Printer Friendly

Determinants of tuberculosis in Brazil: from conceptual framework to practical application/Determinantes de la tuberculosis en el Brasil: del marco conceptual a la aplicacion practica.

Historically, which factors have determined the emergence of new cases of tuberculosis (TB) and poor TB treatment outcomes? Studies that attempt to address this question point to the multicausal framework of the disease and describe the biological and molecular aspects of Mycobacterium tuberculosis, the host-immune response, conditions over the life-course to which individuals are exposed, and relevant political, economic, and sociocultural issues (1, 2).

The causal complex of TB may be expressed as follows: proximal causes are related to aspects of the host-pathogen interaction; intermediate causes are intricately related to health services and TB control policies; and distal causes are those that describe the socioeconomic and cultural contexts that give rise to and exacerbate the disease. Based on these assumptions, Maciel (2) proposes a theoretical model to determine the disease, taking into account factors previously shown to be relevant in seminal studies of the disease in Brazil (3-8). An adaptation of Maciel's model (2) is shown in Figure 1. It is based on a proposal of interdependence that combines three axes of vulnerability: the individual, the health system, and the social context (9). However, in practice, this model is difficult to implement, as there are further temporal and spatial networks that simultaneously act on the individual.

TB is a priority for the Ministry of Health of Brazil. The main strategy for TB control is thorough implementation of Directly Observed Treatment Short-course (DOTS), in addition to a firm political commitment, case detection by sputum smear, standardized treatment, a regular and uninterrupted supply of standardized anti-tubercular drugs, and a consistent and reliable case-reporting system (10). The Sistema de Informacao de Agravos de Notificacao (Notifiable Disease Surveillance System database; SINAN) offers detailed information on patients diagnosed with incident TB and is considered a proxy of disease incidence (10). Thus, this study aimed to propose and implement a conceptual analytical model for TB determination based on data available from SINAN.

MATERIALS AND METHODS

This was a cross-sectional study based on data from SINAN, including all confirmed, incident TB cases reported in Brazil between 2007 and 2011. SINAN was developed in Brazil in the 1990s to collect and disseminate important notifiable disease information to health professionals and government bodies. With respect to TB, SINAN includes sociodemographic and health characteristics, as well as features of TB and information describing TB treatment (11, 12). Several analyses of the completeness and validity of SINAN data have been performed and suggest a sufficient number of cases and a high degree of reliability (13).

Theoretical model

Several models were fit and adapted to describe TB determinants based on the theoretical model developed by Maciel (2) (Figure 1). The model that best explained the causal network for the outcome is shown in Figure 2, where TB determinants are arranged hierarchically according to previous knowledge describing their determination relation and temporal ordering.

Study population

In 2007-2011, SINAN received reports of 432 958 TB cases. Of these, 318 465 individual's treatment outcomes were included in the analysis; 114 493 were excluded due to incomplete treatment outcome or other inadequate data. TB treatment outcome was dichotomized (cure / treatment failure), as well as a multinomial variable (cure / abandonment / death from TB / death from other causes / development of multidrug-resistant tuberculosis [MDR-TB]).

Variables

In order to evaluate the fit of the proposed theoretical model, the following variables, all of which are available in SINAN, were studied and incorporated into both the nucleus and the multiple axes of the model:

* Individual vulnerability axis: patient age; education level; skin color/ ethnicity; the comorbidities of mental illness, diabetes, alcoholism, AIDS/ HIV, and other diseases; and pregnancy.

* Programmatic vulnerability axis: initial smear results (as well as presence or absence of the diagnostic test); initial culture results; tuberculin skin test (TST); thoracic x-ray; histopathology; prescribed anti-TB drugs; whether or not disease transmission was due to the patient's professional environment; number of patient contacts indicated by patient and diagnosis; DOTS; and smear-tracking results.

* Social vulnerability axis: location of patient residence; area of residence; institutionalization; and occupation.

* Nucleus variables were: type of treatment; clinical form of TB; and TB treatment outcome (14).

