- Top of page
- Material and methods
Objectives To analyse the profile of tuberculosis (TB) among Bolivian immigrants, investigate the impact that this population has on the trends of TB and assess equity in access to TB treatment, in the city of São Paulo, Brazil.
Methods Descriptive study of the epidemiological profile of TB in four city districts with large Bolivian populations, comparing cases among Brazilians and Bolivians, during the 1998–2008 period was carried out. We used logistic regression to adjust the treatment outcome for potential confounders.
Results We identified 2056 new TB cases: 65.7% in Brazilians, 32.1% in Bolivians and 2.2% among other nationalities. Although TB incidence remained stable (high) over the study period, the annual proportion of cases among Bolivians increased from 15.0% to 53.0%. In comparison with the Brazilians, the Bolivians were younger (median age, 24 vs. 40 years; P < 0.0001) and presented a lower unemployment rate (3.1%vs. 11.6%; P < 0.0001), a lower rate of HIV co-infection (1.5%vs. 28.5%; P < 0.001), a higher proportion of individuals receiving supervised treatment (81.5%vs. 62.0%; P < 0.0001) and a higher proportion of cures (71.6%vs. 63.2%; P < 0.0001). After having been adjusted for potential confounder, cure after treatment was not associated with nationality.
Conclusions Bolivian immigrants influenced the incidence but not the trends of TB among Brazilians in the study area. We found no significant differences between Bolivians and Brazilians regarding healthcare access or treatment outcome. Guaranteed universal health care access for all, including undocumented individuals, contributes to health equity. Specific intervention strategies are warranted for immigrants with tuberculosis.
Objectifs: Analyser le profil de la tuberculose (TB) chez les immigrants boliviens, investiguer l’impact que cette population a sur les tendances de la TB et évaluer l’équité dans l’accès au traitement contre la TB dans la ville de São Paulo au Brésil.
Méthodes: Etude descriptive du profil épidémiologique de la TB dans 4 districts de la ville avec de vastes populations boliviennes, comparant les cas chez les Brésiliens et les Boliviens, au cours de la période 1998-2008. Nous avons utilisé la régression logistique pour ajuster le résultat du traitement pour les facteurs confusionnels potentiels.
Résultats: Nous avons identifié 2.056 nouveaux cas de TB: 65,7% chez des brésiliens, 32,1% chez des boliviens et 2,2% chez d’autres nationalités. Bien que l’incidence de la TB soit restée stable (élevée) au cours de la période d’étude, la proportion annuelle des cas chez les boliviens a augmenté de 15,0%à 53,0%. En comparaison avec les brésiliens, les boliviens étaient plus jeunes (âge médian, 24 vs 40 ans; P <0,0001); présentaient un taux de chômage inférieur (3,1% vs 11,6%; P <0,0001), un taux de coinfection avec VIH faible (1,5% vs 28,5%; P <0,001), une proportion plus élevée de personnes recevant un traitement supervisé (81,5% vs 62,0%; P <0,0001) et une proportion plus élevée de guérisons (71,6% vs 63,2%; P <0,0001). Après avoir été ajusté pour les facteurs confusionnels potentiels, la guérison après traitement n’était pas liée à la nationalité.
Conclusions: Les immigrés boliviens influencent l’incidence mais pas les tendances de la TB chez les brésiliens dans la zone d’étude. Nous n’avons pas trouvé de différences significatives entre les boliviens et les brésiliens concernant l’accès aux soins de santé ou les résultats du traitement. L’accès universel aux soins de santé garanti pour tous, y compris aux personnes sans-papiers, contribue à l’équité en santé. Des stratégies d’intervention spécifiques sont justifiées pour les immigrants atteints de TB.
Objetivos: Analizar el perfil de la tuberculosis (TB) entre los inmigrantes Bolivianos, investigar el impacto que esta población tiene sobre las tendencias de la TB, y evaluar la equidad en el acceso al tratamiento de la TB en la ciudad de São Paulo, Brasil.
