INFANT AND YOUTH SURVIVAL INDICATORS DISAGGREGATED BY DISTRICT INCOME. SÃO PAULO CITY, BRAZIL

Original article

Running head: Infant and youth survival indicators

Contexto: Este estudo mostra a existente disparidade entre distritos selecionados da cidade de São Paulo considerando-se o status socio-econômico, Mortalidade Infantil e a Mortalidade por Causas Externas.

Objetivo: Atualizar e disseminar informações sobre os problemas urbanos em São Paulo, através da análise dos dados de mortalidade e renda, bem como apresentação de indicadores selecionados nos contrastantes distritos da cidade.

Procedimentos: Realizou-se coleta de dados sobre população, mortalidade e renda da cidade de São Paulo previamente desagregados por distritos. Escolheu-se para comparação, quatro Indicadores de Saúde: Índice de Mortalidade Infantil, Índice de Mortalidade Proporcional em crianças menores de cinco anos, Índice de Mortalidade Específica por Causas Externas em homens com idade entre 15 e 29 anos e Índice de Mortalidade Proporcional por Causas Externas em homens com idade entre 15 e 29 anos. Por fim, examinou-se a relação entre renda e os Indicadores de Saúde para, seis distritos, representando a população mais carente, e os outros seis distritos, a camada mais opulenta da sociedade.

Resultados: Os Índices de Mortalidade Infantil e os Índices de Mortalidade por Causas Externas nos distritos menos favorecidos foram de 3 a 13 vezes maiores e de 6 a 24 vezes maiores, quando comparados com os distritos mais ricos, respectivamente. Em três dos seis distritos selecionados com baixa renda, a maior causa de morte verificada para esta população foi por Causas Externas, superando os índices de morte por Doenças Cardiovasculares. Quatro dos seis distritos menos favorecidos apresentaram Índice de Mortalidade Proporcional entre jovens do sexo masculino atribuídos à Causas Externas, excedendo os Índices de Mortalidade Proporcional em crianças menores de cinco anos de idade.

Conclusões: A disponibilidade de dados desagregados é uma ferramenta poderosa para o direcionamento de intervenções onde são mais urgentemente requeridas, determinando, assim, prioridades.

Palavras-chave: Demografia; Planejamento de Saúde; Saúde Urbana; Epidemiologia; Política de Saúde.

 ABSTRACT:

Context: This study demonstrates wide disparities within São Paulo's selected districts, in terms of socioeconomic status as well as infant mortality and external causes specific mortality rates.

Objective: To update and disseminate information on urban inequalities in São Paulo, through the analysis of obtainable mortality and income statistics and the presentation of selected survival indicators in contrasting districts within the city.

Main mesurements: This was accomplished by collecting available data on population, mortality and income at the disaggregated level within São Paulo. Four Survival Indicators were then chosen for comparison: Infant Mortality Rates, Under Five Proportional Mortality Rates, Male 15-29 External Cause Specific Mortality Rates and Male 15-29 External Cause Proportional Mortality Rates. Finally, the relationship between income and these chosen survival indicators was examined, six representing the poor extreme and six representing the wealthy extreme.

Results: Childhood Mortality Rates in the poorest districts of São Paulo were between 3 and 13 times grater than in the wealthiest district. Mortality Rates due to External Causes in the poorest districts were between 6 and 24 times higher than in the wealthiest districts. In three of the six study low income districts, the leading cause of death for the entire population was External Causes, ahead of even cardiovascular disease. Four of the six low income districts had proportional mortality rates among young males attributable to External Causes that exceeded their Under Five Proportional Mortality Rates

Conclusions: The availability of disaggregated data has the potential to be a powerful tool in directing interventions to where they are most urgently needed and in determining emerging priorities.

Keywords: Demography; Health Planning; Urban Health; Epidemiology; Heath Policies. 

 1.- INTRODUCTION

At the close of the 20^th century, unprecedented numbers of people reside in urban areas, and rapid urbanization has become a fact of life. Just one hundred years ago, the world's urban population numbered 50 million while today just under 3 billion people, representing approximately half of the world's population, reside in urban areas ^7,16.

In that brief span of time, the world has witnessed the emergence of "mega-cities", defined by the United Nations as cities with greater than 8 million inhabitants. In 1970, there were 11 mega-cities, five of which were located in the developing world, whereas in 1995 there were 22 mega-cities, 16 in developing countries ¹⁵.