Data analysis

Adjusted analyses were used to assess the model fit. Hierarchical logistic regression was used to model the dichotomous TB outcome, whereas hierarchical polytomous regression was used for multinomial TB outcome. Polytomous regression is a useful technique to simultaneously model the probabilities of an outcome with multiple categories. The method can uncover associations that are otherwise obscured by transformation of a multinomial outcome variable into a dichotomous outcome (15, 16). The "cure" category was included as an outcome for both analyses. This type of modeling is relatively robust, lacking restrictions on the distribution of the outcome and has been used in models that depict the outcome as either ordinal or nominal polytomous (17). As such, when the outcome of TB treatment is categorized, a multinomial approach also allows for inclusion of multiple covariates in the same model. This allows for the identification of individual risk factors within each subgroup of the outcome (15, 18).

In order to overcome the restrictions of linear models (otherwise rigid and compartmentalized), hierarchical models have been used when distal factors (antecedents) influence the intermediate factors that assert an influence on more proximal factors (those that act directly on the outcome). As such, it is possible to include a variety of factors according to their temporal ordering in addition to their influence on the outcome, without necessitating ignorance of the interconnectedness of the antecedents.

The process of statistical analysis and interpretation of results is different in hierarchical modeling. The introduction of each of the variables occurs in stages, starting with the variables at the most distal level and simultaneously introducing variables at the same level. The effect of each variable on the outcome is interpreted as having been adjusted for all the variables that belong to the hierarchical levels above it (distal), as well as for the variable's effects that coexist on the same level. Table 1 shows the variables introduced at each level as illustrated in Figure 2.

For the various comorbidities of alcoholism, diabetes, mental illness, and HIV/ AIDS, as well as the assignment of DOTS, a field left blank or marked as 'ignored' was considered an absence of the respective comorbidity or not being assigned to DOTS. Results are presented as odds ratios (OR) with 95% Confidence Intervals (95%CI). Analyses were performed using Stata[R] version 12.0 (StataCorp LP, College Station, Texas, United States).

Ethics

The Research Ethics Committee at the Universidade Federal do Espirito Santo's Center for Health Sciences (Vitoria, Espirito Santo, Brazil) approved this study. To ensure patient anonymity and privacy, no personal identifiers were included in the analysis.

RESULTS

This study analyzed 318 465 TB cases reported in Brazil from 2007-2011, of which 222 186 (69.8%) were classified as "cured" and 96 279 (30.2%) as "treatment failure." Among the latter, 37 604 (11.8%) abandoned treatment; 13 193 (4.1%) died due to TB; 15 440 (4.8%) died due to causes other than TB; 28 848 (9.1%) were transferred to another municipality; and 1 194 (0.4%) developed MDR TB.

Adjusted Odds Ratios (OR) for each of characteristics associated with TB outcome are shown in Table 2. Analyses of the outcome with artificially-created grouped categories often resulted in a spurious association.

Individuals more than 60 years of age had higher odds of death from TB (OR 7.97; 95%CI = 7.15-8.89) or death from other causes (OR 4.8; 95%CI = 7.24-8.93). However, these individuals also were less likely to abandon their treatment (OR 0.57; 95%CI = 0.54-0.61) compared to those who were less than 20 years old.

In the dichotomous model, males were less likely to abandon treatment (OR 0.76; 95%CI = 0.75-0.77). In the polytomous model, males were more likely to experience all adverse outcomes, except development of MDR TB (OR 1.1; 95%CI 0.89-1.14), but this result was not statistically significant.

Individuals institutionalized in prisons were less likely to abandon treatment (OR 0.68; 95%CI = 0.64-0.71), die from TB (OR 12.56; 95%CI = 0.50-0.63), or die from any other causes (OR 0.70; 95%CI = 0.64-0.76).

In the polytomous analysis, alcoholism was associated with a greater occurrence of all outcomes. The co-presence of diabetes and tuberculosis was also associated with a higher rate of treatment compliance (OR 0.68; 95%CI = 0.64-0.73). Among individuals with HIV/AIDS, the odds of death from TB were 4.7 (95%CI = 3.88-4.27), and from other causes, 11.10 (95%CI = 10.65-11.54)

When compared to individuals identified as new cases, those who had relapsed after previous treatment (9.43 OR; 95% CI = 8.13-10.95), had returned after abandoning treatment (OR 9.14; 95% CI = 7.67 10.88), or had transferred to another municipality (OR 6.44; 95%CI = 5.437.63) were more likely to develop MDR TB. In contrast, individuals with extrapulmonary TB had 68% lower odds of developing MDR TB (95%CI = 0.21-0.47).