Métodos: Estudio descriptivo del perfil epidemiológico de la TB en 4 distritos de la ciudad con una amplia población boliviana, comparando casos entre Brasileros y Bolivianos, durante el periodo comprendido entre 1998–2008. Hemos utilizado una regresión logística ajustando al resultado del tratamiento a posibles confusores.
Resultados: Identificamos 2,056 nuevos casos de TB: 65.7% en Brasileros; 32.1% en Bolivianos; y 2.2% entre otras nacionalidades. Aunque la incidencia de TB se mantuvo estable (alta) a lo largo del periodo de estudio, la proporción anual de casos entre los Bolivianos aumentó del 15.0% al 53.0%. En comparación con los Brasileros, los Bolivianos eran más jóvenes (edad media, 24 vs. 40 años; P<0.0001); presentaban una menor tasa de desempleo (3.1% vs. 11.6%; P <0.0001), una menor tasa de coinfección con VIH (1.5% vs. 28.5%; P<0.001), una mayor proporción de individuos recibiendo tratamiento supervisado (81.5% vs. 62.0%; P<0.0001) y una mayor proporción de curaciones (71.6% vs. 63.2%; P<0.0001). Después de haber ajustado para posibles confusores, la curación después de haber recibido tratamiento no estaba asociada a la nacionalidad.
Conclusiones: Los inmigrantes Bolivianos tienen influencia sobre la incidencia pero no sobre las tendencias de la TB entre los Brasileros del área de estudio. No encontramos diferencias significativas entre los Bolivianos y los Brasileros en lo que respecta al acceso al tratamiento o a los resultados del mismo. Un acceso universal garantizado a los servicios sanitarios, incluidos los individuos indocumentados, contribuye a la equidad en salud. Las estrategias de intervención específicas están justificadas para inmigrantes con tuberculosis.
- Top of page
- Material and methods
In recent decades, population movements have been a recurrent cause for concern among international organisations, because such movements have economic and social consequences, as well as having an impact on public health (Aagaard-Hansen et al. 2010). It has been estimated that 190 million people emigrate annually; of those, 40% move to neighbouring countries (Carballo & Nerukar 2001; Vázquez et al. 2011). This population is characterised as vulnerable from a socio-economic and health standpoint, because of poor living/working conditions, illegal immigrant status, language barriers and cultural barriers (Stronks et al. 2001). Therefore, migration has become a focus of socio-economic development plans and specific public policies to integrate immigrants into the host society (Vázquez et al. 2011).
Migration also constitutes a major public health challenge because it creates an unforeseen demand for health care and because it changes the epidemiological profile of the host country (Carballo & Nerukar 2001). In recent decades, population movements from regions where tuberculosis (TB) is hyperendemic have been associated with resurgence of the disease in various countries (Gilbert et al. 2009).
Some Brazilian state capitals, especially the city of São Paulo, have received, since the 1990s, an increasing number of immigrants from countries in South America, Africa and Asia (Patarra 2005). However, no studies have consistently estimated the size and characteristics of immigrant communities. Most immigrants are undocumented and are therefore excluded from official statistics. Although imprecise, the available data indicate that the Bolivian community is the largest of all of the communities of immigrants who have recently arrived in the city of São Paulo (Patarra & Baeninger 2006).
The high TB incidence in Bolivia (WHO 2009) and the lack of published data regarding the impact that the Bolivian immigration has on the epidemiological profile of the disease in the city of São Paulo warrant an exploratory study of the issue. The objectives of this study were to describe the profile of TB among Bolivian immigrants residing in four districts in the city of São Paulo, Brazil, and to determine the impact of the Bolivian community on TB trends in the study area, as well as to investigate possible disparities between Brazilians and Bolivian immigrants regarding access to health care and the quality of the treatment provided.
- Top of page
- Material and methods
Between 1998 and 2008, 2056 new TB cases were reported in the study area. Of those, 1351 (65.7%) occurred among Brazilians, 660 (32.1%) occurred among Bolivians, and 45 (2.2%) occurred among individuals of other nationalities. From 1998 to 2008, the annual proportion of TB cases among Bolivians increased from 15.0% to 53.0%. In addition, while the number of new TB cases among Brazilians decreased by 40.5%, the number of new TB cases among Bolivians increased by 260.0% (Figure 2). The mean annual TB incidence rates in the study area ranged from 149.1/100 000 population in 1998 to 167.2/100 000 population in 2008 (Figure 2).
