Tuberculosis among Brazilian Indigenous people: temporal trends and factors associated with treatment abandonment and death, 2013-2022

Abstract

This study analyzed the temporal trend and factors associated with unfavorable tuberculosis outcomes in Indigenous populations in Brazil during 2013-2022. An ecological and cross-sectional study was conducted on tuberculosis cases among Indigenous people. Prais-Winsten regression was used for temporal trends and logistic regression for factors associated with outcomes using the odds ratio (OR). A total of 7,446 cases among Indigenous people were analyzed; 82.9% were cured, 9.1% abandoned treatment, and 8.0% died. The North region showed a 5.6% (95%CI=3.1; 8.1) annual increase in abandonment and a 1.5% (95%CI=-2.6; -0.4) reduction in cure rates during 2013-2022, along with a 4.5% (95%CI=1.0; 8.2) increase in new cases and a 24.4% (95%CI=18.6; 30.6) rise in deaths from 2013 to 2019. Factors associated with treatment abandonment were insufficient monitoring (OR=14.07; 95%CI=8.25; 23.99) and lack of observed treatment (OR=1.82; 95%CI=1.32; 2.52). Predictors of death were age over 60 (OR=8.35; 95%CI=4.93; 14.15), living in the North (OR=3.25; 95%CI=1.68; 6.27), and insufficient monitoring (OR=42.94; 95%CI=20.75; 88.82). In summary, the temporal trend was unfavorable in all North scenarios, with multiple predictors of abandonment and death identified, particularly insufficient monitoring of cases.

Key words:
Tuberculosis; Indigenous Peoples; Death; Patients who dropped out of treatment

Resumo

Objetivou-se analisar a tendência temporal e os fatores associados aos desfechos desfavoráveis da tuberculose em indígenas no Brasil no período de 2013-2022. Utilizou-se da regressão de Prais-Winsten para tendência temporal e regressão logística para os fatores associados aos desfechos usando a Odds Ratio (OR). Foram analisados 7.446 casos em indígenas, sendo que 82,9% evoluíram para cura, 9,1% abandonaram o tratamento e 8,0% foram a óbito. A região Norte apresentou crescimento do abandono de 5,6% (IC95%=3,1; 8,1) ao ano e redução das proporções de cura de 1,5% (IC95%=-2,6; -0,4) no período de 2013-2022, além de crescimento do número de casos novos de 4,5% (IC95%=1,0; 8,2) e óbitos 24,4% (IC95%=18,6; 30,6) de 2013 a 2019. Os fatores associados ao abandono de tratamento foram: casos com indicador de acompanhamento insuficiente (OR=14,07; IC95%=8,25; 23,99) e ausência do tratamento observado (OR=1,82; IC95%=1,32; 2,52). Os preditores de óbito foram: idade acima de 60 anos (OR=8,35; IC95%=4,93; 14,15), residir na região Norte (OR=3,25; IC95%=1,68; 6,27) e ter indicador de acompanhamento insuficiente (OR=42,94; IC95%=20,75; 88,82). Em suma, a tendência foi desfavorável em todos os cenários no Norte, enquanto múltiplos preditores do abandono e óbito foram identificados como o acompanhamento insuficiente dos casos.

Palavras-chave:
Tuberculose; Indígenas; Morte; Pacientes desistentes do tratamento

Resumen

Se analizó la tendencia temporal y los factores asociados a los desenlaces desfavorables de la tuberculosis en poblaciones indígenas de Brasil durante el período 2013-2022. Se utilizó regresión de Prais-Winsten para evaluar las tendencias y regresión logística para identificar los factores asociados, utilizando la razón de probabilidades (OR). Se analizaron 7.446 casos: el 82,9% evolucionaron hacia la curación, el 9,1% abandonaron el tratamiento y el 8,0% fallecieron. La región Norte presentó un aumento anual en el abandono del tratamiento del 5,6% (IC95%=3,1; 8,1) y una reducción en las tasas de curación del 1,5% (IC95%=-2,6; -0,4) durante 2013-2022. Además, entre 2013 y 2019, se registró un crecimiento en los casos nuevos del 4,5% (IC95%=1,0; 8,2) y en los fallecimientos del 24,4% (IC95%=18,6; 30,6). Los factores asociados al abandono fueron seguimiento insuficiente (OR=14,07; IC95%=8,25; 23,99) y ausencia de supervisión del tratamiento (OR=1,82; IC95%=1,32; 2,52). Entre los predictores de mortalidad se identificaron: edad mayor a 60 años (OR=8,35; IC95%=4,93; 14,15), residencia en la región Norte (OR=3,25; IC95%=1,68; 6,27) y seguimiento insuficiente (OR=42,94; IC95%=20,75; 88,82). En resumen, la tendencia fue desfavorable en el Norte, y el seguimiento insuficiente destacó como el principal factor asociado a desenlaces negativos.

Palabras clave:
Tuberculosis; Pueblos Indígenas; Muerte; Pacientes Desistentes del Tratamiento

Introduction

The indigenous population is present throughout nearly the entire Brazilian territory. According to data from the 2022 census of the Brazilian Institute of Geography and Statistics (IBGE), there are approximately 1,694,836 indigenous people in Brazil, representing 0.83% of the country’s population, with the highest concentration (51.2%) located in the Legal Amazon¹. The indigenous population faces significant health disparities, such as inadequate healthcare systems, lower life expectancy, higher infant and maternal mortality rates, nutritional problems, and socioeconomic disadvantages².

Historically, infectious and parasitic diseases have been the leading causes of morbidity and mortality among indigenous populations in Brazil. Tuberculosis stands out as one of the primary endemic diseases affecting indigenous peoples due to its broad current distribution³.

In Brazil, it is estimated that between 2011 and 2017, more than 600,000 tuberculosis cases occurred, approximately 1.1% of which were among Indigenous peoples (N=6,520). The average incidence rate in this population was 109.0 cases per 100,000 Indigenous individuals. Regions such as the Central-West and North present the highest coefficients, with 202.5 per 100,000 and 124.9 per 100,000 Indigenous individuals, respectively⁴.

The Brazilian Ministry of Health considers the indigenous population as vulnerable to tuberculosis, with a disease risk three times higher compared to the general population⁵. Global literature indicates that factors such as advanced age, HIV history, male gender, recurrence of cases, and inadequate treatment follow-up increase the likelihood of treatment abandonment and mortality from the disease⁶.

Thus, despite national efforts to control tuberculosis, disparities in healthcare access and socioeconomic and cultural complexities faced by indigenous peoples result in challenges related to the disease’s prevention, diagnosis, and treatment. Current scientific literature offers limited insights into the influence of sociodemographic and clinical characteristics on tuberculosis treatment outcomes. Furthermore, there are few studies on this topic; among available studies, there is a specific focus on children and adolescents⁷, and the presence of small samples with only local representativeness^8,9. Therefore, this study aimed to analyze the temporal trend and factors associated with unfavorable tuberculosis treatment outcomes among indigenous peoples in Brazil from 2013 to 2022.

Methods

Study design

This study employed a mixed design, combining ecological time series analysis (aggregated data) and cross-sectional analysis (individual data) with a descriptive, exploratory, analytical, and quantitative approach. The research followed the guidelines of the Reporting of Studies Conducted using Observational Routinely-collected Health Data (RECORD)¹⁰.

Context

Public data from the Brazilian Information System for Notifiable Diseases (Sistema de Informação de Agravos de Notificação, SINAN) were studied, specifically, cases of tuberculosis notified in Brazil and its macro-regions from January 2013 to December 2022.

Participants

The study included cases of tuberculosis among individuals who self-identified as Indigenous. Cases were excluded if information regarding treatment outcome was missing or if they were classified as “treatment regimen change,” “treatment failure,” “diagnostic change,” “drug-resistant tuberculosis,” or “transferred.” The last two categories refer to cases transferred for treatment to specialized units and recorded in the Special Tuberculosis Treatment System (SITE-TB), preventing accurate determination of treatment outcomes.