In the polytomous model, individuals assigned to DOTS were more likely to be cured than those with other TB treatment outcomes: abandonment (OR 0.52; 95%CI = 0.51-0.54), death from TB (OR 0.50; 95%CI = 0.48-0.53), death from other causes (OR 0.68; 95%CI = 0.650.71), and development of MDR TB (OR 0.84; 95%CI = 0.75-0.95).

DISCUSSION

The use of conceptual models has become widespread. Models are widely considered to be a fundamental element in the research process and in epidemiology because they simplify and facilitate a thorough understanding of the various contexts and multi-dimensional mechanisms that lead to disease (19). Although the first conceptual models were based on the natural history of communicable diseases (1), today they are used in fields that range from health policy evaluation (20) to both communicable (21, 22) and non-communicable (3) diseases, and other disciplines that examine the influence of natural phenomena on the life of the individual (23).

Since identification of M. tuberculosis as the causal agent for the disease, various models for its determination have been proposed. These have ranged from basic models based on a single etiology (24, 25) to more advanced models that attempt to account for multiple causes, such as the social determinants of health (2, 26). However, most studies do not consider these models and lack details on how the analysis was carried out. This fact has dramatically hindered a comprehensive understanding of the causal chain of determination of TB, as well as the best interpretation of the model's results and respective measures of association.

TB treatment outcomes in Brazil

TB incidence in Brazil has decreased roughly 1.4% annually since 2001. In 2010, TB incidence was reported as 37.7 / 100 000 individuals and TB mortality as 2.4 / 100 000 (10). TB treatment outcome has been thoroughly studied by researchers in Brazil (4, 27, 28) as well as those hailing from countries and regions with dramatically different TB disease burdens (6, 29). Despite this, few studies have explicitly considered different categories of TB treatment outcomes (30, 31).

The findings of the present study showed several factors were associated with being less likely to abandon TB treatment: being older in age; self-identifying as indigenous or of Asian ethnicity; having participated in or completed higher education (beyond high school); not living in a rural area; being institutionalized in a prison; having diabetes; having extrapulmonary TB; and being assigned to DOTS.

In contrast, the following factors increased the likelihood of abandoning treatment: male gender; being institutionalization in a place other than a prison (e.g., orphanage or psychiatric clinic); having alcoholism or HIV/AIDS; being a relapsed case; returning to treatment after having abandoned it; having unknown treatment completion status; and a positive sputum culture.

Death from TB was more likely among those who were older than 20 years; male; of black or brown ethnicity; not being a new TB case; being illiterate; living in a rural area; being institutionalized in a place other than prison (e.g., orphanage or psychiatric clinic); having alcoholism, mental illness, or HIV/AIDS; a positive sputum culture; the pulmonary form of the disease; and not assigned to DOTS.

With respect to non-TB related causes of death, the odds were higher for those who were older than 20 years; male; of Caucasian ethnicity; illiterate; living in a rural area; being institutionalized in a place other than a prison (e.g., an orphanage or psychiatric clinic); having alcoholism, diabetes, mental illness, or HIV / AIDS; returning to treatment after having abandoned it; unknown treatment completion status; a positive sputum culture; having a concomitant form of the disease; and not being assigned to DOTS.

Factors for those more likely to develop MDR TB were being 20-59 years of age; living in a rural area; having alcoholism, diabetes, or HIV/AIDS; being a relapsed case; having returned after abandoning previous treatment; having transferred away from the treatment center; a positive sputum culture; having pulmonary TB; and not being assigned to DOTS.

