INTRODUCTION

Adolescence is understood as a period of transition from childhood to adulthood, representing a phase of intense human growth and development that involves multiple and simultaneous transformations of physical, psychological, and social aspects ^1,2 . This can be understood both as a subjective response to the onset of puberty and as a cultural and social construction linked to age ^3,4,5 .

The Ministry of Health (MS) chronologically defines adolescents as individuals in the age group of 10 to 19 years, and can be classified as “early adolescents” (10 and 14 years old) and “young adolescents” (15 to 19 years old) ^2,5 .

According to the Department of Information Technology of the Unified Health System (DATASUS) ⁶ , in 2016, Bahia had 2,598,291 adolescents. Of these, approximately 51% are male and 49% are female, 50.2% are between 10 and 14 years of age and 49.8% are between 15 and 19 years of age. Despite the demographic importance associated with their vulnerability to health problems, adolescents are a group excluded from health care and attention that considers the specificities of their demands.

Adolescents’ health is influenced by factors that are associated with both the causes of diseases and deaths in children and young adults. In this sense, prevention and control measures initially depend on the identification of the central determinants of diseases and injuries in this group, made possible by “establishing patterns of distribution of diseases and injuries” ⁷ . These patterns can be generated from the use of collective health measures that describe the health situation, among which mortality indicators stand out ^7,8 . Statistics enable a situational diagnosis that supports planning and evaluation of health promotion and disease prevention actions ⁷ .

The objective of this article is to identify the pattern and trends in adolescent mortality in the State of Bahia, from 1989 to 2016, according to age group, sex, regional health centers and groups of causes.

METHODOLOGY

This is an ecological study of temporal trends in adolescent mortality in Bahia from 1989 to 2016. The State of Bahia is composed of 417 municipalities distributed in nine Regional Health Centers (NRS) and has an estimated population of 15,339,922 inhabitants, of which 16.77% correspond to the adolescent population ^6,9 .

The period chosen was 1989, based on the year of enactment of the Adolescent Health Program (PROSAD) ¹⁰, up to 2016. The data used are in the public domain and correspond to registered deaths of adolescents, extracted from the Mortality Information System (SIM) of the State of Bahia in the period from November to December 2016, from the websites of DATASUS and DIS/SUVISA (Health Information Directorate of the Superintendence of Surveillance and Health Protection of the State of Bahia). From this system, which is currently being replaced by e-SUS, variables such as age group, sex, health macro-regions, coinciding with the NRS, and causes of death were extracted.

The causes are named according to the International Statistical Classification of Diseases and Related Health Problems (ICD) and the study period encompasses the ninth and tenth revisions. Thus, in order to avoid biases in the collection and analysis of information, a grouping of the large groups of causes was performed between ICD-BR-9 and ICD-BR-10.

Information on the resident population in each year of the study period was collected based on estimates from the Brazilian Institute of Geography and Statistics (IBGE). It is noteworthy that until the submission of this study, there were no available data on the resident population by macro-regions in 2016.

Data analysis was based on Specific Mortality Coefficients (SMC), which are defined as the number of deaths by sex, age or cause, divided by the population of that same group, considering the area and year, multiplied by one hundred thousand inhabitants. The race/color variable was excluded due to the large amount of ignored information.

The linear regression method was used to analyze the trajectory of the SMC. To analyze the difference in SMC means by sex and age group over time, the Student’s t- test for paired data was used. The analysis of the SMC by region and by causes was performed using the F-test – Analysis of Variance (ANOVA) for repeated measures and for the comparison of means in pairs, the Bonferroni test. The significance level used was 5%.

Line graphs and trend adjustment by the least squares method were used, and a box plot was also used to visualize the behavior of the mean CME level in the various categories. The data were compiled in the EXCEL© Electronic Spreadsheet and statistical analyses were performed with the Statistical Package for the Social Sciences (SPSS), version 22.

This research is part of the Project “Adolescent Health Care in the State of Bahia: unveiling public policies and indicators” approved by the Human Research Ethics Committee of the State University of Santa Cruz (CEP-UESC) with Certificate of Presentation for Ethical Appreciation number 60045716.2.0000.5526, approved in November 2016 and funded by the Bahia State Research Support Foundation (FAPESB).

RESULTS

During the study period, 52,915 adolescent deaths were recorded in the State of Bahia, of which 73.7% (39,021 deaths) were deaths of adolescents aged between 15 and 19 years. In this group, 2016 stood out as the year with the highest mortality rate: 174.9 per 100,000 inhabitants. The CME of adolescents aged 15 to 19 years was always higher than the CME of adolescents aged 10 to 14 years, almost constant around their averages until 2006. From 2007 onwards, the CME of adolescents aged 15 to 19 years began a linearly increasing trajectory, justifying the establishment of a cut-off point between these years in order to facilitate statistical analysis (Figure 1).

