will be detailed in the next section. Estimating Mortality Rates In order to establish a proxy for the color or race of the deceased individual, we tested different typologies: (1) at least one indigenous living at the household; (2) all members of households classified as indigenous; (3) person responsible for the household classified as indigenous. Figure 1 shows the unadjusted mortality rates for the different typologies, indicating that mortality estimates are similar regardless the household color/race classification scheme. The only exception is the mortality curve for males when all members of the household are indigenous. Estimates under this definition is noisier, since the population in this group is smaller. We opted to carry out the more detailed analyses for this paper based on the last alternative (3).

8

Figure 1 – Unadjusted Age-Specific Mortality Rates by sex and definition of indigenous household – Brazil, 2010

Source: IBGE, 2010.

Mortality levels were calculated for three age groups (< 5 years old, 5-14 years old and 15-59 years old), providing probabilities of dying estimations (5q0,

10q5, 45q15) for indigenous and non-indigenous households using the adjustment

levels proposed by Queiroz and Sawyer (2012). No estimates were generated for age groups beyond age 60 because of the previously mentioned limitations that bias mortality rates for this group (Queiroz and Sawyer 2012). Preliminary results There are significant differences between indigenous and non-indigenous population in Brazil in all age groups investigated, more pronounced for females (Table 1 and Figure 1). The most pronounced differentials are observed in the 0-4 age group, with indigenous children presenting mortality rates almost twice as high as the rates for the non-indigenous. Adult mortality among indigenous women is 28% higher than among non-indigenous. The difference is less pronounced in adult indigenous men, yet being 19% higher.

9

Table 1 – Specific Mortality Rates by sex, category and selected age groups – indigenous and non-indigenous – Brazil, 2010

Source: IBGE, 2010.

Figure 1 shows the differences in the rates by age, illustrating the differences found in the selected indicators. Figure 1 - Specific Mortality Rates by sex – indigenous and non-indigenous – Brazil, 2010

Source: IBGE, 2010.

Preliminary conclusions Since the 1990s, there has been a significant increase in the number of studies on the health profile of indigenous populations in Brazil, often showing that they are worse off than the non-indigenous (see reviews in Coimbra et al. 2013). Most of the investigations carried out so far have focused on topics related to nutrition and infectious diseases. This is the first investigation that, based on national census data, presents findings on mortality levels for children and adults comparing indigenous to non-indigenous segments in Brazil. The results point to striking differences, showing much higher mortality rates for the indigenous population in all age groups considered.

Selected age group Indigenous

• female

Non-Indigenous

• female

Indigenous - male Non-Indigenous

• male

5q0 (‰) 34.6 17.4 38.0 19.9 10q5 (‰) 7.0 3.7 8.3 4.7 45q15 (‰) 147.0 114.8 255.1 214.7

10

References COIMBRA JR. CEA, SANTOS RV. (2004) Emerging health needs and epidemiologic research in indigenous peoples in Brazil. In: Salzano FM, Hurtado M, editors. Lost paradise and the ethics of research and publication. Oxford: Oxford University Press; p. 89-109. COIMBRA CEA, SANTOS RV, WELCH JR, CARDOSO AM, SOUZA MC, GARNELO L, RASSI E, FOLLER MJ, HORTA BL. (2013) The First National Survey of Indigenous People´s health and nutrition in Brazil: rationale, methodology and overview of results. BMC Public Health, 13:52. GARNELO L, MACEDO G, BRANDÃO LC. (2003) Os povos indígenas e a construção da política de saúde no Brasil. Brasília: Organização Pan- Americana da Saúde. Metodologia do Censo Demográfico 2000. Rio de Janeiro: IBGE. HILL, KENNETH, DANZHEN YOU, AND YOONJOUNG CHOI. "Death distribution methods for estimating adult mortality: sensitivity analysis with simulated data errors." Demographic Research 21 (2009): 235-254. MCSWEENEY K & ARPS SA. (2005) Demographic turnaround: the rapid growth of indigenous populations in lowland Latin America. Lat Am Res Rev, 40:3-29. MONTENEGRO RA, STEPHENS C. (2006) Indigenous health. 2 – Indigenous health in Latin America and the Caribbean.Lancet; 367:1859-69. QUEIROZ, B. L.; SAWYER, D. (2012) O que os dados de mortalidade do censo de 2010 podem nos dizer? Revista Brasileira de Estudos de População (Impresso), v. 29, p. 225-238. SANTOS RV, CARDOSO AM, GARNELO L, COIMBRA JR. CEA, Chaves

• MBG. (2008) Saúde dos povos indígenas e políticas públicas no Brasil. In:

Giovanella L, Escorel S, Lobato L, Noronha JC, Carvalho AI, organizadores. Políticas e sistema de saúde no Brasil. Rio de Janeiro: Editora Fiocruz. p. 1035- 56.

11

SILVA, L. E. D., FREIRE, F. H. M. D. A., & PEREIRA, R. H. M. (2016). Diferenciais de mortalidade por escolaridade da população adulta brasileira, em

• 2010. Cadernos de Saúde Pública, 32(4).

TIMÆUS IM, RE DORRINGTON AND KH HILL. 2013. "Introduction to adult mortality analysis". In Moultrie TA, RE Dorrington, AG Hill, K Hill, IM Timæus and B Zaba (eds). Tools for Demographic Estimation. Paris: International Union for the Scientific Study

• f

Population. http://demographicestimation.iussp.org/content/introduction-adult-mortality-

• analysis. Accessed 26/09/2017.

WONG, L.L.R ; BARROS, Juliana V. S. ; SANTOS, W. (2014). Mortalidad infantil e infanto juvenil en Brasil según sexo y color de la piel. In: Laura L. Rodriguez Wong, Jhon Antón Sánchez. (Org.). Situación de la población afro- descendiente e indígena en América Latina ? puntos de reflexión para el debate sobre Cairo + 20. 1ed.Belo Horizonte: ALAP, v. 1, p. 79-97.