Variables

The following variables from the tuberculosis notification form were selected for individual-level estimates: age (in years); age group (<19 years; 20-39 years; 40-59 years; ≥60 years); sex (male, female); macro-region of residence (North, Northeast, Southeast, South, and Central-West); year of notification (2013-2022); educational level (illiterate, incomplete primary education, complete primary education, incomplete secondary education, complete secondary education, incomplete higher education, complete higher education); case entry type (new case, recurrence, re-entry after abandonment, post-death, unknown); clinical form (pulmonary, extrapulmonary, mixed); chest radiography (normal, suspicious, other pathology, not performed); AIDS (yes, no); diabetes (yes, no); illicit drug use (yes, no); mental illness (yes, no); smoking (yes, no); directly observed treatment (yes, no); sputum smear microscopy (positive, negative, not performed); sputum culture (positive, negative, not performed); anti-HIV serology (positive, negative, not performed); rapid molecular test - TRM-TB (positive for rifampicin-sensitive tuberculosis, rifampicin-resistant, negative, not performed); and treatment outcome (cure, treatment abandonment, primary abandonment, and death from tuberculosis). The treatment outcome was considered the dependent variable, with the remaining independent variables (sociodemographic and clinical).

An empirical classification system for monitoring treatment adherence was also used as an independent variable (follow-up indicator). According to the Brazilian Society of Pulmonology and Phthisiology (SBPT)¹¹, this system combines five recommendations: sputum smear microscopy at months two, four, and six of treatment, examination of contacts, and treatment supervision. Cases meeting none or only one of these recommendations were classified as insufficient follow-up. Cases meeting two recommendations were classified as regular follow-up. Those meeting three recommendations were classified as a good follow-up, and those meeting at least four were classified as excellent.

Additionally, for the time series analysis, the following epidemiological indicators were calculated according to Ministry of Health standards¹²: cure proportion, abandonment proportion, incidence rate, and mortality rate. The first two indicators were calculated: number of cured or abandoned cases / total number of cases * 100. Incidence and mortality rates were calculated as follows: number of cases or deaths from tuberculosis among Indigenous people in a specific location and year/resident Indigenous population in the exact location and year * 100,000. It should be noted that only new cases, post-death cases, and unknown cases were considered for incidence calculations.

Data sources

Tuberculosis case data were sourced from SINAN, made available by the Department of Informatics of the Unified Health System (DATASUS). Data access occurred in April 2023 through the following steps:

• 1) Downloading annual tuberculosis notification data files (2013-2022) from the Datasus website - data in .dbc format.

• 2) Convert downloaded data files from .dbc to .dbf format using TabWin software.

• 3) Combining annual databases (2013-2022)

• 4) Data processing using Microsoft Excel® spreadsheets.

Additionally, indigenous population data from Brazil were obtained from census results provided by the Brazilian Institute of Geography and Statistics (IBGE) for 2010 and 2022. Since intercensal estimates for the Indigenous population were unavailable, geometric (linear) interpolation¹³ was applied to estimate the population for the years 2011 to 2021 using the following formula:

$\begin{array}{c}\frac{\text{Y}-\text{C}}{\text{X}-\text{A}}=\frac{\text{D}-\text{C}}{\text{B}-\text{A}}\\ \text{Y}=\frac{\text{D}-\text{C}}{\text{B}-\text{A}}x(\text{X}-\text{A})+\text{C}\end{array}$

Where: Y = interpolated population estimate for a specific year and location; X = year of population estimate; C = observed population estimate for 2010; D = observed population estimate for 2022; A = year 2010; B = year 2022. Data from 2013 to 2022, by Brazil and macro-regions, were used to calculate the indicators for this research.

Variables management

The dependent variable was classified as follows: “treatment abandonment” and “primary abandonment” were merged into “treatment abandonment,” while “death from tuberculosis” and “death from other causes” were merged into “death.”

The following independent variables underwent modifications: detailed age was categorized into age groups (<19 years; 20-39 years; 40-59 years; ≥60 years); for anti-HIV serology, sputum culture, and sputum smear microscopy, classifications of “inconclusive” and “not applicable,” as well as “unknown” classifications in other variables, were treated as missing data.

Data analysis

Ecological study

Tuberculosis epidemiological indicators stratified by year and macro-region were analyzed for temporal trends using Prais-Winsten regression, according to recommendations by Antunes and Cardoso¹⁴. Study years were entered as independent variables, while crude occurrences and indicators were dependent variables. All indicators were log-transformed (base 10). Based on regression results, annual percentage change (APC) and respective 95% confidence intervals (95% CI) were estimated. Percentage variations were interpreted as follows:

• Increasing trend: Positive APC and statistically significant model (p<0.05)

• Decreasing trend: Negative APC and statistically significant model (p<0.05)

• Stable trend: Non-significant model (p>0.05).

Cross-sectional study

The Chi-square test was used to detect associations between independent and dependent variables. Cramer’s V was employed as an effect size estimate for these associations, using cutoff points suggested by Serdar et al. ¹⁵

Multinomial logistic regression was used to analyze predictors associated with outcomes. The “cure” category was the reference, compared with categories “treatment abandonment” and “death.” To improve the regression model, the entry type variable was aggregated, merging “re-entry after abandonment” and “recurrence” into “retreatment.” Variables showing statistical significance (p<0.05) in the Chi-square test were included in the multivariate model. Multiple logistic regression models were proposed, adjusting for potential confounding and interaction effects. The model with the lowest Bayesian Information Criterion (BIC) and highest Nagelkerke (pseudo-R-square) was selected. Odds ratios (OR) and their respective 95% CIs were used as measures of association.

Data organization and processing were performed using Microsoft Excel (Office 365). Analyses were conducted using Statistical Package for the Social Sciences (SPSS), version 25.

Ethical considerations

The data used in this study are openly accessible, free, and without personal identification, thus exempting the need for Research Ethics Committee approval, supported by Resolution 510 (April 7, 2016) and the Information Access Law No. 12.527 (November 18, 2011).

Results

From 2013 to 2022, 907,436 tuberculosis cases were reported in Brazil, of which 1% (N=9,007) involved indigenous individuals. After data processing, 7,446 confirmed tuberculosis cases among indigenous people in Brazil were considered for this study. Of these, 82.9% resulted in cure, 9.1% abandoned treatment, and 8.0% resulted in death.

Approximately 37.5% (N=2,794) of cases occurred in individuals aged 20-39 years, 58.8% (N=4,381) were male, and 42.7% (N=3,182) resided in the northern region. Clinically, 89.7% (N=6,678) of cases were pulmonary tuberculosis, with 86.9% (N=6,415) classified as new cases. Among behavioral factors, alcohol use predominated at 11.8% (N=6,169), followed by tobacco use at 9.7% (N=5,323). Positive sputum smear microscopy was observed in 47.4% (N=3,420) of cases, while 26.5% (N=1,406) did not undergo the test. Over 70% (N=4,225) of cases underwent directly observed therapy (DOT). Regarding treatment monitoring quality, 31.5% (N=2,347) had insufficient monitoring, 27.2% (N=2,022) regular, and 27.8% (N=2,073) excellent.

Table 1 presents tuberculosis incidence and mortality rates and proportions of cure and treatment abandonment. The highest abandonment proportions and lowest cure proportions were observed in 2022, with the Central-West region leading in abandonment and the South region showing the lowest cure rates.

Table 1
Tuberculosis incidence and mortality rates, cure proportion, and treatment abandonment among indigenous peoples in Brazil, 2013-2022.