The means and statistics of the Student’s t- test for paired samples showed that these differences were statistically significant. In the period from 1989 to 2006, the CME of the 15- to 19-year-old group was more than double that of the 10- to 14-year-old group and quadrupled in the period from 2007 to 2016. The 10- to 14-year-old age group maintained a stable behavior throughout the analysis period.

The linear adjustment of the trends showed that the CME of the 10- to 14-year-old group is not related to time, since the slope could be considered zero ( p = 0.483). That is, the CME tended to a constant around a CME of 33.5, very close to the average value of the entire period, which was 32.8. The same occurred with the trend for the 15 to 19 age group until 2006, with CME independent of time, since the angular coefficient was considered zero ( p = 0.174). In other words, the CME tended towards the value of 72.3, which was very close to the average of 74 corresponding to the period. However, from 2007 onwards, the CME of the 15 to 19 age group began a linearly increasing trend, at a rate of 6.72 per year ( p = 0.001), which dramatically described the situation of this age group (Figure 1).

Regarding the distribution of mortality by sex, it was found that male mortality was always statistically higher than female mortality, according to the result of the Student’s t- test (10 to 14 years – 17.366; 15 to 19 years – 8.801). In the 10 to 14 age group, male mortality was 60% higher than female mortality. In the 15 to 19 age group, from 1989 to 2006, male mortality was 2.6 times higher than female mortality, and this doubled from 2007 to 2016. It was found that the CME of males in the 15 to 19 age group was the only one that deviated from the patterns of the other categories (Figure 2).

The CME of the male and female groups aged 10 to 14 years and the female group aged 15 to 19 years showed a trend that could be considered constant over time. For the male group aged 15 to 19 years, the linear trend was increasing, and in the period up to 2006, the angular coefficient was no more than one unit per year (y = 0.7326), becoming 13.098 each year ( p = 0.000) from 2007 onwards (Figure 2).

Regarding the distribution of the CME in the age group of 10 to 14 years by NRS, it was observed that there was a significant difference between the regions (F _(1,26) = 1796.71; p = 0.000), with the Far South, East and South standing out in relation to the others, according to the Bonferroni test.

In the 15-19 age group, there was also a significant difference between the NRS (F _(1,26) = 385.79; p = 0.000), with these same regions occupying the first three positions. However, in this case, the East NRS occupied the first position, followed by the Extreme South and South, according to the Bonferroni test. It is important to mention that the Central North Region showed a reversal in the trend from 2007 onwards and such a large variability that it compromises the comparison between the averages.

Figure 3 shows the average CME of each NRS, by age group, and in the 15-19 age group, the average was calculated in the two separate periods (1989 to 2006 and 2007-2015). When comparing age groups, it was observed that in all NRS the death rates were higher for adolescents aged 15 to 19, with emphasis on the Far South, East and South. It was also found that there was a high correlation between the CME of the age group 10 to 14 and the CME of the age group 15 to 19. That is, regions with a higher rate in one age group tend to reproduce this high rate in the other age group, except for the Central North region, which was the only one that changed the trend by suffering a strong drop in the second study period, becoming a point outside the trend.

When analyzing the CME by causes in the State of Bahia, deaths due to “external causes of morbidity and mortality” remained in first place throughout the study period, in both age groups. In the case of the 10 to 14 age group, the average level by causes differed significantly (F _(1,27) = 2561.08; p = 0.000). External causes reached a CME of 14.7, against 4.53 for the second cause, with the other three being below 2.5 and not differing statistically, according to the pairwise comparison of the Bonferroni test.

In this age group, the second cause of death corresponded to “symptoms, signs and abnormal findings of clinical and laboratory tests not classified elsewhere”. In other words, ill-defined causes, followed by “neoplasms”, “diseases of the circulatory system” and “infectious and parasitic diseases” (Figure 4).

In the case of the 15 to 19 age group, the differences were also significant (F _(1,27) = 156.21; p = 0.000); the average number of external causes was 67.12, compared to 7.78 for the second cause, which was ill-defined causes. The third leading cause of death in this group was “diseases of the circulatory system”, followed by “neoplasms” and “endocrine, nutritional, metabolic, blood and hematopoietic organ diseases and some immune disorders” (Figure 4).

Regarding deaths from external causes, it was observed that in both age groups, aggression was identified as the leading cause of death, with a considerable increase from 2007 onwards. However, in the 15 to 19 age group, this increase became even more significant (Figure 5).

DISCUSSION

Regarding age group, it is noteworthy that in Bahia, the deaths of young adolescents have always been higher than those of early adolescents, especially when the rates began to rise dramatically.

According to Horta and Sena ¹¹ , public policies for adolescents nationwide increased between 1999 and 2002, but they were isolated projects that did not contribute to changing the health of adolescents. In Bahia, this reality is reaffirmed by observing the trajectory of CME over time, which categorically demonstrates the low effectiveness of the social and public health protection network, which should be guaranteed by the State ⁴ .