The relentless urban expansion has overwhelmed the capacity of city administrators to effectively provide either existing or emerging needs ¹⁶, and in the process left many hopeful migrants on the margins of society. This is especially the case in developing countries, which often lack the resources, infrastructure, planning and commitment necessary to provide basic services at the municipal level to match the pace of growth. In many cities, poverty, unemployment and frustration lead to a dangerously violent social environment, adding to existing miseries at the margins of urban life.

The expansion of the urban population, and its accompanying social dislocation poses challenges that the public health community will have to grapple with for the foreseeable future. Indeed, the sheer quantity of people living in cities and the proportion living in urban slums will constitute the biggest environmental health challenge of the 21^st century ².

A 1997 World Health Organization¹⁵ report concluded that the world's poorest people are predominantly found in two types of areas: remote and ecologically fragile rural areas, and at the margins of expanding urban areas. The pre-existing infrastructure in many cities is frequently unable to absorb the incessant influx of newcomers, who are forced to settle in precarious housing. It is estimated that between 30 and 60 percent of the urban population in low income countries lives in poor quality housing. As a consequence, there is projected to be one billion people, fully one sixth of the world’s population, living in urban poverty by the year 2000. Twenty-four percent of whom, reside in Latin America ⁶.

With the trend towards urbanization showing no signs of slowing down, the unique health problems associated with the urban environment, especially those that burden the urban poor, warrant special and increased attention. Yet the urban poor have been largely forgotten. Perhaps this is due to the fact that there is a paucity of accurate statistics regarding life chances and health risks in the slums of major cities in the developing world. Another explanation is the historical perception (now being challenged) that urban areas are better off than rural areas simply because the better off live in them. It has been precisely for this reason that many anti-poverty programs have traditionally focused on rural areas ².

While in many instances urban statistics may look better than rural statistics or even national averages, it is often because slum inhabitants are not included in studies or considered in their analyses or conclusions. It is often the case that these anonymous urban inhabitants, deprived of an official identity, are not legal residents. And when these groups are included, their conditions can easily be offset by the prosperity that surrounds them. Reliance upon statistics that present city averages can lead to a grave misinterpretation of reality. "Properly compiled and disaggregated information reveals a quite different and more truthful picture" ⁷.

As recently as 1988, however, data comparing groups from different geographic areas or income levels within a given city were quite rare ⁷. A 1984 World Bank study found Infant Mortality Rates in the core areas of São Paulo Brazil to be 42 per 1,000 live births but 175 per 1,000 live births in marginalized, peri-urban municipalities ³. A study from Manila, Philippines¹⁷ (1996), also found infant mortality rates to be 3 times greater in poor squatter settlements in comparison to non squatter areas. On a similar note, a 1990 study by McCord and Freeman ³ found that mortality rates in Harlem were the highest in New York, and 50% higher than that of all U.S. blacks. They concluded that Harlem men were less likely to reach age 65 than men in Bangladesh.

Despite the startling results of such studies and the increasing realization of the importance of disaggregated urban data, the availability of data in the literature demonstrating intra-urban differentials in health status had not markedly improved by the mid 1990s ⁸. Although scarce, disaggregated urban health data can be powerful tools for municipal health offices and policy makers. Indeed intra-urban differentials in health status demand specific attention, so that interventions can be more effectively planned, targeted and implemented ⁸ .

The sheer size of many large cities today argues for more geographically precise divisions of routine indicators. For instance, the city of São Paulo alone, with it 10 million inhabitants is roughly the size of Senegal or Mali, and is bigger than European countries such as Switzerland, Denmark and Sweden, African nations such as Niger and Chad, and other Latin American countries such as the Dominican Republic and Haiti.

Greater São Paulo alone, with its 17 million inhabitants, has a population roughly equivalent to Syria or Australia, has more people than Cameroon, and more people than Israel, Jordan and Lebanon combined. Greater São Paulo alone is also larger than several of its Latin American neighbors, such as Chile and Ecuador. It also has more people than Bolivia, Paraguay and Uruguay combined ⁴.

Therefore, both the collection and dissemination of disaggregated urban health data in São Paulo is essential if effective interventions for high-risk groups are to be appropriately developed and implemented ¹².

São Paulo is one of the few megacities in which disaggregated indicators are available and in which intra-urban differentials in health status have been examined ⁹.

In 1994, Stephens et al.¹² conducted a study examining intra-urban differentials in health status in the cities of Accra, Ghana and São Paulo, Brazil. They collected indicators on education, income, sewage, water and housing density. Based on this cumulative data they created index and from which they divided the city into 4 socio-environmental zones, each comprising 14 districts. At the time of their study, the city of São Paulo consisted of 56 districts, the smallest units of disaggregation within the city. Since that time, the city has restructured the organization of its districts, and has created more, smaller districts occupying the same physical space. There are currently 96 districts in the city of São Paulo.