Indicators		2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Treatment abandonment proportion per 100,000
	North	7.4	6.1	8.6	5.2	11.7	7.3	10.5	8.8	8.3	14.0
	Northeast	13.6	7.9	5.1	3.3	9.0	6.1	9.0	9.3	6.6	15.8
	Southeast	2.8	12.4	3.8	5.2	3.6	6.7	1.8	6.7	4.9	4.0
	South	14.7	17.9	5.7	7.5	13.0	11.1	13.9	10.3	7.1	18.8
	Center-West	4.3	5.1	6.0	7.1	8.0	6.9	6.6	13.1	10.7	16.4
	Brazil	7.0	7.5	6.5	5.7	9.7	7.1	8.9	9.9	8.5	14.2
Cure proportion per 100,000
	North	87.7	89.1	85.8	86.1	81.5	78.8	81.7	84.8	83.0	72.1
	Northeast	68.6	78.7	83.8	84.3	82.0	80.9	83.5	76.6	77.6	71.1
	Southeast	83.0	74.3	82.9	83.5	73.8	78.7	83.6	76.7	70.7	88.0
	South	64.7	61.5	65.7	82.5	75.9	88.9	63.9	86.2	85.7	50.0
	Center-West	91.8	92.6	85.7	86.4	87.5	89.3	90.9	84.6	77.4	65.7
	Brazil	84.2	85.1	84.3	85.5	81.9	82.3	83.5	82.9	79.9	70.5
Incidence rate per 100,000
	North	63.96	57.73	52.13	50.52	55.79	51.56	56.68	41.73	35.86	14.73
	Northeast	33.89	32.88	26.39	25.99	22.93	22.77	23.51	18.14	12.49	6.24
	Southeast	119.76	83.93	83.29	76.38	60.01	56.36	38.34	40.21	28.83	17.01
	South	30.80	36.60	25.37	39.49	47.42	25.84	34.90	28.82	14.85	13.58
	Center-West	119.28	121.97	178.97	167.64	100.27	113.11	93.56	79.71	76.76	28.48
	Brazil	66.40	60.54	62.88	60.35	51.60	49.55	48.48	38.29	31.87	13.81
Mortality rate per 100,000
	North	3.59	3.13	3.31	4.92	4.46	8.59	5.07	3.06	3.47	2.39
	Northeast	6.84	5.13	3.65	3.94	2.47	3.49	2.20	3.13	2.38	0.94
	Southeast	19.01	13.06	12.81	9.88	16.77	9.54	6.82	8.38	8.24	1.62
	South	8.62	9.76	12.08	4.79	7.11	0.00	9.31	1.15	1.14	5.66
	Center-West	5.08	3.08	15.56	11.64	5.10	4.97	2.69	2.10	10.23	6.00
	Brazil	6.57	5.07	6.51	5.94	5.14	6.16	4.28	3.27	4.18	2.48

• Source: Authors.

Table 2 shows temporal trends in epidemiological indicators and raw occurrences of tuberculosis among the indigenous Brazilian population. Data from 2020 to 2022 showed alterations, likely due to the COVID-19 pandemic. Therefore, trends were analyzed in two distinct series: 1) 2013-2022 (complete series); and 2) 2013-2019 (excluding the pandemic period). Additionally, models using raw occurrences were proposed to avoid bias due to significant growth in the indigenous population between 2010 and 2022.

Table 2
Temporal trend of epidemiological indicators, new tuberculosis cases, and deaths in Brazil and macro-regions from 2013 to 2022 and 2013 to 2019 (before the COVID-19 pandemic).

Characteristics		2013-2022 (with COVID-19)				2013-2019 (without COVID-19)
		APC	95%CI	p-value	D-W	APC	95%CI	p-value	D-W
Treatment abandonment proportion
	North	5.6 ↑	3.1; 8.1	<0.001	2.558	6.6 ↑	2.0; 11.4	0.013	2.432
	Northeast	3.6	-7.8; 16.4	0.506	1.562	-4.1	-24.8; 22.1	0.671	1.659
	Southeast	-3.4	-9.9; 3.6	0.283	2.110	-9.9	-19.5; 0.8	0.062	2.101
	South	0.0	-9.1; 10.0	0.998	1.943	-1.1	-20.4; 22.9	0.901	1.865
	Center-West	13.0 ↑	8.7; 17.5	<0.001	1.910	7.4	-1.8; 17.4	0.094	1.037
	Brazil	6.0 ↑	2.9; 9.1	0.002	2.084	3.5	-2.4; 9.7	0.194	2.279
Cure proportion
	North	-1.5 ↓	-2.6; -0.4	0.013	1.670	-2.1 ↓	-2.8; -1.3	<0.001	2.143
	Northeast	0.0	-2.4; 2.4	0.998	1.412	2.6	-0.8; 6.1	0.107	1.378
	Southeast	-0.3	-1.4; 0.9	0.585	2.654	0.1	-2.1; 2.2	0.941	2.401
	South	1.1	-2.7; 5.0	0.521	1.677	5.0 ↑	1.1; 9.1	0.022	1.630
	Center-West	-3.3 ↓	-6.2; -0.3	0.037	0.962	-0.3	-2.1; 1.6	0.735	1.590
	Brazil	-1.5 ↓	-2.8; -0.1	0.034	1.195	-0.5	-1.1; 0.0	0.057	2.125
Incidence rate per 100,000
	North	-12.6 ↓	-22.4; -1.6	0.031	1.199	-1.9	-5.6; 2.1	0.275	1.518
	Northeast	-16.0 ↓	-24.7; -6.3	0.006	0.885	-6.8 ↓	-10.1; -3.4	0.004	1.669
	Southeast	-17.0 ↓	-20.1; -13.7	<0.001	1.584	-13.9 ↓	-16.5; -11.2	<0.001	1.899
	South	-8.2	-16.6; 1.0	0.072	1.789	0.8	-7.8; 10.3	0.829	2.396
	Center-West	-12.5 ↓	-20.3; -4.0	0.011	1.366	-5.0	-16.0; 7.4	0.335	1.716
	Brazil	-14.2 ↓	-22.7; -4.8	0.009	1.087	-5.4 ↓	-7.4; -3.3	0.001	2.055
Mortality rate per 100,000
	North	-2.5	-14.0; 10.5	0.651	1.785	16.7 ↑	10.8; 22.8	<0.001	2.295
	Northeast	-12.9 ↓	-17.3; -8.4	<0.001	1.846	-14.2 ↓	-18.8; -9.3	<0.001	1.891
	Southeast	-13.7 ↓	-19.8; -7.3	0.002	1.859	-10.5 ↓	-18.9; -1.1	0.035	2.055
	South	-18.9 ↓	-28.9; -7.7	0.007	2.069	-4.5	-20.7; 15.1	0.534	1.744
	Center-West	-2.5	-18.9; 17.2	0.760	1.929	-7.4	-34.8; 31.5	0.599	1.774
	Brazil	-8.1 ↓	-11.7; -4.3	0.001	1.794	-2.2	-5.1; 0.7	0.109	2.508
New cases
	North	-7.5	-18.4; 4.7	0.184	1.176	4.5 ↑	1.0; 8.2	0.021	1.561
	Northeast	-10.8 ↓	-20.5; 0.0	0.049	0.872	-0.6	-3.5; 2.4	0.628	1.721
	Southeast	-15.5 ↓	-18.7; -12.1	<0.001	1.580	-12.3 ↓	-14.9; -9.6	<0.001	1.894
	South	-7.4	-15.8; 2.0	0.103	1.789	1.8	-7.0; 11.3	0.638	2.395
	Center-West	-10.2 ↓	-18.3; -1.3	0.031	1.362	-2.3	-13.7; 10.5	0.648	1.711
	Brazil	-10.1	-19.3; 0.2	0.054	1.069	-0.3	-2.7; 2.0	0.728	1.949
Crude death
	North	3.3	-9.3; 17.6	0.586	1.771	24.4 ↑	18.6; 30.6	<0.001	2.418
	Northeast	-7.5 ↓	-12.0; -2.7	0.007	1.856	-8.3 ↓	-12.9; -3.5	0.007	1.920
	Southeast	-12.2 ↓	-18.4; -5.6	0.003	1.853	-8.8	-17.5; 0.7	0.062	2.055
	South	-18.2 ↓	-28.2; -6.8	0.008	2.069	-3.6	-19.9; 16.2	0.619	1.744
	Center-West	0.1	-16.7; 20.3	0.988	1.929	-4.8	-33.0; 35.2	0.735	1.772
	Brazil	-3.6	-7.7; 0.7	0.086	1.762	2.9	-0.2; 6.2	0.064	2.396

• Note: APC = Annual Percent Change (%); 95% CI = 95% Confidence Interval (lower; upper); D-W = Durbin-Watson after autocorrelation correction by Prais-Winsten; ↓ = Decreasing trend; ↑ = Increasing trend.
• Source: Authors.