Figure 2. Annual incidence of tuberculosis, for all forms and cases of the disease, by nationality, in the districts of Belém, Bom Retiro, Brás and Pari, in the city of São Paulo, Brazil (1998–2008).
Download figure to PowerPoint
Between 2006 and 2008, 557 new TB cases were reported in the study area. Of those, 272 (48.8%) occurred among Brazilians, 273 (49.0%) occurred among Bolivians, and 12 (2.2%) occurred among individuals of other nationalities.
Table 1 shows that the median age was higher among Brazilians than among Bolivians (40 vs. 24 years; P < 0.0001). The proportion of men was also higher among Brazilians (73.5%vs. 61.9%; P < 0.0001), as was the proportion of unemployed individuals (11.6%vs. 3.1%; P < 0.0001). There are no differences among Brazilians and Bolivians in the level of education; the district of Belem had the higher proportion of TB among Bolivians. The TB patients with HIV co-infection were higher among Brazilians than among Bolivians (28.5%vs. 1.5%; P < 0.001), as were the prevalence of alcoholism (15.4%vs. 2.6%; P < 0.0001), the prevalence of diabetes (4.4%vs. 1.5%; P = 0.04) and the prevalence of mental illness (1.5%vs. 0.0%; P = 0.04).
Table 1. New cases of tuberculosis among Brazilians and Bolivians, by socio-demographical characteristic and comorbidities, in the districts of Belém, Bom Retiro, Brás and Pari, in the city of São Paulo, Brazil (2006–2008)
|Characteristic||Nationality||Total (n = 545†) n (%)|| P*|
|Brazilian (n = 272†) n (%)||Bolivian (n = 273†) n (%)|
| Female||72 (26.5)||104 (38.1)||176 (32.2)||0.004|
| Male||200 (73.5)||169 (61.9)||369 (67.7)|
| <15 years||7 (2.6)||15 (5.5)||22 (4.1)||<0.0001|
| 15–29 years||60 (22.1)||204 (75.0)||264 (48.6)|
| 30–39 years||62 (22.9)||41 (15.1)||103 (19.0)|
| 40–49 years||63 (23.2)||4 (1.5)||67 (12.3)|
| ≥50 years||79 (29.2)||4 (2.9)||87 (16.0)|
|Median age (years)||40.0||24.0|| ||<0.0001|
|Level of education|
| 0 years of schooling||4 (3.3)||5 (3.7)||9 (3.5)||0.412|
| 1–4 years of schooling||19 (15.4)||22 (16.4)||41 (16.0)|
| 4–8 years of schooling||49 (39.8)||64 (47.8)||113 (44.0)|
| 9–12 years of schooling||41 (33.3)||30 (22.4)||71 (27.6)|
| >12 years of schooling||10 (8.1)||13 (9.7)||23 (8.9)|
| Employed||189 (78.1)||242 (94.5)||431 (86.5)||<0.0001|
| Unemployed||28 (11.6)||8 (3.1)||36 (7.2)|
| Homemaker||15 (6.2)||5 (2.0)||20 (4.0)|
| Retired||10 (4.1)||1 (0.4)||11 (2.2)|
|District of residence|
| Pari||36 (13.2)||26 (9.5)||62 (11.4)||0.01|
| Bom Retiro||69 (25.4)||68 (24.9)||137 (25.1)|
| Brás||87 (32.0)||64 (23.4)||151 (27.7)|
| Belém||80 (29.4)||115 (42.5)||195 (35.5)|
| No||143 (71.5)||198 (98.5)||341 (85.0)||<0.001|
| Yes||57 (28.5)||3 (1.5)||60 (15.0)|
| No||260 (95.6)||269 (98.5)||529 (97.1)||0.04|
| Yes||12 (4.4)||4 (1.5)||16 (2.9)|
| No||230 (84.6)||266 (97.4)||496 (91.0)||<0.0001|
| Yes||42 (15.4)||7 (2.6)||49 (9.0)|
| No||268 (98.5)||273 (100.0)||541 (99.3)||0.04|
| Yes||4 (1.5)||0 (0)||4 (0.7)|