They determined that 44percent of the population lived in the worst zone, as determined by the lowest per capita income, highest densities, least adequate sewage and the least water consumption. They found that mortality Rates due to respiratory infections and diarrhea among children under 5 years of age were four times higher in the most deprived zone than the most privileged zone. Their study also noted the emergence of homicide, affecting the poor most severely, as a major killer, alongside cardiovascular disease, and stressed this altered the health profile within the city.

This article is an effort to update and disseminate recent information on urban inequalities in São Paulo through the analysis of obtainable health (mortality) and income statistics.

2. METHODOLOGY

São Paulo routinely collects mortality data, disaggregated by district and organized by age, sex and cause of death. Mortality data is 99 percent complete, of good quality and available for each of São Paulo’s 96 districts ^10,11. Morbidity and environmental indicators are not consistently and comprehensively available, and are therefore not included in this study. However, average income data by district was also available for 1997-8 through a special study carried out by the METRO (regional transport authority).

After reviewing the mortality profiles for each of the 96 districts, this study included the following survival indicators: Infant Mortality Rate; Under Five Proportional Mortality Rate; Age (15 – 29), Sex (male) and External Causes Specific Mortality Rate; Age (15 – 29), Sex (male) and External Causes Proportional Mortality Rate.

We selected these indicators for analysis because they all share a focus on survival, and reflect major public health outcomes in São Paulo. Furthermore, they incorporate the main causes of premature and preventable mortality in São Paulo. And finally, they dramatically illustrate intra-urban differentials in health status.

In addition, Infant Mortality Rates are widely accepted as an indication of overall health status within a community and can depict the relative quality of the living environment between the districts, as well as their relative access to quality health care. The External Causes Mortality Rates were chosen because São Paulo is entrenched in a well documented epidemic of violence. These indicators highlight the urgency and extent of the public health problem of violence in São Paulo by focusing on the population most at risk: young males aged 15-29, living in poor areas.

POPULATION DATA

In 1996, The Population Division of Fundação SEADE conducted a population count to update the population of São Paulo, last estimated by the 1991 CENSUS. Based on the 1996 count, population estimates were projected to 1998 for each of the 96 municipal districts, divided by gender and age group. A further adjustment was made in the 0-4 age group, in order to compensate for underestimation of this age group population at the household level.

MORTALITY DATA

By law, when someone dies, the family is required to immediately notify a registry office. Each death is classified in groups of Major Causes according to the Tenth Revision of the CID ⁵. SEADE’s mortality statistics, presented by age and gender, are listed by a Major Group Cause of Death. The mortality figures correspond to events transpired in 1998.

With 1998 population data for each of the 96 districts matching the 1998 mortality records for the same districts, mortality rates which take into consideration cause, age and gender can easily be calculated. It is therefore possible to determine each district’s mortality profile, the leading causes of death, and the share of mortality for which a given cause is responsible.

LIVE BIRTH DATA

In a similar fashion, Fundação SEADE also collects the total number of Live Births in the São Paulo municipality, and disaggregates them by district. The process of collection is similar to that of mortality. Although notification of live births in not mandatory, as is the case of mortality.

SEADE then groups the data into regular place of mother’s residence and then compiles the figures for the 96 districts. It is worth noting that in order to register a newborn child, a parent also must be officially registered or have an official identity. This poses a problem for many of the low income districts with recent, often unregistered, immigrants and may lead to underreporting. The effect on the results would be higher infantile Mortality Rates in the low income districts. The underrreporting of births have no influence on Under Five Proportional Mortality Rate. Disaggregated Infant Mortality Rates were calculated first by taking all deaths in the first year of life for each district and then dividing by the number of live births in that district.

INCOME DATA

Within metropolitan São Paulo, the METRO periodically performs a Origen Destination (OD) survey. It is considered an important instrument to urban planning activities as well as to metropolitan São Paulo’s regional transport. The most recent METRO survey was conducted in 1997, and analysis was completed in 1998. The survey included 39 municipalities in the Greater São Paulo area. 120.000 people in 30.000 residences were interviewed for a variety of socio-economic indicators. METRO classified each respondent according to their place of residence, and divided the survey into 389 zones of research (called OD zones). None of these METRO zones overlap or occupy territory in more than one of the city of São Paulo’s 96 districts. In other words, each one of the city of São Paulo’s 96 districts comprises one or more of the 389 Metro zones. Since each district shares identical borders with the a group of METRO zones, it was possible to total the data for all of the METRO zones that fall within each district and aggregate METRO findings into our principal geographical unit of analysis, the 96 city districts.