The North region showed increased abandonment proportions and reduced cure proportions in both series, as well as increased new cases and deaths from 2013-2019. Mortality rates and deaths decreased in the Northeast region in both series, while the Southeast region saw reductions in incidence and new cases.

Table 3 cross-references sociodemographic and clinical characteristics of tuberculosis cases among indigenous individuals with treatment outcomes, alongside chi-square test results. All variables showed statistical significance (p<0.001), except diabetes mellitus, mental illness, and sputum culture.

Table 3
Cross-analysis of sociodemographic and clinical variables with tuberculosis case outcomes among indigenous peoples in Brazil, 2013-2022.

Características		Treatment outcome				VCramer		p-value
		Abandonment		Death		Cure
		N	%	N	%	N	%
Age group
	≥60 years	77	6.4	213	17.6	921	76.1	0.128	<0.001
	40-59 years	168	10.1	151	9.1	1,348	80.9
	20-39 years	311	11.1	155	5.5	2,328	83.3
	<19 years	115	6.5	80	4.5	1,579	89.0
Sex
	Male	475	10.8	382	8.7	3,524	80.4	0.085	<0.001
	Female	196	6.4	217	7.1	2,652	86.5
Macro-region of residence
	North	251	7.9	273	8.6	2,658	83.5	0.089	<0.001
	Northeast	133	12.5	86	8.1	844	79.4
	Center-West	118	5.6	159	7.6	1,812	86.7
	South	50	15.4	39	12.0	235	72.5
	Southeast	119	15.1	42	5.3	627	79.6
Case entry type
	Transferred	17	9.8	6	3.5	150	86.7	0.144	<0.001
	Recurrence	33	7.4	41	9.2	374	83.5
	Re-entry after abandonment	121	34.6	35	10.0	194	55.4
	New case	500	7.8	473	7.4	5,442	84.8
Chest radiography
	Suspicious	451	8.4	457	8.5	4,485	83.2	0.049	<0.001
	Not performed	172	11.4	92	6.1	1,247	82.5
	Other pathology	7	9.7	14	19.4	51	70.8
	Normal	31	9.5	20	6.1	277	84.5
Clinical form
	Mixed	24	13.5	41	23.0	113	63.5	0.071	<0.001
	Extrapulmonary	56	9.5	67	11.4	467	79.2
	Pulmonary	591	8.8	491	7.4	5,596	83.8
AIDS
	Yes	51	23.9	61	28.6	101	47.4	0.176	<0.001
	No	550	8.3	476	7.2	5,590	84.5
Illicit drugs
	Yes	71	34.6	16	7.8	118	57.6	0.168	<0.001
	No	469	8.3	426	7.5	4,764	84.2
Smoking
	Yes	97	17.0	52	9.1	423	74.0	0.091	<0.001
	No	451	8.5	392	7.4	4,480	84.2
Alcoholism
	Yes	150	18.2	95	11.5	579	70.3	0.133	<0.001
	No	484	7.8	440	7.1	5,245	85.0
Diabetes
	Yes	26	7.4	37	10.6	286	81.9	0.025	0.087
	No	606	9.1	508	7.6	5,565	83.3
Mental illness
	Yes	10	12.5	10	12.5	60	75.0	0.025	0.118
	No	618	8.9	526	7.6	5,773	83.5
Sputum smear microscopy
	Positive	325	9.5	221	6.5	2,874	84.0	0.050	<0.001
	Not performed	198	10.4	177	9.3	1,538	80.4
	Negative	127	6.7	178	9.5	1,578	83.8
Sputum culture
	Positive	127	9.2	100	7.3	1,152	83.5	0.013	0.675
	Not performed	476	9.1	431	8.2	4,341	82.7
	Negative	59	8.9	46	7.0	555	84.1
anti-HIV serology
	Positive	58	23.6	65	26.4	123	50.0	0.137	<0.001
	Not performed	149	10.6	176	12.5	1,081	76.9
	Negative	456	8.0	352	6.2	4,904	85.9
Directly observed treatment
	No	282	16.3	154	8.9	1,290	74.7	0.206	<0.001
	Yes	203	4.8	235	5.6	3,787	89.6
Follow-up indicator
	Insufficient	411	17.5	390	16.6	1,546	65.9	0.241	<0.001
	Regular	172	8.5	163	8.1	1,687	83.4
	Good	62	6.2	32	3.2	910	90.6
	Excellent	26	1.3	14	0.7	2,033	98.1

• Source: Authors.

Treatment abandonment rates were higher among males, while cure rates were higher among females. The Southern region showed high abandonment and mortality rates, while the Central-West region had the lowest abandonment rates. Cases classified as re-entry after abandonment and illicit drug users exhibited abandonment rates around 34%. Mixed clinical forms showed higher mortality rates compared to other forms. Additionally, elevated abandonment and mortality rates were noted among AIDS patients and those with positive HIV serology at diagnosis (Table 3).

Table 4 presents the final multinomial logistic regression model, which was statistically significant [χ²=864.32(52), p<0.001, Nagelkerke R²=0.311].

Table 4
Adjusted odds ratio (OR) of factors associated with death and treatment abandonment among indigenous peoples in Brazil, 2013-2022.

Characteristics		Treatment abandonment			Death
		OR	95%CI		OR	95%CI
			Lower	Upper		Lower	Upper
Age group
	≥60 years	1.27	0.80	2.02	8.35*	4.93	14.15
	40-59 years	1.25	0.83	1.89	3.39*	1.96	5.87
	20-39 years	1.54*	1.07	2.21	1.92*	1.11	3.33
	<19 years	1.00			1.00
Sex
	Male	1.75*	1.32	2.31	1.08	0.79	1.47
	Female	1.00			1.00
Macro-region of residence
	North	0.98	0.68	1.42	3.25*	1.68	6.27
	Northeast	0.79	0.51	1.22	1.29	0.60	2.78
	Center-West	0.59*	0.39	0.89	2.17*	1.10	4.29
	South	1.04	0.57	1.89	2.64*	1.09	6.40
	Southeast	1.00			1.00
Case entry type
	Retreatment	2.60*	1.88	3.59	1.12	0.70	1.79
	New case	1.00			1.00
Chest radiography
	Suspicious	0.71	0.39	1.31	1.11	0.47	2.58
	Not performed	0.87	0.46	1.64	0.83	0.33	2.05
	Other pathology	1.82	0.51	6.46	2.37	0.57	9.92
	Normal	1.00			1.00
AIDS
	Yes	0.67	0.13	3.49	1.85	0.37	9.33
	No	1.00			1.00
Alcoholism
	Yes	2.27*	1.58	3.24	2.17*	1.40	3.37
	No	1.00			1.00
Sputum smear microscopy
	Positive	1.62*	1.14	2.29	0.82	0.56	1.19
	Not performed	1.41	0.97	2.05	0.71	0.48	1.07
	Negative	1.00			1.00
anti-HIV serology
	Positive	2.30	0.49	10.87	4.59	0.98	21.38
	Not performed	1.20	0.84	1.72	1.27	0.82	1.94
	Negative	1.00			1.00
Illicit drugs
	Yes	2.07*	1.24	3.44	0.79	0.29	2.20
	No	1.00			1.00
Smoking
	Yes	1.25	0.84	1.86	0.72	0.42	1.24
	No	1.00			1.00
Clinical form
	Mixed	2.15	0.98	4.72	2.68*	1.21	5.95
	Extrapulmonary	0.56*	0.32	0.96	0.53*	0.28	0.99
	Pulmonary	1.00			1.00
Follow-up indicator
	Insufficient	14.07*	8.25	23.99	42.94*	20.75	88.82
	Regular	7.27*	4.46	11.85	14.66*	7.50	28.67
	Good	3.23*	1.82	5.74	5.07*	2.33	11.04
	Excellent	1.00			1.00
Directly observed treatment
	No	1.82*	1.32	2.52	0.50*	0.34	0.76
	Yes	1.00			1.00