Limitations

This study's importance and relevance come from its thorough assessment of TB treatment outcomes using a nationally representative TB case database and its detailed, conceptual, model-based analysis that identified TB determinants in Brazil. There are several important limitations, however, that should be considered. Of the total TB cases, 26% of notified individuals were excluded for insufficient TB treatment outcome data. That said, it is also unlikely that this affected the reported associations since missing data were not differentially distributed among the variables. There were several other variables that presented missing data as well, and those cases were also excluded due to the study's already high statistical power.

The proposed theoretical model has described the multiple TB determinants, however the absence of several important variables in the SINAN database precludes evaluation of several important aspects of this causal framework. Future studies that combine data on the characteristics of the bacillus (such as genotyping) and information from social programs, governmental or nongovernmental, with existing income data on TB patients could allow for the model to be used to identify other important determinants and/or modify those already identified.

Conclusions

The main takeaway from these findings is the need for analysis plans that are supported by theoretical models. Theoretical models are dynamic structures that must be re-evaluated as new information is obtained; they are subject to continual adaptation according to new findings, as well as more complete and discrete information describing outcomes or exposures. Furthermore, the models may vary according to an emphasis on certain desired aspects, being general or more specific. As such, this article is not a definitive model for TB determination or analysis, but rather a proposal that is currently adequate in Brazil, yet can also be extrapolated or adapted to other contexts.

Acknowledgements. This study was supported by Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (Brasilia, Brazil) and an International Clinical, Operational and Health Services Research Training Award (ICOHRTA grant: 5 U2R TW006883-02) from the National Institutes of Health (Bethesda, Maryland, United States).

Conflicts of interest. None.

Disclaimer. Authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the RPSP/ PAJPH and/or PAHO.

REFERENCES

(1.) Brookmeyer R. Statistical problems in epidemiologic studies of the natural history of disease. Environ Health Perspect. 1990;87:43-9.

(2.) Maciel EL. A promocao da saude e os determinantes socais da tuberculose: elementos para a acao. Promocao da saude na diversidade humana e na pluralidade de itinerarios terapeuticos. Campinas: Saberes; 2012. Pp. 429-8.

(3.) Hosking J, Ameratunga S, Morton S, Blank D. A life course approach to injury prevention: a "lens and telescope" conceptual model. BMC Public Health. 2011; 11:695.

(4.) Cantalice Filho JP. Food baskets given to tuberculosis patients at a primary health care clinic in the city of Duque de Caxias, Brazil: effect on treatment outcomes. J Bras Pneumol. 2009;35(10):992-7.

(5.) Belo MT, Luiz RR, Hanson C, Selig L, Teixeira EG, Chalfoun T, et al. Tuberculosis and gender in a priority city in the state of Rio de Janeiro, Brazil. J Bras Pneumol. 2010;36(5):621-5.

(6.) Gegia M, Cohen T, Kalandadze I, Vashakidze L, Furin J. Outcomes among tuberculosis patients with isoniazid resistance in Georgia, 2007-2009. Int J Tuberc Lung Dis. 2012;16(6):812-6.

(7.) Monroe AA, Gonzales RI, Palha PF, Sassaki CM, Ruffino Netto A, Vendramini SH, et al. Envolvimento de equipes da Atencao Basica a Saude no Controle da Tuberculose. Rev Esc Enferm USP. 2008; 42(2):262-7.

(8.) Mazzei AMA, Monroe AA, Sassaki CM, Gonzales RIC, Villa TCS. Suporte social para portador de tuberculose no servico de saude e na comunidade. Boletim de Pneumologia Sanitaria. 2003;11(2):41-6.

(9.) Ayres J, Franca Junior I, Calazans G, Filho H. O conceito de vulnerabilidade e as praticas de Saude: novas perspectivas e desafios. In: Czeresnia D, Freitas CM. Promocao da saude: conceitos, reflexoes e tendencias. Rio de Janeiro: Fiocruz; 2003. Pp. 117-39.

(10.) de Oliveira GP, Torrens AW, Bartholomay P, Barreira D. Tuberculosis in Brazil: last ten years analysis, 2001-2010. Braz J Infect Dis. 2013;17(2):218-33.

(11.) Ministry of Health of Brazil. Sistema de Informacao de Agravos de Notificacao Brasilia. Available from: http://dtr2004. saude.gov.br/sinanweb/ Accessed on 20 July 2012.

