Tuberculosis associated with the living conditions in an endemic municipality in the North of Brazil*

Objective: to analyze the association between the occurrence of new tuberculosis cases and the Adapted Living Condition Index, and to describe the spatial distribution in an endemic municipality. Method: this is an analytical and ecological study that was developed from new cases in residents of an endemic municipality in the North Region of Brazil. The data were obtained from the Notifiable Diseases Information System and from the 2010 Demographic Census. The Adapted Living Conditions Index was obtained by factor analysis and its association with the occurrence of the disease was analyzed by means of the chi-square test. The type I error was set at 0.05. Kernel estimation was used to describe the density of tuberculosis in each census sector. Results: the incidence coefficient was 97.5/100,000 inhabitants. The data showed a statistically significant association between the number of cases and socioeconomic class, with the fact that belonging to the highest economic class reduces the chance of the disease occurring. The thematic maps showed that tuberculosis was distributed in a heterogeneous way with a concentration in the Southern region of the municipality. Conclusion: tuberculosis, associated with precarious living conditions, reinforces the importance of discussion on social determinants in the health-disease process to subsidize equitable health actions in risk areas, upon a context of vulnerability.


Tuberculosis (TB) is an infectious disease which
is caused by Mycobacterium tuberculosis, with a high impact on global public health (1) . It is a millennial, singleagent disease that has caused the most fatalities and affects thousands of people around the world, ranking among the 10 diseases with the highest mortality rates on the planet (2)(3) .
For 2018, 10 million new cases were estimated in the world, with an incidence in the countries varying from 5 to more than 500 cases per 100,000 inhabitants.
Brazil is among the 30 countries with the highest TB loads (3) . For 2019, the incidence coefficient was 35.0 cases per 100,000 inhabitants. The incidence in the country had decreased between 2010 and 2016 but, from 2017 to 2018, this measure increased (4) .
In 2019, the state of Pará was among the federal units with an incidence rate close to or above the national coefficient, and its capital city, Belém, was among the 5 capitals with most incidence of the disease in 2018 (62.7 cases/100,000 inhab.) (4)(5) .
The epidemic in the country does not have a heterogeneous character, but it has centralized on vulnerable populations such as street people, individuals deprived of their freedom, indigenous population, and individuals living with the Human Immunodeficiency Virus (HIV). In this sense, clinical and epidemiological management is a challenge for health professionals, managers, TB patients, families, and organized civil society to implement inclusive, focused, and coaccountability policies (3)(4)(5)(6) .
There are several factors that boost the occurrence of TB, among them, the socioeconomic conditions and the difficulties of access to the health services. Such conditions express the precarious living conditions related to poverty, low schooling, unhealthy housing, population thickening, and abusive drug use (7) .
In this sense, TB has been considered a marker of social inequities in health (8) . The persistence of unequal social models interferes in the health-disease process, especially in the chain of transmissibility, and it predicts the multi-causal dynamics of illness based on the social determinants of health, regarding low living conditions and its impact on the individual/society relationship in the different regions of the country.
In order to provide satisfactory answers on the density of TB and its distribution in relation to the living conditions, this study proposes a smoothed innovative spatial analysis, independent of the geographical limits for the visualization of the disease. Studies that consider the spatial and temporal diffusion of diseases make it possible to understand how the occurrence of health adverse events affects population groups and spread in territories (9) .
It is understood that a broad look at the needs of the population can support public policies and guidelines for planning actions and conducting Primary Health Care services, based on emancipatory practices, aimed at attaining global goals to combat TB.
Thus, the hypothesis of this study is that the occurrence of new TB cases is associated with the strata of the municipality with more precarious living conditions, and the objectives are the following: to analyze the association between the occurrence of new tuberculosis cases and the Adapted Living Condition Index, and to describe the spatial distribution in an endemic municipality.    (15) .

Method
To estimate a territorial distribution surface of TB from geocoded addresses, the Kernel Density Estimator (KDE) was used. The main objective of the KDE is to generate a regular grid where each cell represents a density value (16) . This is a non-parametric technique that promotes statistical smoothing, giving rise to chromatic gradients with "hot areas" to the extent that in that region there is a vast density of cases (16) .
The KDE method is based on search radios that can be prefixed or adaptive. Due to the unequal distribution of the cases, the quartic-function adaptive radio was used.
Cluster detection techniques tend to have a spatial distribution similar to the population distribution in the health events. This distribution may derive from social, historical, and economic organizations. However, the Kernel estimator does not predict only the distribution of clusters but explores the behavior pattern of the health data points. Thus, it generates a continuous surface from point data, which allows for a quick visualization of the areas that deserve more attention, being an important tool for the analysis of events and for the rapid implementation of strategies in the area of public health (17) . Throughout this period, the incidence coefficient of TB was higher in men (12.4/10,000 men) than in women (7.3/10,000 women), and the age group most affected was that of the older adults aged 60 or over (13.8/10,000).
To identify possible social disparities in the geographic space of Belém, the ALCI was built with seven variables, whose data were obtained in the IBGE electronic portal, using descriptive statistics (Table 1).   (Table 2). The variables that expressed higher factor loads were The chi-square test showed a statistically significant association between the occurrence of TB cases and socioeconomic class (χ 2 3; 0.05 = 104.51; p < 0.001).  to inconsistency in the addressing system. From the geocoding it was possible to produce Kernel maps, expressing the density of TB cases, which is higher in the darker regions ( Figure 1).