• Note: 1.00 = Reference group; * p<0.05; OR = Odds Ratio; 95% CI = 95% Confidence Interval.
• Source: Authors.

Individuals aged 20-39 had a 54% higher risk of treatment abandonment compared to those under 19 years, while the risk of death increased progressively with age, with elderly individuals facing nearly eight times higher risk. Males had a 75% higher likelihood of treatment abandonment compared to females. Residing in the Central-West reduced abandonment risk by 41%, whereas living in the North, South, and Central-West increased death risk by 3.25, 2.64, and 2.17, respectively, compared to the Southeast region (Table 5).

Table 5
Resident Indigenous population in Brazil, macro-regions, and states: results from the 2010 and 2022 censuses (complete enumeration) and geometric (linear) interpolation for the years 2011 to 2021.

Macro-region		2010*	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022*
North		342,836	377,081	411,327	445,572	479,817	514,063	548,308	582,553	616,799	651,044	685,289	719,535	753,780
	Rondônia	13,076	13,749	14,421	15,094	15,766	16,439	17,111	17,784	18,456	19,129	19,801	20,474	21.146
	Acre	17,578	18,754	19,931	21,107	22,283	23,460	24,636	25,812	26,989	28,165	29,341	30,518	31.694
	Amazonas	183,514	209,132	234,751	260,369	285,988	311,606	337,225	362,843	388,461	414,080	439,698	465,317	490.935
	Roraima	55,922	59,401	62,880	66,359	69,837	73,316	76,795	80,274	83,753	87,232	90,710	94,189	97.668
	Pará	51,217	53,697	56,178	58,658	61,138	63,618	66,099	68,579	71,059	73,539	76,020	78,500	80.980
	Amapá	7,411	7,738	8,065	8,392	8,719	9,046	9,373	9,699	10,026	10,353	10,680	11,007	11.334
	Tocantins	14,118	14,610	15,102	15,594	16,086	16,578	17,071	17,563	18,055	18,547	19,039	19,531	20.023
Northeast		232,739	257,438	282,137	306,836	331,535	356,234	380,934	405,633	430,332	455,031	479,730	504,429	529,128
	Maranhão	38,831	40,359	41,887	43,415	44,943	46,471	47,999	49,526	51,054	52,582	54,110	55,638	57.166
	Piauí	2,944	3,299	3,654	4,009	4,363	4,718	5,073	5,428	5,783	6,138	6,492	6,847	7.202
	Ceará	20,697	23,670	26,643	29,616	32,589	35,562	38,535	41,507	44,480	47,453	50,426	53,399	56.372
	Rio Grande do Norte	2,597	3,358	4,118	4,879	5,639	6,400	7,161	7,921	8,682	9,442	10,203	10,963	11.724
	Paraíba	25,043	25,468	25,893	26,317	26,742	27,167	27,592	28,016	28,441	28,866	29,291	29,715	30.140
	Pernambuco	60,995	64,799	68,604	72,408	76,212	80,016	83,821	87,625	91,429	95,233	99,038	102,842	106.646
	Alagoas	16,291	17,077	17,863	18,650	19,436	20,222	21,008	21,794	22,580	23,367	24,153	24,939	25.725
	Sergipe	5,221	5,178	5,136	5,093	5,051	5,008	4,966	4,923	4,880	4,838	4,795	4,753	4.710
	Bahia	60,120	74,230	88,341	102,451	116,561	130,671	144,782	158,892	173,002	187,112	201,223	215,333	229.443
Southeast		99,137	101,162	103,187	105,211	107,236	109,261	111,286	113,310	115,335	117,360	119,385	121,409	123,434
	Minas Gerais	31,677	32,096	32,514	32,933	33,351	33,770	34,188	34,607	35,025	35,444	35,862	36,281	36.699
	Espírito Santo	9,585	9,987	10,389	10,791	11,193	11,595	11,998	12,400	12,802	13,204	13,606	14,008	14.410
	Rio de Janeiro	15,894	15,986	16,077	16,169	16,261	16,352	16,444	16,536	16,627	16,719	16,811	16,902	16.994
	São Paulo	41,981	43,094	44,206	45,319	46,431	47,544	48,656	49,769	50,881	51,994	53,106	54,219	55.331
South		78,773	79,570	80,368	81,165	81,962	82,760	83,557	84,354	85,152	85,949	86,746	87,544	88,341
	Paraná	26,559	26,885	27,210	27,536	27,861	28,187	28,513	28,838	29,164	29,489	29,815	30,140	30.466
	Santa Catarina	18,213	18,510	18,806	19,103	19,400	19,696	19,993	20,290	20,586	20,883	21,180	21,476	21.773
	Rio Grande do Sul	34,001	34,176	34,351	34,526	34,701	34,876	35,052	35,227	35,402	35,577	35,752	35,927	36.102
Center-West		143,432	148,159	152,886	157,612	162,339	167,066	171,793	176,519	181,246	185,973	190,700	195,426	200,153
	Mato Grosso do Sul	77,025	80,312	83,599	86,886	90,173	93,460	96,747	100,034	103,321	106,608	109,895	113,182	116.469
	Mato Grosso	51,696	52,251	52,806	53,361	53,916	54,471	55,026	55,581	56,136	56,691	57,246	57,801	58.356
	Goiás	8,583	9,494	10,405	11,317	12,228	13,139	14,050	14,961	15,872	16,784	17,695	18,606	19.517
	Federal District	6,128	6,102	6,075	6,049	6,022	5,996	5,970	5,943	5,917	5,890	5,864	5,837	5.811
Brazil		896,917	963,410	1,029,904	1,096,397	1,162,890	1,229,383	1,295,877	1,362,370	1,428,863	1,495,356	1,561,850	1,628,343	1,694,836

• *Observed population from IBGE censuses.
• Source: Brazilian Institute of Geography and Statistics (IBGE) - 2010 and 2022 censuses.

Cases classified as retreatment showed 2.60 times higher chances of treatment abandonment compared to new cases. Alcohol use significantly predicted treatment abandonment and death, with approximately twice the risk compared to non-users. Illicit drug use doubled the risk of abandonment, while positive sputum smear microscopy increased risk by 62%. Indigenous individuals not receiving DOT had an 82% higher chance of abandonment and a lower risk of death (Table 4).

Discussion

The present study provided substantial evidence regarding characteristics associated with unfavorable tuberculosis outcomes among the indigenous population and temporal trends of epidemiological indicators. Through the findings, it was possible to identify that the situation in the North region is concerning, showing annual growth in new cases and deaths, along with reductions in cure and treatment abandonment proportions. It was also observed that the proportion of treatment abandonment in Brazil exceeded the level recommended by the Ministry of Health¹⁶ (<5%), while the cure rate was very close (>85%). Additionally, it was noted that most cases occurred among adults, predominantly from the North region. It was possible to identify further that treatment abandonment and death were associated with age, region of residence, alcohol, and illicit drug use, case retreatment, lack of directly observed therapy (DOT), and inadequate follow-up.