The data also show that there was an excess of male mortality, especially among adolescents aged 15 to 19, which showed increasing rates from 2007 onwards, corroborating other studies ^12,13 . It is known that the State of Bahia has the second largest difference in life expectancy between men and women (9.1 in favor of women) in the Federative Units ¹³ .

Excess male mortality is considered to be the result of the biological difference between the sexes that give women a naturally higher life expectancy than men. In addition, testosterone drives them to violence and more frequent exposure to risks ¹⁴ .

The IBGE ¹³ , in the document “Complete mortality table for Brazil – 2015”, states that excess male mortality occurred through a rapid process of urbanization and metropolization starting in the 1940s, when the population ceased to be essentially rural with precarious sanitary conditions. Currently, this phenomenon is concentrated in the group of young adolescents and young adults (15 to 29 years old) due to a higher incidence of deaths from preventable causes that intensely affect this population ¹³ .

The profile of adolescent mortality due to NRS reveals a concern for the Far South, which has the highest coefficients. The differences between the health regions of the state highlight the variability of the risks to which adolescents are exposed, with exposure to vulnerabilities related to the territory ¹⁵ .

It is noted that the three health regions (Far South, East and South) that stood out in relation to the others in terms of CME coincide with the three regions that had the highest Homicide Mortality Rates (HMR) in all age groups, in another study in Bahia ¹⁵ . Although it is possible to state that there is no uniform distribution of the injuries in the regions ¹⁵, it is emphasized that the determining elements of the mortality profile specific to each region are not easily identified.

When observing the causes of death, we find a worrying emphasis on external causes. This, as the main cause of death, has been highlighted since the publication of the PROSAD text ¹⁰ . Bustamante-Teixeira et al. ¹² state that among adolescents, Brazil has shown an increase in mortality rates due to external causes, with similar results being identified in Bahia since 2007.

Some studies reinforce this result by stating that preventable causes are the main responsible for the increase in adolescent deaths in Brazil ^4,12,16 , reaching more significant proportions when compared to the rest of the population and being related to marginality, drug trafficking and alcohol abuse ^16,17 .

Among external causes, aggression stands out, similarly to other studies ^4,15,18 . For young adolescents, these MEC have always been significantly higher compared to early adolescents. However, over the last 10 years, it has been observed that CME has been increasing among early adolescents, revealing the need for urgent intervention.

It is worth noting that in 2007, when mortality rates began to rise in Bahia, the Lethal Violence Reduction Program was created, which, through the Adolescent Homicide Index, aims to “contribute to monitoring the phenomenon of lethal violence in adolescence and to evaluating public policies aimed at prevention” ¹⁹ . It is noted that there was no positive influence on the change in mortality rates in Bahia, which was the state with the most municipalities that lead the national ranking of the incidence of adolescent homicides ¹⁹ .

Other causes that deserve attention are ill-defined causes, neoplasms and diseases of the circulatory system, which occupy the second, third and fourth positions of causes of death among adolescents, respectively. Results similar to the causes of death among adolescents at the national level ¹⁰ .

It is worth highlighting the initiative of the Ministry of Health to implement, since 2005, the program “Reduction of the percentage of deaths due to ill-defined causes” to reduce this proportion, which remains persistently high. Thus, monitoring and evaluation of the SIM information generation process must be present in the set of attributions of those responsible for health surveillance ²⁰ .

Thus, the analysis of adolescent mortality indicators in the State of Bahia reveals that the field of citizenship production for this age group needs to be further explored in order to promote public policies that take into account universality, access, equity and comprehensive care. It is necessary to incorporate the political dimension, in addition to the technical dimension, based on the construction of meanings for making and operationalizing decisions to transform this reality. Therefore, it is expected that this study will enable the creation, implementation and/or improvement of policies in order to offer improvements in care, and support managers, professionals and others involved in qualifying comprehensive care for adolescent health in the State.

CONCLUSION

Adolescent mortality in Bahia is an alarming health indicator that reflects the importance of focusing on the health needs of this population. The significant increase in CME since 2007 deserves to be investigated to identify whether these indicators correspond to the introduction of new determinants of health and safety, or whether they are the result of other factors such as a reduction in public policies and an intensification of the SIM feed in the state.

This study helps to elucidate the issue of adolescent mortality in Bahia, but leaves open other possibilities for investigation, such as a deeper look at determinants and the perspective of documents on the policies implemented in the state so that they can more assertively support management in tackling this multifaceted and transdisciplinary problem.

This study makes it possible to establish the diagnostic situation that seeks to subsidize new perspectives on health care. It is clear that government interventions must permeate the elements that make up the broader concept of health, ranging from health care in primary care links to public safety.

NOTE

Funded by the Bahia State Research Support Foundation (FAPESB).