Discussion
The results of this study showed a spatial dependence in the occurrence of TB cases, with higher density in the Southern region of this municipality, over the years studied. It is important to consider that the tendency of TB is associated with multiple historical and social processes that involve social determinants of the health-disease process and demand individual, collective, and programmatic strategies of the social actors to eliminate it, especially in vulnerable populations (18)(19) .
The results showed greater predominance of cases in young adults and in individuals over 60 years of age, with 13.3 cases/100,000 inhabitants in the 20-29 age group, and 13.8 for those over 60 years of age. This to information, knowledge benefits, consumer goods, and health services (25) .
A higher prevalence of the pulmonary clinical form was observed, which characterizes a higher risk of transmissibility among the people living in the studied municipality, due to its high infectivity. The interruption André SR, Nogueira LMV, Rodrigues ILA, Cunha TN, Palha PF, Santos CB.
of transmission requires immediate intervention by the health services to promptly diagnose and treat the disease (21) , in addition to notification and active search for the patients' contacts. Each patient diagnosed with TB tends to infect 10 to 15 people within a year, and one or two get ill, maintaining the cycle of the endemic (26) . The results also show that the "heads of home with a monthly income less than or equal to two minimum wages" variable portrays the precarious socioeconomic panorama of the population in Belém, since it presented a mean of 69.6%, while in municipalities like Ribeirão Preto this mean was 23%, revealing the influence of the economic factor as a conditioning factor for the development of TB in low-income individuals (28) . TB is associated with low living conditions and income, related to problems such as population growth, street people, chemical dependency, poor housing conditions, poor nutrition, low income, lack of sanitation, and other determinants (29) .
The ALCI obtained the "picture" of the living conditions of the population since the variables studied concern the socioeconomic aspects related to TB, which, even associated with underprivileged conditions, reach expressively strata with better living conditions (9) . to the specific treatment, which is a consequence of restricted access to information, knowledge benefits, consumer goods, and health services (27) .
The location and geographical analysis of areas considered at risk for TB development were presented in this study through spatial analysis techniques, which contributed to the understanding of the current health context and its trends, building approaches directed at health surveillance practices, such as the identification of risk areas, population concentration, and prioritization of actions and resources, as well as the possible association of local conditions in the social environment where the patients live (30) .
A study conducted in the municipality of Belém showed that the spatial analysis exhibited areas with similar TB incidence with a tendency to clusters, and the same profile was found in this study, where neighborhoods with similar TB rates were close by.
Although the distribution density is given in a variety of ways, the geographic regions affected by TB showed a predictable distribution pattern regarding the affected neighborhoods, which leads to questions about the effectiveness of disease control actions in these places (31) .
A study conducted in Ethiopia concluded that, despite different intervention programs aimed at reducing disease transmission and improving diagnosis, abnormal incidence rates persisted in the same locations with the most likely spatial clusters (32)  associating the high demographic density of the urban municipality with the high occurrence of the disease (33) .
Regarding the clusters formed by the association between TB and the ALCI, it can be highlighted in this study that the municipality of Belém presented similar characteristics to the findings in the literature (28,34) .
The higher concentration of TB in strata of poorer living conditions shows that the disease is associated with underprivileged conditions; however, even in strata of better living conditions a significant number of cases of the disease are still found (9) . Findings of a research conducted in Campina Grande revealed a higher mean incidence rate in strata of "worse" living conditions; however, the "best" living condition stratum had a higher incidence rate than the "regular" and "bad" living condition strata (9) . Understanding the ways in which the disease spreads and how health actions are implemented impacts on the planning measures focused on the diversities.
The planning actions to combat and control TB must be assured so that the health service is prepared, offering quality and accessible assistance that presents better health outcomes throughout the country (35) . The interventions should be targeted at underprivileged areas as these are regions most affected by the disease, in order to reduce transmission (36) .
In a study conducted in South Africa, the distance to the diagnostic health unit in a cohort of patients with resistant TB was assessed, and it was found that a large proportion of patients sought the health service outside their home district (37) . Such a situation may reveal the stigma related to the disease still present in society, which is capable of contributing to the low adherence to the treatment and the search for services outside the area of their territory. More importantly, the spatial analysis was based on the geocoding of the cases by addresses obtained from the aforementioned database, with inevitable losses due to absences and inconsistencies.
Since one of the study objectives was to correlate TB with ALCI in the urban space, no population-based rates were calculated; however, the spatialization of the ALCI is done by Census Sector. Hence the adoption of the density estimator by Kernel to infer the correspondence of TB occurrence with the ALCI. It is suggested that this study be extended in the future to bring analyses by a spatial structure such as Census Sector or Neighborhood.
Additionally, we signal the difficult separation into lower economic class strata in a municipality where the majority of the population is socially segregated, but with territorial concentrations not always clear. In order to highlight the territorial discrimination in these cases, more refined indicators would be needed, taking into account other aspects of the social inequalities.

Conclusion
The description of the TB spatial pattern allowed the intensity of the disease to be visualized from the behavior of point patterns, which constitutes a refined first-order analysis to subsidize equitable public health actions in risk areas.
In addition, the statistically significant association between the occurrence of TB and the strata representing worse living conditions reasserts that the disease remains associated with social vulnerability reaching more people in situations of exclusion. These findings reinforce the importance to effectively discuss the health social determinants, which are essential for planning and formulating intervention measures for combating and controlling the disease in this context.