Most tuberculosis cases reported in this research occurred among Indigenous adults aged 20-39 years, followed by young individuals aged 0-19 years, collectively accounting for over 60% of cases. These findings are similar to those reported in the general population, where the disease incidence among youth is lower than among adults in Brazil¹⁷ and internationally¹⁸. Among the indigenous population, the disease burden in children and adolescents is lower compared to adults, with a disproportionately higher percentage of cases among adults relative to other age groups⁷. The significant tuberculosis burden among children and adolescents stands out as a challenge for public health due to difficulties in diagnosis and treatment, given the nonspecific nature of symptoms and challenges in performing sputum smear microscopy and sputum culture tests¹⁹. However, it should not be overlooked that the disease burden among adults is equally concerning, as disparities in tuberculosis burden across all ages reinforce inequalities, highlighting the need for inclusive interventions tailored to the needs of vulnerable populations.

High rates and greater likelihood of treatment abandonment were observed among men, whereas women presented higher cure proportions. This pattern was also identified in the general population²⁰ and a study among indigenous populations²¹. Men’s low interest in caring for their own health, possibly influenced by culture or lifestyle, might explain this phenomenon²². Additionally, in some Indigenous communities, men often undertake physically demanding responsibilities, such as deforestation and clearing areas for planting²³.

Approximately 20% to 25% of tuberculosis cases did not undergo chest radiography or sputum smear microscopy, more than 70% did not perform sputum culture, and around 19% did not undergo anti-HIV serology testing. These findings are similar to those reported in another study¹⁷. This gap in diagnostic methods can compromise effective tuberculosis control and hinder achieving goals set by the World Health Organization and the Ministry of Health. According to the Manual of Recommendations for Tuberculosis Control, diagnosing Indigenous populations requires a comprehensive approach due to the prevalence of non-tuberculous mycobacteria and fungi, necessitating sputum culture or TRM-TB, chest radiography, and HIV testing¹². However, the absence of testing in specific scenarios does not necessarily indicate diagnostic failures, considering the clinical peculiarities of particular groups, such as children and extrapulmonary tuberculosis cases.

Directly observed therapy (DOT) was implemented in more than 70% of the studied population, with an 82% higher risk of abandonment among those who did not receive DOT. This strategy is crucial for reducing abandonment rates; thus, the results of this study align with existing literature²⁴. Notably, DOT coverage is higher among indigenous populations than the general population²⁵. Importantly, this approach fosters and strengthens the professional-patient relationship, encouraging treatment completion, especially within vulnerable populations. Therefore, intensifying this strategy should be prioritized due to its significant role in reducing tuberculosis cases, drug-resistant tuberculosis, and, consequently, mortality.

Since 2012, the National Tuberculosis Control Program (PNCT) has strengthened specific actions for the indigenous population through partnerships with CGAPSI/Sesai, CGPDS, municipal and state tuberculosis control programs, and the Special Indigenous Health Districts. The PNCT aims to improve access and provide differentiated care for disease control by training Indigenous healthcare professionals, empowering presidents of Indigenous health councils, and including Funai professionals in tuberculosis control monitoring²⁶. These aspects are essential for controlling this disease among indigenous peoples and may justify the higher rates of directly observed therapy (DOT).

This study identified an annual reduction in tuberculosis incidence rates. Between 2011 and 2017, seven states showed a decrease in incidence rates of up to 38.48% per year⁴. The significant increase in the indigenous population recorded in the 2022 census compared to 2010, nearly doubling in size, may explain this phenomenon. This change is primarily attributed to methodological factors such as more excellent census coverage, expanded questionnaire content, and an increased number of individuals identifying as indigenous²⁷. Additionally, it is possible that the COVID-19 pandemic impacted the quality of tuberculosis notifications^28,29. Excluding these two factors, it was observed that only the Southeast region showed a reduction in new cases and the Northeast in deaths, whereas the North region showed increases in both. Therefore, the increase in the estimated Indigenous population and the decrease in notifications during the pandemic period influenced the epidemiological indicators and temporal trends. Advancing census techniques better to map indigenous peoples’ demographic and epidemiological situation is crucial, and future studies should exercise caution when using demographic data.

Treatment abandonment and mortality rates were reduced as the quality of case monitoring improved, whereas cure rates increased accordingly. Moreover, the risk of unfavorable tuberculosis outcomes decreased as the monitoring indicator improved. When considering national recommendations for tuberculosis case follow-up, this operational indicator, despite being empirical and not formally validated, demonstrated a logical and expected pattern consistent with the national tuberculosis control program guidelines¹². This finding indicates that patients receiving adequate follow-up have lower chances of experiencing adverse outcomes and higher chances of being cured, as demonstrated in another study³⁰.

Another risk factor for treatment abandonment identified in this study was case retreatment. These findings are similar to those observed in the general population, such as a Mato Grosso do Sul study, which reported a higher risk of abandonment among retreatment cases³¹. The literature highlights that retreatment predisposes individuals to treatment abandonment, drug resistance development, and increased disease severity³². Therefore, greater efforts are needed to actively search for individuals who miss treatment and intensify the supervision of indigenous patients in this situation to prevent further complications.

In this study, Indigenous individuals who consumed alcohol had higher chances of treatment abandonment and death from tuberculosis compared to those who did not drink alcohol. A systematic review reported findings similar to those observed here³³. Coimbra Júnior et al. ³⁴ highlighted in their book that although infectious diseases remain significant causes of morbidity and mortality among Indigenous peoples, in recent years, an increase in so-called “social diseases,” such as depression and alcoholism, has been observed. The study by Garcia and Freitas³⁵ found that the overall proportion of alcohol consumption among Indigenous individuals aged 18 years and older was 12.6%. Alcohol consumption can weaken the immune system, impair cognitive function, and exacerbate social and economic challenges, making treatment adherence more complex and consequently increasing the risk of death. Therefore, it is crucial to adopt social mobilization measures and health education initiatives that promote the prevention and reduction of alcohol consumption among indigenous populations.

A higher risk of treatment abandonment was also identified among Indigenous individuals who used illicit drugs. The literature still presents gaps regarding the association between substance use and tuberculosis outcomes in this population. However, it is well known that this combination has severe public health consequences. The inhalation of drugs such as cocaine exacerbates pulmonary complications, while injectable drug use increases the risk of HIV co-infection. Drug users are more likely to experience delays in diagnosis and treatment, treatment abandonment, drug resistance, and overall health deterioration³⁶. Additionally, indigenous cultural practices may include the use of local psychoactive substances, possibly unknown, which could also impact health. Therefore, it is essential to implement awareness and education initiatives while facilitating access to rehabilitation programs for these users.

A higher risk of death from tuberculosis was observed among Indigenous individuals residing in the North region compared to those in the Southeast. At the same time, the South had the highest proportion of treatment abandonment and mortality among all macro-regions. This may be explained by the demographic distribution of indigenous peoples in Brazil. According to the latest IBGE census (2022), the North region accounts for 44.5% (N=753,357) of the Indigenous population, whereas the South represents only 5.2% (N=88,097)¹. Another possible explanation is related to social and sanitary conditions, as well as the limited access to secondary and tertiary healthcare services, which hinder adequate treatment for indigenous peoples in the North. On the other hand, the South has a higher concentration of children and adolescents, which contributes to the spread of the disease³⁷. Combined with deficiencies in case monitoring and identification, this may justify the higher rates of treatment interruptions and deaths in this region³⁸.