The predominant clinical form was pulmonary TB, which was seen in 80–86% of the cases, and there was no significant difference between Brazilians and Bolivians in terms of this aspect (P > 0.05). As can be seen in Table 2, there were no significant differences between Brazilians and Bolivians regarding the time elapsed between symptom onset and treatment initiation (P > 0.05); however, the proportion of patients who underwent supervised treatment was higher among Bolivians (81.5%vs. 62.0%; P < 0.0001). The proportions of patients diagnosed at primary healthcare facilities were 63.3% and 42.1% among the Bolivians and the Brazilians, respectively, whereas the proportions of patients treated at such facilities were 72.1% and 43.5%, respectively (P < 0.0001 for both). Comparing Bolivians and Brazilians, we found that the former showed higher proportion of cure rates (71.6%vs. 63.2%; P < 0.0001). In addition, the rates of all unfavourable outcomes were lower among the Bolivians than among the Brazilians: treatment abandonment (16.0%vs. 20.7%; P < 0.0001), deaths from TB (1.2%vs. 5.0%; P < 0.001) and deaths from other causes (1.6%vs. 10.7%; P < 0.001). There were no documented cases of treatment failure.
Table 2. New cases of tuberculosis among Brazilians and Bolivians, by characteristic related to diagnosis/treatment, in the districts of Belém, Bom Retiro, Brás and Pari, in the city of São Paulo, Brazil (2006–2008)
|Characteristic||Nationality||Total (n = 545†) n (%)|| P*|
|Brazilian (n = 272†) n (%)||Bolivian (n = 273†) n (%)|
|Time from symptom onset to treatment initiation|
| ≤4 weeks||72 (54.5)||100 (62.9)||172 (59.1)||>0.05|
| 5–8 weeks||32 (24.2)||23 (14.5)||55 (18.9)|
| 9–12 weeks||15 (11.4)||21 (13.2)||36 (12.4)|
| >12 weeks||13 (9.8)||15 (9.4)||28 (9.6)|
|Type of treatment|
| Supervised||103 (62.0)||185 (81.5)||288 (73.3)||<0.0001|
| Self-administered||63 (38.0)||42 (18.5)||105 (26.7)|
| Cure||153 (63.2)||174 (71.6)||327 (67.4)||<0.0001|
| Abandonment||50 (20.7)||39 (16.0)||89 (18.4)|
| Transfer||1 (0.4)||23 (9.5)||24 (4.9)|
| Death from tuberculosis||12 (5.0)||3 (1.2)||15 (3.1)|
| Death from HIV||26 (10.7)||4 (1.6)||30 (6.2)|
|Place of diagnosis|
| Primary healthcare clinic||110 (42.1)||162 (63.3)||272 (52.6)||<0.0001|
| Emergency room||79 (30.3)||54 (21.1)||133 (25.7)|
| Hospital||57 (21.8)||30 (11.7)||87 (16.8)|
| Other||15 (5.8)||10 (3.9)||25 (4.8)|
|Place of treatment|
| Primary healthcare clinic||118 (43.5)||196 (72.1)||314 (72.1)||<0.0001|
| Outpatient clinic||96 (35.4)||58 (21.3)||154 (28.4)|
| Hospital||57 (21.0)||18 (6.6)||75 (13.8)|
|Mean number of contacts per TB case||1.8||4.4|| ||<0.0001|
|Household contacts||437 (27.0)||1182 (73.0)||1619 (100%)|| |
|Household contacts examined|
| Yes||282 (64.5)||719 (60.8)||1001 (61.8)||>0.05|
| No||155 (35.5)||463 (39.2)||618 (38.2)|
As shown in Table 2, the mean number of contacts per patient was higher among Bolivians than among Brazilians (4.4 vs. 1.8; P < 0.0001). The total number of contacts was 1182 among the Bolivian patients – 719 (60.8%) of those 1182 had been examined – compared with 437 among the Brazilian patients – 282 (64.5%) of those 437 had been examined. The difference between Bolivians and Brazilians regarding the proportion of contacts examined was not significant (P > 0.05). Among the contacts examined, new cases of TB were identified in 2 (0.7%) of the 282 Brazilians and 13 (1.8%) of the 719 Bolivians.