(12.) Laguardia J, Domingues CMA, Carvalho C, Lauerman CR, Macario E, Glatt R. Sistema de Informacao de Agravos de Notificacao (SINAN): desafios no desenvolvimento de um sistema de informacao em saude. Epidemiol Serv Saude. 2004; 13(3):135-47.

(13.) Medeiros D, Sucupira ED, Guedes RM, Costa AJL. Analise da qualidade das informacoes sobre tuberculose no municipio de Belford Roxo, Rio de Janeiro, 2006 a 2008. Cad Saude Colet. 2012;20(2): 146-52.

(14.) Ministerio da Saude da Brasil. Secretaria de Vigilancia em Saude. Departamento de Vigilancia Epidemiologica. Manual de recomendacoes para o controle da tuberculose no Brasil. Brasilia: Ministerio da Saude; 2011. Pp. 284.

(15.) Kleinbaum DG, Klein M. Logistic regression: a self-learning text. Statistics for biology and health. New York: Springer; 2010. Pp. 429-62.

(16.) Biesheuvel CJ, Vergouwe Y, Steyerberg EW, Grobbee De, Moons KG. Polytomous logistic regression analysis could be applied more often in diagnostic research. J Clin Epidemiol. 2008;61(2):125-34.

(17.) Ananth CV, Kleinbaum DG. Regression models for ordinal responses: a review of methods and applications. Int J Epidemiol. 1997;26(6):1323-33.

(18.) Mafra AC, Nucci LB, Cordeiro R, Stephan C. Estudo caso-controle com resposta multinomial: uma proposta de analise. Cad Saude Publica. 2010;26(3):451-9.

(19.) Facchini LA. Teoria. Trabalho materno e ganho de peso infantil. Pelotas: UFPel; 1995. Pp. 23-40.

(20.) Hernandez-Giron C, Orozco-Nunez E, Arredondo-Lopez A. Modelos conceptuales y paradigmas en salud publica. Rev Salud Publica (Bogota). 2012;14(2): 315-24.

(21.) Bastos JL, Gigante DP, Peres KG, Nedel FB. Determinacao social da odontalgia em estudos epidemiologicos: revisao teorica e proposta de um modelo conceitual. Cien Saude Colet. 2007;12(6):1611-21.

(22.) Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. Int J Epidemiol. 1997;26(1): 224-7.

(23.) Weekes AA, Torgersen CE, Montgomery DR, Woodward A, Bolton SM. A process-based hierarchical framework for monitoring glaciated alpine headwaters. Environ Manage. 2012;50(6):982-97.

(24.) Burnet FM. The natural history of tuberculosis. Med J Aust. 1948;1(3):57-63.

(25.) Myers JA. The natural history of tuberculosis in the human body; forty-five years of observation. JAMA. 1965;194(10): 1086-92.

(26.) Hargreaves JR, Boccia D, Evans CA, Adato M, Petticrew M, Porter JD. The social determinants of tuberculosis: from evidence to action. Am J Public Health. 2011;101(4): 654-62.

(27.) Picon PD, Rizzon CF, Bassanesi SL, Silva LC, Della Giustina Mde L. Retreatment of tuberculosis patients in the city of Porto Alegre, Brazil: outcomes. J Bras Pneumol. 2011;37(4):504-11.

(28.) Belo MT, Luiz RR, Teixeira EG, Hanson C, Trajman A. Tuberculosis treatment outcomes and socio-economic status: a prospective study in Duque de Caxias, Brazil. Int J Tuberc Lung Dis. 2011;15(7): 978-81.

(29.) Antoine D, Che D. Treatment outcome monitoring of pulmonary tuberculosis cases notified in France in 2009. Euro Surveill. 2013;18(12):20434.

(30.) Escobar AL, Coimbra CE Jr, Camacho LA, Portela MC. Tuberculose em populacoes indigenas de Rondonia, Amazonia, Brasil. Cad Saude Publica. 2001;17(2):285-98.

(31.) Sanchez M, Bartholomay P, Arakaki-Sanchez D, Enarson D, Bissell K, Barreira D, et al. Outcomes of TB treatment by HIV status in national recording systems in Brazil, 2003-2008. PLoS One. 2012;7(3): e33129.