The main limitation of this study relates to the use of secondary data, which is subject to underreporting of cases, data incompleteness, errors in tuberculosis case diagnosis and classification, and underreporting of deaths. Another limitation was the addition, exclusion, or modification of variables such as illicit drug use, smoking, and DOT³⁹, which may have contributed to increased data incompleteness. However, the variables and analyses employed in this study help mitigate these challenges. Misclassification errors in the race/ethnicity variable may also be present, as it is self-reported. Additionally, the COVID-19 pandemic, from 2020 to 2022, may have negatively impacted the quality of notifications, influencing trend estimates. To minimize interpretation bias, models that excluded this period were proposed. Furthermore, concerns arise regarding the reliability of information in the variables “mental illness” and “alcoholism,” considering the culturally distinct nature of this population. Finally, the increase in the estimated Indigenous population due to methodological improvements in the IBGE census influenced tuberculosis indicators in the present study. This factor may have also affected the observed trends; however, analyses using raw case numbers were performed to minimize bias.

Despite these limitations, the findings of this study are of great value to public health and epidemiology. The approach adopted provides national representation and includes all tuberculosis cases reported among indigenous populations over a decade. The data analysis models allowed for the identification of factors associated with treatment abandonment and mortality in an integrated manner, enhancing the accuracy and reliability of the findings. These aspects have not been explored in previous studies. Therefore, this research is expected to contribute to expanding current knowledge about tuberculosis among indigenous populations and indirectly assist in achieving the goals set by the National Tuberculosis Control Plan and the Sustainable Development Goals of the 2030 Agenda.

Conclusion

It was observed that, during the period from 2013 to 2022, the cure rate for confirmed tuberculosis cases among indigenous individuals was close to the national target, while treatment abandonment and mortality rates were below the recommended thresholds. The North region of the country presented the most concerning temporal scenario, with an increase in the proportion of treatment abandonment, a decrease in the cure rate, and a rise in the number of new cases and deaths. The risk of treatment abandonment was higher among individuals aged 20-39 years, males, those undergoing retreatment, those who consumed alcohol or illicit drugs, those with positive sputum smear microscopy, those who did not undergo directly observed therapy (DOT), and those with poor treatment monitoring quality. The risk of death was higher among older indigenous individuals, those residing in the North, Central-West, and South regions, those who consumed alcohol, those affected by the mixed clinical form (pulmonary + extrapulmonary), and those with reduced treatment monitoring quality.

Acknowledgments

This study was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Funding Code 001.