It is worth noting one fundamental limitation with respect to the income data: not everyone declared their personal monthly income. For the purposes of our study, we totalled the income that was declared in all of the zones that made up a given district. We then totalled the population that declared for that same district. Average family monthly income was thus determined by taking the total income declared for each district divided by the population within each district that declared.

In order to correlate income with survival indicators in the 12 selected districts, we devised a proportional income scale. The lowest monthly per capita income, Grajaú at R$ 156.00, was given a value of 1. The monthly per capita income of the remaining districts were then given a value based on the number of times their income exceeded the minimum in Grajaú. Jardim Paulista, at the other end of the spectrum, with a monthly per capita income of R$ 1.466,00 has a monthly per capita income 9.4 times greater than that of Grajaú.

SELECTION OF 12 STUDY DISTRICTS

With 1998 population, live birth and mortality data, we created mortality profiles for each of São Paulo’s 96 districts. After reviewing the mortality profiles, the selected survival indicators, and income for each of the districts, we chose 12 districts at opposite ends of the spectrum to highlight the wide disparity in health status within the city of São Paulo.

The study districts were selected by their overall ranking for all four of the survival indicators. The individual ranking of four survival indicators for each district were totalled and a composite ranking was developed. The six poor study districts selected for this study had the six lowest (worst) composite rankings, and the six wealthy districts were among the twelve highest (best) composite rankings.

One district, Marsilac, was omitted from this analyses despite its exceptionally high mortality rates for each of the survival indicators, because its population, only 7.000, was deemed too small in comparison with the other districts.

We compared 12 of the 96 districts in the São Paulo municipality with the following survival indicators: Infant Mortality Rates, Under Five Proportional Mortality Rates, Age and Cause Specific Mortality Rates, Age and Cause Specific Proportional Mortality Rates ¹¹. We also compared each of the districts in terms of their Average Family Monthly Income. Wide variation was observed between the districts, both in terms of income and survival. The 12 districts included in our analysis represent 17 percent of São Paulo’s population, and had an average 1998 population of 140.345. The 12 districts varied in size of population, from 67.809 in Pinheiros to 289.418 in Grajaú. Average monthly family income for the 12 districts was 683 reais, in contrast with the average of 493 reais among all 96 districts. The higher average income among the 12 selected districts is due to the fact that six of the wealthiest districts were included in the analysis are only balanced by six of the poorest districts. The six poor study districts had an average income of 189 reais per month; the six wealthy study districts had an average income of 1.178 reais per month. At the time of the study, 1 Brazilian Real was approximately equal to 1 U.S. Dollar.

In Table 2, with respect to Infant Mortality Rates, there was substantial variation between districts, although not as wide as the other indicators. The average Infant Mortality Rate was 16.1 per 1,000 live births in the city of São Paulo; however, the average was 21.1 per 1,000 live births among the six poor study districts and 9.4 among the six wealthy study districts. The extremes were to be found in Jardim Paulista, with a rate of 6.5, and in Cidade Tiradentes, with a rate of 24.4.

The risk of death to an infant in Cidade Tiradentes was almost four times as great as in Jardim Paulista as shown in Figure 1. Infant Mortality Rates are consistently over three times greater in the poorest areas of the periphery, such as Brasilândia, Jardim Ângela and Grajaú, than in the wealthier district of Jardim Paulista. Each of the poor districts had Infant Mortality Rates between 3 and 3.8 times as high as Jardim Paulista. And each of the wealthy districts had Infant Mortality Rates less than twice as high as Jardim Paulista. In that sense, the relationship between Infant Mortality Rates and income appears quite strong. As income rises, Infant Mortality rates steadily decline, and vice versa.

An even wider differential between the districts was found in terms of Under Five Proportional Mortality Rates (<5 P.M.R). The city wide Under Five Proportional Mortality Rate was 6.4. In the six poor study districts, the proportion of all deaths occurring Under Five was 12.9 percent in contrast to 1.7 percent among the six wealthy study districts. Jardim Paulista’s <5 P.M.R. of 1.2 was the lowest in São Paulo, while the highest rate of 15.4 was found in Cidade Tiradentes. The six wealthy study districts had rates very close to the lowest - all within 1.7 times that of Jardim Paulista. Conversely, the six poor study districts had <5 P.M.R. between 8 and 13 in excess of the best district, Jardim Paulista. (Figure 1)