Manuscript received on 18 July 2014. Revised version accepted for publication on 14 April 2015.

Ethel Leonor Maciel [1] and Barbara Reis-Santos [1]

[1] Epidemiology Laboratory of the Universidade Federal do Espirito Santo, Vitoria, Espirito Santo, Brazil. Send correspondence to Ethel Leonor Maciel, ethel.maciel@gmail.com

TABLE 1. Distribution of variables in the
Information System for Notifiable Diseases
database (SINAN) in a hierarchical
model of tuberculosis (TB) determinants,
Brazil, 2007-2011

Hierarchy level    SINAN variable

Level 1            Age group
                   Skin color
                   Gender
                   Schooling

Level 2            Area of residence
                   Institutionalization

Level 3            Alcoholism
                   Diabetes mellitus
                   Mental disease
                   HIV/AIDS

Level 4            Type of SINAN notification
                   Initial smears results
                   Initial culture results
                   Clinical form of TB

Level 5            If under Directly Observed
                   Treatment Short-course

TABLE 2. Logistic and polytomous hierarchical models of
association between sociodemographic and clinical
characteristics of individuals with tuberculosis (TB)
and tuberculosis treatment outcome, Brazil, 2007-2011

                                                        Cured vs
                              Cured vs Failure         Abandonment

    Characteristic (n)       OR (a) (95% CI (b))       OR (95% CI)

Level 1
Age group (318 465)
  < 20 years                      Ref. (d)                Ref.
  20-39 years                 0.67 (0.65-0.69)      1.57 (1.50-1.64)
  40-59 years                 0.72 (0.70-0.74)      0.95 (0.91-1.00)
  [less than or equal to]     0.59 (0.57-0.61)      0.57 (0.54-0.61)
    60 years
Gender (318 465)
  Female                            Ref.                  Ref.
  Male                        0.76 (0.75-0.77)      1.48 (1.44-1.51)
Skin color (275 109)
  Caucasian                         Ref.                  Ref.
  Black                       0.79 (0.77-0.81)      1.49 (1.44-1.54)
  Brown                       0.86 (0.84-0.87)      1.17 (1.14-1.20)
  Asian/ indigenous           1.05 (0.99-1.11)      0.91 (0.83-0.99)
Schooling (196 869)
  Illiterate                        Ref.                  Ref.
  1-4 years                   1.12 (1.08-1.16)      0.91 (0.86-0.96)
  5-8 years                   1.08 (1.04-1.12)      0.95 (0.89-1.00)
  > 8 years                   1.88 (1.81-1.96)      0.46 (0.43-0.49)
  Not applicable              1.25 (1.17-1.34)      0.55 (0.49-0.61)
Level 2
Area of residence
    (238 067)
  Rural                             Ref.                  Ref.
  Urban                       1.36 (1.32-1.40)      0.55 (0.53-0.58)
  Peri-urban                  1.12 (1.01-1.23)      0.63 (0.54-0.75)
Institutionalization
    (266 136)
  None                              Ref.                  Ref.
  Prison                      1.23 (1.19-1.27)      0.68 (0.64-0.71)
  Other                       0.78 (0.75-0.81)      1.28 (1.21-1.36)
Level 3
Alcoholism (318 465)
  No                                Ref.                  Ref.
  Yes                         0.55 (0.54-0.56)      2.00 (1.93-2.05)
Diabetes mellitus
    (318 465)
  No                                Ref.                  Ref.
  Yes                         1.06 (1.03-1.10)      0.68 (0.64-0.73)
Mental illness (318 465)
  No                                Ref.                  Ref.
  Yes                         0.86 (0.82-0.91)      1.00 (0.92-1.08)
HIV/AIDS (318 465)
  No                                Ref.                  Ref.
  Yes                         0.31 (0.31-0.32)      2.07 (2.00-2.14)
Level 4
Type of SINAN (e)
    notification
    (318 465)
  New case                          Ref.                  Ref.
  Relapse                     0.78 (0.76-0.81)      1.48 (1.41-1.55)
  Return after                0.35 (0.34-0.36)      5.64 (5.44-5.85)
    abandonment
  Unknown                     0.33 (0.29-0.37)      2.15 (1.65-2.80)
  Transfer                    0.74 (0.72-0.76)      1.05 (0.99-1.11)
Smear test (258 648)
  Negative                          Ref.                  Ref.
  Positive                    1.04 (1.02-1.06)      1.07 (1.03-1.10)
Culture test (61 130)
  Negative                          Ref.                  Ref.
  Positive                    0.74 (0.71-0.77)      1.51 (1.42-1.61)
Clinical form of TB
    (318 465)
  Pulmonary                         Ref.                  Ref.
  Extrapulmonary              1.15 (1.12-1.17)      0.69 (0.67-0.72)
Pulmonary +                   0.71 (0.68-0.74)      0.97 (0.91-1.04)
  Extrapulmonary
Level 5 DOTS (f)
    (318 465)
  No                                Ref.                  Ref.
  Yes                         1.58 (1.55-1.60)      0.52 (0.51-0.54)