Bivariate analysis (Table 3), using Brazilians as the category of reference, revealed a positive association between cure and being Bolivian (crude OR = 2.2; 95% CI: 1.4–3.3). However, that association did not remain statistically significant when we adjusted simultaneously for gender, age, type of treatment and HIV co-infection (adjusted OR = 1.0; 95% CI: 0.4–2.5).
Table 3. Association between cure after tuberculosis treatment and nationality after adjusted by sex, age, co-infections TB/HIV and type of treatment. Districts of Belém, Bom Retiro, Brás and Pari, in the city of São Paulo, Brazil (2006–2008)
|Characteristic||Cure after treatment||OR crude (95% CI)||OR adjusted (95% CI)|
|No (n = 134*) n (%)||Yes (n = 327*) n (%)|
| Female||36 (26.9)||117 (35.8)||1||1|
| Male||98 (73.1)||210 (64.2)||0.7 (0.4–1.0)||0.4 (0.2–1.1)|
| <15 years||2 (1.5)||15 (4.6)||1||1|
| 15–29 years||54 (40.3)||161 (49.5)||0.4 (0.1–1.7)||1.7 (0.2–16.1)|
| 30–39 years||20 (14.9)||68 (20.9)||0.4 (0.1–1.9)||1.2 (0.1–12.5)|
| 40–49 years||24 (17.9)||34 (10.5)||0.2 (0.1–0.8)||0.8 (0.1–9.6)|
| ≥50 years||34 (25.4)||47 (14.5)||0.2 (0.1–0.8)||1.2 (0.1–13.4)|
| Yes||30 (38.0)||23 (8.8)||1||1|
| No||49 (62.0)||238 (91.2)||6.3 (3.4–11.8)||5.3 (2.0–14.2)|
|Type of treatment|
| Self-administered||28 (41.8)||64 (23.7)||1||1|
| Supervised||39 (58.2)||206 (76.3)||2.3 (1.3–4.0)||1.5 (0.6–3.2)|
| Brazilian||88 (65.7)||153 (46.8)||1||1|
| Bolivian||46 (34.3)||174 (53.2)||2.2 (1.4–3.3)||1.0 (0.4–2.5)|
- Top of page
- Material and methods
The study area is located in the oldest sector of the city of São Paulo. Although there is ample access to public transportation, schools and health care in those districts, living conditions have deteriorated in recent decades, which have led to a worsening of the social, economic and health indicators (Ferreira et al. 2006). It should be highlighted that even when the number of Bolivian immigrants in those districts was low, the TB incidence was higher there than in the city as a whole.
Two of the findings of the present study are relevant to public health. One is that Bolivian immigrants have influenced the incidence of TB, as evidenced by the rapid increase (in absolute and relative values) of TB among Bolivians immigrants, without affecting the trends of TB or its characteristics in the districts under study, as evidenced by the consistent decrease in the number of TB cases among Brazilians. Nevertheless, considering the mean number of contacts per patient and the proportion of cases identified among contacts were far higher among Bolivians than among Brazilians, together with the fact that the number of Bolivian immigrants is increasing, shows that in the city of São Paulo, we perceive a potential risk of the resurgence of TB, whose occurrence in other countries has been attributed to immigration (Gilbert et al. 2009).