Figure 1

The average external causes specific mortality rate among males age 15-29 was 26.7 per 10,000 inhabitants in the city of São Paulo. Among the six poor study districts, the rate was 45.3 per 10,000 in contrast to 8.8 per 10,000 in the six wealthy study districts. The more prosperous regions lost between 6 and 13 males between 15 and 29 per 10,000 to external causes, while the destitute districts lost between 35 and 54 males each year. (Table 2)

The wealthy districts each had rates of male, 15-29 external causes specific mortality less than 2 times that of Jardim Paulista, while the poorest districts had risks between 6.1 and 9.2 times as high. ( Figure 2)

Examining those same deaths – males 15 to 29 due to External Causes – as a proportion of all deaths within the district, the gap widens even further. On average the proportion of mortality due to external causes among males 15-29 in the city of São Paulo was 5.7 percent. Among the six poor study districts the proportion was 13.2 percent, in contrast to 1.2 percent among the six wealthy study districts. The share of mortality attributed to males 15-29 from external causes in districts with the highest income ranged from 0.7 to 1.5 percent, whereas they accounted for 10.6 to 16.9 percent of all mortality in the districts with the least income. ( Table2).

The share of External Causes Mortality among young males is almost 25 times higher in Jardim Ângela, than in Jardim Paulista. The wealthiest six districts had risks ranging from only 1.1 to 2.4 times higher than the lowest rate, while the poorest six districts had risks from 15.1 to 24.1 times as high. (Figure 2).

Figure 2

This study concluded that there was wide disparity between São Paulo's districts with respect to the four survival indicators. Among the 12 study districts representing the extremes of São Paulo, adverse health outcomes were between 3-24 times higher in the worst districts than in the best districts. The differences were 3-4 fold in the case of Infant Mortality Rates, 8-13 fold for the Under Five Proportional Mortality Rates, 6 - 9 fold for the male 15-29 External Causes Specific Mortality Rates and 15-24 fold for the Male 15-29 External Causes Proportional Mortality Rates.

The findings also highlight the differences at the extremes of São Paulo with respect to the average monthly family income level. The relationship between income and the four survival indicators was quite strong among the twelve study districts. It is worth noting, as Richard Wilkinson¹⁴ (1996) argues in his book "Unhealthy Societies: The Afflictions of Inequality", that there is ample evidence indicating a strong relationship between poverty and elevated mortality rates, throughout history and across countries, age and ethnic groups.

The availability of quality mortality data made this study possible. There are, however, still limitations to São Paulo’s Health Information Systems that, if addressed, could substantially contribute to an improved understanding of public health in São Paulo. For example, morbidity data is currently not available throughout the city because of the discrepancy of reporting regulations between public and private institutions. Likewise, health expenditure data would contribute to our understanding of where resources allocated to health are deployed and concentrated. Therefore, São Paulo should expand its capacity to systematically collect and disseminate other types of disaggregated health data.

Despite its limitations, São Paulo is further along than many of the cities in the developing world. It is one of the few cities in which a study like this may be carried out because of the relative strength of its information system infrastructure. Other cities that are currently unable to determine precise health profiles at the disaggregated level should invest in systems able to routinely collect accurate disaggregated data on health conditions and commit to develop focused urban health initiatives.

This is an urgent priority so that the stream of urban poor can be identified and their needs explicitly planned for and met. Disaggregated data will be essential to improving the lot of the urban poor. Simply put, the mortality data compiled by SEADE in São Paulo enables the calculation of age, gender and cause specific mortality rates from which comprehensive mortality profiles can be created for each district. This can inform the public health community of which adverse health outcomes are occurring where and within which segment of the population -- a gold mine for program managers and public health decision and policy makers. Disaggregated data can greatly help identify high risk groups within the city as well as identify emerging districts with severe mortality burdens. It can also be used to as baseline data by which progress can be monitored and impact of interventions evaluated over time.

With the mechanism to do this already in place, such data must then be used to formulate interventions intended to serve those in most need. Yet although this data is available, collected and occasionally disseminated, there is no evidence in the literature documenting the constructive utilization of disaggregated data to benefit groups and areas in urgent need.

Disaggregation of data in the São Paulo study served its purpose by dramatically contrasting one district with another and with an aggregate view of the city for each of the four survival indicators. Examining the conditions at the extremes revealed a more dramatic and truthful distribution of health outcomes.

We thank Eliana Monteiro Rodrigues, Paulo Roberto Campanário and Paulo Borlina Maia for their kindness on facilitating our access to SEADE data files.