                               Cured vs Death        Cured vs Death
                                   from TB          from other cause

    Characteristic (n)           OR (95% CI)           OR (95% CI)

Level 1
Age group (318 465)
  < 20 years                        Ref.                  Ref.
  20-39 years                 1.92 (1.72-2.14)      2.37 (2.14-2.63)
  40-59 years                 3.93 (3.53-4.38)      4.16 (3.76-4.62)
  [less than or equal to]     7.97 (7.15-8.89)      8.04 (7.24-8.93)
    60 years
Gender (318 465)
  Female                            Ref.                  Ref.
  Male                        1.39 (1.34-1.45)      1.25 (1.20-1.29)
Skin color (275 109)
  Caucasian                         Ref.                  Ref.
  Black                       1.24 (1.17-1.32)      0.94 (0.89-1.00)
  Brown                       1.17 (1.12-1.23)      0.79 (0.76-0.82)
  Asian/ indigenous           0.99 (0.87-1.13)      0.65 (0.57-0.75)
Schooling (196 869)
  Illiterate                        Ref.                  Ref.
  1-4 years                   0.83 (0.77-0.89)      0.87 (0.81-0.94)
  5-8 years                   0.79 (0.73-0.85)      0.88 (0.81-0.96)
  > 8 years                   0.50 (0.45-0.55)      0.56 (0.51-0.61)
  Not applicable              1.23 (1.06-1.42)      1.38 (1.20-1.58)
Level 2
Area of residence
    (238 067)
  Rural                             Ref.                  Ref.
  Urban                       0.67 (0.62-0.72)      0.90 (0.84-0.96)
  Peri-urban                  0.77 (0.60-0.98)      0.78 (0.61-0.99)
Institutionalization
    (266 136)
  None                              Ref.                  Ref.
  Prison                      0.56 (0.50-0.63)      0.70 (0.64-0.76)
  Other                       1.46 (1.34-1.59)      1.35 (1.23-1.47)
Level 3
Alcoholism (318 465)
  No                                Ref.                  Ref.
  Yes                         2.07 (1.98-2.16)      1.39 (1.32-1.45)
Diabetes mellitus
    (318 465)
  No                                Ref.                  Ref.
  Yes                         1.01 (0.94-1.08)      1.10 (1.03-1.17)
Mental illness (318 465)
  No                                Ref.                  Ref.
  Yes                         1.41 (1.28-1.56)      1.44 (1.31-1.60)
HIV/AIDS (318 465)
  No                                Ref.                  Ref.
  Yes                         4.07 (3.88-4.27)     11.10 (10.65-11.54)
Level 4
Type of SINAN (e)
    notification
    (318 465)
  New case                          Ref.                  Ref.
  Relapse                     1.32 (1.23-1.41)      1.00 (0.93-1.07)
  Return after                2.10 (1.95-2.26)      1.33 (1.22-1.44)
    abandonment
  Unknown                    11.57 (9.54-14.03)     5.02 (3.97-6.36)
  Transfer                    1.26 (1.16-1.36)      0.81 (0.75-0.89)
Smear test (258 648)
  Negative                          Ref.                  Ref.
  Positive                    0.88 (0.84-0.93)      0.64 (0.61-0.66)
Culture test (61 130)
  Negative                          Ref.                  Ref.
  Positive                    1.53 (1.37-1.71)      1.35 (1.24-1.48)
Clinical form of TB
    (318 465)
  Pulmonary                         Ref.                  Ref.
  Extrapulmonary              0.69 (0.65-0.74)      0.95 (0.90-0.99)
Pulmonary +                   2.03 (1.88-2.20)      1.96 (1.83-2.10)
  Extrapulmonary
Level 5 DOTS (f)
    (318 465)
  No                                Ref.                  Ref.
  Yes                         0.50 (0.48-0.53)      0.68 (0.65-0.71)