Our results show that migration from countries where the TB incidence is high can have a major impact on host cities, even cities such as São Paulo, where the TB incidence is already far higher than is that reported for large cities in developed countries (Iñigo et al. 2007; Gilbert et al. 2009). One aspect that merits attention is the possibility of a change in the pattern of TB transmission as a result of such migration flows, a phenomenon reported in studies conducted in developed countries receiving immigrants from hyperendemic areas (Iñigo et al. 2007). Investigations conducted in those countries showed that immigrants rarely become infected after having arrived in the host country, which is evidence that infection occurs before their arrival, that is, in their country of origin (Dahle et al. 2001). Although the risk of developing a disease is higher during the first years following emigration, it can remain high for longer when the country of origin is highly endemic (Lillebaek et al. 2002; Vos et al. 2004). However, studies conducted in developed countries, using epidemiological data and DNA genotyping, suggest that immigration from hyperendemic areas has no significant influence on the pattern of TB transmission within the host country population. Nevertheless, that could be attributed to limited contact between the two populations and to the good performance of immigrant-screening programmes targeting TB (Dahle et al. 2007; Barniol et al. 2009). Those issues are relevant to the development of new control strategies and require further investigation in the context of a large Brazilian city, where the risk of TB infection is far higher than is that reported for developed countries.
Another noteworthy finding of the present study is the equity in access to health care and in the quality of the treatment provided. This equity was evidenced by the lack of differences between Brazilians and Bolivians regarding the time elapsed between symptom onset and treatment initiation, as well as the proportions of patients who received home visits, patients who were cured and patients whose outcome was unfavourable (treatment abandonment or death). The most striking finding was that the proportion of patients submitted to supervised treatment was higher among the Bolivians than among the Brazilians.
These results suggest that, even in the absence of public health policies targeting immigrants, the guarantee of universal access to free health care – one of the cornerstones of the Brazilian public health system – promotes equity in health (Menicucci 2009; Guanais 2010). It is possible that the changes made by the Brazilian National Tuberculosis Control Programme in the last decade, which strengthened the role of municipal primary health care (Hijjar et al. 2007), have increased the effectiveness of the programme in vulnerable populations.
The fact that Bolivian immigration in São Paulo typically results from labour migration (Patarra & Baeninger 2006; Cymbalista & Xavier 2007) largely explains the differences between Bolivian and Brazilian patients regarding their socio-demographical profile. In our study, we found that the Bolivians were far younger than the Brazilians, the proportion of women being higher and unemployment rates being lower among the Bolivians, although the level of education was comparable between the two populations. In general, similar characteristics were observed among immigrants with TB in Madrid, Spain and most of whom were from South or Central America (Iñigo et al. 2007).
Unlike TB among Bolivians, TB among Brazilians has the characteristics of TB in areas of moderate endemicity. The disease primarily affects individuals at a more advanced age (Chaimowicz 2001), which can partially explain the higher prevalence of comorbidities, especially diabetes, among Brazilians in the present study. In contrast, the low prevalence of HIV co-infection among the Bolivians with TB in the present study was probably due to the incidence of AIDS in Bolivia, which is among the lowest in the Americas (USAID Bolivia 2008).
Our results should be interpreted in view of the fact that they reflect the situation only in the four districts under study and not in the city of São Paulo as a whole. In addition, the quite real possibility that there are undocumented Bolivians in São Paulo who are subject to abusive working conditions should be taken into consideration. The situation in which those immigrants live might prevent them from seeking medical assistance, which would represent a bias in the present study. In 2009, the Brazilian government signed an amnesty bill affecting workers from countries that are members of the Southern Common Market, as well as of those from Chile and Bolivia, allowing illegal immigrants to seek legalisation (Brasil 2010).
The limitations of passive surveillance itself – especially under-reporting – and the lack of data regarding immigrant populations should also be taken into consideration because it prevents us from estimating the risk of TB among Bolivian immigrants.
Despite the aforementioned limitations, the results of the present study are consistent with those in the literature (Gilbert et al. 2009) and are therefore useful for the formulation of public policies, as well as for the planning of interventions. Our results indicate the need for demographical censuses, and national surveys should include specific questions regarding immigration. Furthermore, intersectoral public policies are required to facilitate the integration of immigrants into Brazilian society. The results of the present study underscore the need for evaluating the advantages and disadvantages of screening programmes for immigrants from countries where TB is hyperendemic, to identify and treat individuals with latent TB infection.