References

• ¹ Instituto Brasileiro de Geografia e Estatística (IBGE). Censo 2022 | População residente por cor ou raça e total de pessoas indígenas [Internet]. [acessado 2024 mar 5]. Disponível em: https://www.ibge.gov.br/estatisticas/sociais/trabalho/22827-censo-demografico-2022.html?edicao=38698&t=resultados
  » https://www.ibge.gov.br/estatisticas/sociais/trabalho/22827-censo-demografico-2022.html?edicao=38698&t=resultados
• ² Anderson I, Robson B, Connolly M, Al-Yaman F, Bjertness E, King A, Tynan M, Madden R, Bang A, Coimbra CE Jr, Pesantes MA, Amigo H, Andronov S, Armien B, Obando DA, Axelsson P, Bhatti ZS, Bhutta ZA, Bjerregaard P, Bjertness MB, Briceno-Leon R, Broderstad AR, Bustos P, Chongsuvivatwong V, Chu J, Deji, Gouda J, Harikumar R, Htay TT, Htet AS, Izugbara C, Kamaka M, King M, Kodavanti MR, Lara M, Laxmaiah A, Lema C, Taborda AM, Liabsuetrakul T, Lobanov A, Melhus M, Meshram I, Miranda JJ, Mu TT, Nagalla B, Nimmathota A, Popov AI, Poveda AM, Ram F, Reich H, Santos RV, Sein AA, Shekhar C, Sherpa LY, Skold P, Tano S, Tanywe A, Ugwu C, Ugwu F, Vapattanawong P, Wan X, Welch JR, Yang G, Yang Z, Yap L. Indigenous and tribal peoples' health (The Lancet-Lowitja Institute Global Collaboration): a population study. Lancet 2016; 388(10040):131-157.
• ³ Paiva BL, Azeredo JQ, Nogueira LMV, Santos BO, Rodrigues ILA, Santos MNA. Distribuição espacial de tuberculose nas populações indígenas e não indígenas do estado do Pará, Brasil, 2005-2013. Esc Anna Nery 2017; 21(4):e20170135.
• ⁴ Ferreira TF, Santos AM, Oliveira BLCA, Caldas AJM. Tendência da tuberculose em indígenas no Brasil no período de 2011-2017. Cien Saude Colet 2020; 25(10):3745-3752.
• ⁵ Brasil. Ministério da Saúde (MS). Populações mais vulneráveis - Tuberculose [Internet]. [acessado 2023 jun 6]. Disponível em: https://www.gov.br/saude/pt-br/assuntos/saude-de-a-a-z/t/tuberculose/populacoes-mais-vulneraveis
  » https://www.gov.br/saude/pt-br/assuntos/saude-de-a-a-z/t/tuberculose/populacoes-mais-vulneraveis
• ⁶ León-Giraldo H, Rivera-Lozada O, Castro-Alzate ES, Aylas-Salcedo R, Pacheco-López R, Bonilla-Asalde CA. Factors Associated with Mortality with Tuberculosis Diagnosis in Indigenous Populations in Peru 2015-2019. Int J Environ Res Public Health 2022; 19(22):15019.
• ⁷ Viana PVS, Codenotti SB, Bierrenbach AL, Basta PC. Tuberculose entre crianças e adolescentes indígenas no Brasil: fatores associados ao óbito e ao abandono do tratamento. Cad Saude Publica 2019; 35(Supl. 3):e00074218.
• ⁸ Malacarne J, Rios DP, Silva CM, Braga JU, Camacho LA, Basta PC. Prevalence and factors associated with latent tuberculosis infection in an indigenous population in the Brazilian Amazon. Rev Soc Bras Med Trop 2016; 49(4):456-464.
• ⁹ Rios DP, Malacarne J, Alves LC, Sant'Anna CC, Camacho LA, Basta PC. Tuberculosis in indigenous peoples in the Brazilian Amazon: an epidemiological study in the Upper Rio Negro region. Rev Panam Salud Publica Pan Am J Public Health 2013; 33(1):22-29.
• ¹⁰ Benchimol EI, Smeeth L, Guttmann A, Harron K, Moher D, Petersen I, Sørensen HT, von Elm E, Langan SM; RECORD Working Committee. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) Statement. PLoS Med 2015; 12(10):e1001885.
• ¹¹ Sociedade Brasileira de Pneumologia e Tisiologia. III Diretrizes para Tuberculose da Sociedade Brasileira de Pneumologia e Tisiologia. J Bras Pneumol 2009; 35:1018-1048.
• ¹² Brasil. Ministério da Saúde (MS). Manual de recomendações para o controle da tuberculose no Brasil [Internet]. 2ª ed. 2019 [acessado 2024 mar 18]. Disponível em: https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/svsa/tuberculose/manual-de-recomendacoes-e-controle-da-tuberculose-no-brasil-2a-ed.pdf/view
  » https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/svsa/tuberculose/manual-de-recomendacoes-e-controle-da-tuberculose-no-brasil-2a-ed.pdf/view
• ¹³ Hazewinkel M. Linear interpolation. In: Hazewinkel M, editor. Encyclopaedia of Mathematics. Dordrecht: Springer Netherlands; 1997. p. 102-140.
• ¹⁴ Antunes JLF, Cardoso MRA. Uso da análise de séries temporais em estudos epidemiológicos. Epidemiol Serv Saude 2015; 24:565-576.
• ¹⁵ Serdar CC, Cihan M, Yücel D, Serdar MA. Sample size, power and effect size revisited: simplified and practical approaches in pre-clinical, clinical and laboratory studies. Biochem Med (Zagreb) 2021; 31(1):010502.
• ¹⁶ Brasil. Ministério da Saúde (MS). Brasil livre da tuberculose - Plano nacional pelo fim da tuberculose como problema de saúde pública: estratégias para 2021-2025 [Internet]. [acessado 2023 ago 8]. Disponível em: https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/svsa/tuberculose/plano-nacional-pelo-fim-da-tuberculose-como-problema-de-saude-publica_-estrategias-para-2021-2925.pdf/view
  » https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/svsa/tuberculose/plano-nacional-pelo-fim-da-tuberculose-como-problema-de-saude-publica_-estrategias-para-2021-2925.pdf/view
• ¹⁷ Tavares CM, Cunha AMS, Gomes NMC, Lima ABA, Santos IMR, Acácio MS, Santos DM, Souza CDF. Tendência e caracterização epidemiológica da tuberculose em Alagoas, 2007-2016. Cad Saude Colet 2020; 28(1):107-115.
• ¹⁸ Dhamnetiya D, Patel P, Jha RP, Shri N, Singh M, Bhattacharyya K. Trends in incidence and mortality of tuberculosis in India over past three decades: a joinpoint and age-period-cohort analysis. BMC Pulm Med 2021; 21(1):375.
• ¹⁹ World Health Organization (WHO). WHO consolidated guidelines on tuberculosis: Module 5 - Management of tuberculosis in children and adolescents [Internet]. [acessado 2023 jun 23]. Disponível em: https://www.who.int/publications/i/item/9789240046764
  » https://www.who.int/publications/i/item/9789240046764
• ²⁰ Ibrahim LM, Hadejia IS, Nguku P, Dankoli R, Waziri NE, Akhimien MO, Ogiri S, Oyemakinde A, Dalhatu I, Nwanyanwu O, Nsubuga P. Factors associated with interruption of treatment among Pulmonary Tuberculosis patients in Plateau State, Nigeria. 2011. Pan Afr Med J 2014; 17:78.
• ²¹ Escobar AL, Coimbra Jr. CEA, Camacho LA, Portela MC. Tuberculose em populações indígenas de Rondônia, Amazônia, Brasil. Cad Saude Publica 2001; 17(2):285-298.
• ²² Damasceno JCTRA, Rodrigues ES, Borges WD, Camargo CL, Sousa AR. Masculinidades indígenas e saúde: análise qualitativa do cuidado tradicional no contexto pandêmico e pós-pandêmico da COVID-19. Cien Saude Colet 2025; 30(11):e14652023.
• ²³ Gama EVS, Marques CTS, Carvalho AJA, Silva F. Divisão de trabalho entre homens e mulheres na Aldeia Indígena Tupinambá de Serra do Padeiro, Buerarema-BA. Rev Bras Agroecol 2007; 2(2):1669-1673.
• ²⁴ Furlan MCR, Oliveira SP, Marcon SS. Fatores associados ao abandono do tratamento de tuberculose no estado do Paraná. Acta Paul Enferm 2012; 25:108-114.
• ²⁵ Brasil. Ministério da Saúde (MS). Boletim epidemiológico Tuberculose [Internet]. 2022 [acessado 2022 ago 25]. Disponível em: https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/boletins/epidemiologicos/especiais/2022/boletim-epidemiologico-de-tuberculose-numero-especial-marco-2022.pdf
  » https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/boletins/epidemiologicos/especiais/2022/boletim-epidemiologico-de-tuberculose-numero-especial-marco-2022.pdf
• ²⁶ Brasil. Ministério da Saúde (MS). Boletim epidemiológico: tuberculose [Internet]. [acessado 2023 jun 29]. Disponível em: https://bvsms.saude.gov.br/bvs/periodicos/boletim_epidemiologico_numero_2_2014.pdf
  » https://bvsms.saude.gov.br/bvs/periodicos/boletim_epidemiologico_numero_2_2014.pdf
• ²⁷ Simoni AT, Guimarães BN, Santos RV. "Nunca mais o Brasil sem nós": povos indígenas no Censo Demográfico 2022. Cad Saude Publica 2024; 40(4):e00232223.
• ²⁸ Borges TS, Santos AL, Osaki SC, Dias CHFZ. Notificações de tuberculose no período pré-pandêmico e pandêmico da Covid-19 no Estado do Paraná. Arq Cien Saude UNIPAR 2023; 27(4):1825-1844.
• ²⁹ Batista JFC, Santos VSO, Jesus CVF, Lima SO. Série temporal dos casos e dos desfechos do tratamento contra tuberculose em Sergipe, 2012-2021. Rev Bras Epidemiol 2023; 26:e230041.
• ³⁰ Orellana JDY, Gonçalves MJF, Basta PC. Características sociodemográficas e indicadores operacionais de controle da tuberculose entre indígenas e não indígenas de Rondônia, Amazônia Ocidental, Brasil. Rev Bras Epidemiol 2012; 15:714-724.
• ³¹ Bezerra WSP, Lemos EF, Prado TN, Kayano LT, Souza SZ, Chaves CEV, Paniago AMM, Souza AS, Oliveira SMDVL. Risk Stratification and Factors Associated with Abandonment of Tuberculosis Treatment in a Secondary Referral Unit. Patient Prefer Adherence 2020; 14:2389-2397.
• ³² Silva TC, Matsuoka PFS, Aquino DMC, Caldas AJM. Fatores associados ao retratamento da tuberculose nos municípios prioritários do Maranhão, Brasil. Cien Saude Colet 2017; 22(12):4095-4104.
• ³³ Rehm J, Samokhvalov AV, Neuman MG, Room R, Parry C, Lönnroth K, Patra J, Poznyak V, Popova S. The association between alcohol use, alcohol use disorders and tuberculosis (TB). A systematic review. BMC Public Health 2009; 9:450.
• ³⁴ Coimbra Junior CEA, Santos RV, Escobar AL, organizadores. Epidemiologia e saúde dos povos indígenas no Brasil. Rio de Janeiro: Editora Fiocruz, Abrasco; 2003.
• ³⁵ Garcia LP, Freitas LRS. Consumo abusivo de álcool no Brasil: resultados da Pesquisa Nacional de Saúde 2013. Epidemiol Serv Saude 2015; 24(2):227-237.
• ³⁶ Silva DR, Muñoz-Torrico M, Duarte R, Galvão T, Bonini EH, Arbex FF, Arbex MA, Augusto VM, Rabahi MF, Mello FCQ. Fatores de risco para tuberculose: diabetes, tabagismo, álcool e uso de outras drogas. J Bras Pneumol 2018; 44(2):145-152.
• ³⁷ Mendes AM, Bastos JL, Bresan D, Leite MS. Situação epidemiológica da tuberculose no Rio Grande do Sul: uma análise com base nos dados do Sinan entre 2003 e 2012 com foco nos povos indígenas. Rev Bras Epidemiol 2016; 19(3):658-669.
• ³⁸ Gava C, Malacarne J, Rios DPG, Sant'Anna CC, Camacho LA, Basta PC. Tuberculosis in indigenous children in the Brazilian Amazon. Rev Saude Publica 2013; 47(1):77-85.
• ³⁹ Canto VBD, Nedel FB. Completude dos registros de tuberculose no Sistema de Informação de Agravos de Notificação (Sinan) em Santa Catarina, Brasil, 2007-2016. Epidemiol Serv Saude 2020; 29(3):e2019606.

• Data availability statement
  The data sources adopted in the research are indicated in the article’s body.

• Chief editors:
  Maria Cecília de Souza Minayo, Romeu Gomes, Antônio Augusto Moura da Silva, Vania de Matos Fonseca

Data availability

The data sources adopted in the research are indicated in the article’s body.

Publication Dates

• Publication in this collection
  03 Aug 2026
• Date of issue
  July 2026

History

• Received
  19 July 2024
• Accepted
  28 Feb 2025
• Published
  02 Mar 2025