                             Cured vs MDR TB (c)

    Characteristic (n)           OR (95% CI)

Level 1
Age group (318 465)
  < 20 years                        Ref.
  20-39 years                 2.09 (1.58-2.78)
  40-59 years                 2.30 (1.73-3.07)
  [less than or equal to]     1.50 (1.07-2.09)
    60 years
Gender (318 465)
  Female                            Ref.
  Male                        1.01 (0.89-1.14)
Skin color (275 109)
  Caucasian                         Ref.
  Black                       1.01 (0.85-1.20)
  Brown                       0.91 (0.80-1.03)
  Asian/ indigenous           0.38 (0.21-0.70)
Schooling (196 869)
  Illiterate                        Ref.
  1-4 years                   1.19 (0.90-1.58)
  5-8 years                   1.59 (1.20-2.12)
  > 8 years                   0.82 (0.60-1.11)
  Not applicable              0.27 (0.11-0.70)
Level 2
Area of residence
    (238 067)
  Rural                             Ref.
  Urban                       0.66 (0.52-0.83)
  Peri-urban                  0.90 (0.46-1.74)
Institutionalization
    (266 136)
  None                              Ref.
  Prison                      0.82 (0.63-1.08)
  Other                       0.86 (0.62-1.20)
Level 3
Alcoholism (318 465)
  No                                Ref.
  Yes                         1.79 (1.54-2.07)
Diabetes mellitus
    (318 465)
  No                                Ref.
  Yes                         1.53 (1.24-1.90)
Mental illness (318 465)
  No                                Ref.
  Yes                         0.66 (0.42-1.05)
HIV/AIDS (318 465)
  No                                Ref.
  Yes                         1.53 (1.28-1.84)
Level 4
Type of SINAN (e)
    notification
    (318 465)
  New case                          Ref.
  Relapse                     9.43 (8.13-10.95)
  Return after                9.14 (7.67-10.88)
    abandonment
  Unknown                        2.02 (---)
  Transfer                    6.44 (5.43-7.63)
Smear test (258 648)
  Negative                          Ref.
  Positive                    2.41 (2.00-2.91)
Culture test (61 130)
  Negative                          Ref.
  Positive                    6.46 (4.59-9.08)
Clinical form of TB
    (318 465)
  Pulmonary                         Ref.
  Extrapulmonary              0.32 (0.21-0.47)
Pulmonary +                   0.73 (0.49-1.09)
  Extrapulmonary
Level 5 DOTS (f)
    (318 465)
  No                                Ref.
  Yes                         0.84 (0.75-0.95)

(a) Odds Ratio.

(b) 95% Confidence Interval.

(c) Multidrug-resistant tuberculosis.

(d) Reference.

(e) Notifiable Disease Surveillance System database

(f) Directly Observed Treatment Short-course.
COPYRIGHT 2015 Pan American Health Organization
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Original research/Investigacion original
Author:Maciel, Ethel Leonor; Reis-Santos, Barbara
Publication:Revista Panamericana de Salud Publica
Date:Jul 1, 2015
Words:4835
Previous Article:Equity-oriented monitoring in the context of universal health coverage/El monitoreo orientado a la equidad en el contexto de la cobertura universal...
Next Article:Validity of the Mexican version of the combined Foot Care Confidence/Foot-Care Behavior scale for diabetes/Validez de la version mexicana del...
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters