Effect of the Bolsa Familia Programme on tuberculosis treatment outcomes – Authors' reply

Poor individuals and sick populations. Already in 1985, Rose1Rose G Sick individuals and sick populations.Int J Epidemiol. 1985; 14: 32-38Crossref PubMed Scopus (2254) Google Scholar declared that the focus of health interventions needed change. Measures with the potential to affect the course of diseases and promote prolonged modifications are those that have population focus. The Bolsa Familia Programme (BFP) is an example of a population approach that can support health promotion and, thus, could effectively modify the course of diseases. However, weak causal forces, the potential for social selection, and complex causal paths are major concerns of upstream social epidemiology determinants such as BFP. In this context, our Article2Oliosi JGN Reis-Santos B Locatelli RL et al.Effect of the Bolsa Familia Programme on the outcome of tuberculosis treatment: a prospective cohort study.Lancet Glob Health. 2018; 7: e219-e226Summary Full Text Full Text PDF PubMed Scopus (26) Google Scholar proposed to evaluate the existence of an independent direct effect of BFP on tuberculosis treatment outcome as a first step in the understanding of this complex causal mechanism, as pointed out by Leonardo de Paula Martins and colleagues. We built a causal diagram on the basis of theoretical models, which allowed us to identify a minimal set of variables, including individual, contextual, and programmatic characteristics, that were needed for our model's adjustment. However, a concomitant initiative to improve socioeconomic status could bias the study results. Outside of the National Tuberculosis Programme, there are few other initiatives of social protection that could help people with tuberculosis in Brazil, and those that do often have little continuity and are ephemeral, a scenario that is reproduced in our study population. The small sample size and low proportion of people who are beneficiaries of other social benefits make further analysis impossible (table). Nevertheless, the statistical strategy that we used (propensity score matching) allowed us to overcome the minor differences between the few people that received other social protection initiatives (table), and, mainly, differences in other unmeasured background determinants, by estimating conditional probabilities.3Hirano K Imbens GW Ridder G Efficient estimation of treatment effects using the estimated propensity score.Econometrica. 2003; 71: 1161-1189Crossref Scopus (1146) Google ScholarTableDistribution of the study population, according to social protection benefits other than BFPBaselineSecond month of treatmentSixth month of treatmentNot a beneficiaryBeneficiaryNot a beneficiaryBeneficiaryNot a beneficiaryBeneficiaryOther social programmes*Programmes of states and municipalities. BFP=Bolsa Familia Programme.1214 (98%)25 (2%)1199 (97%)40 (3%)1219 (98%)20 (2%)Non-governmental benefit1183 (96%)53 (4%)1212 (98%)27 (2%)1225 (99%)14 (1%)* Programmes of states and municipalities. BFP=Bolsa Familia Programme. Open table in a new tab In addition, the existence of a direct effect of BFP didn't preclude the hypothesis that other causal pathways were involved in the causal model and the importance of other determinants, as showed in previously published studies.4Nery JS Rodrigues LC Rasella D et al.Effect of Brazil's conditional cash transfer programme on tuberculosis incidence.Int J Tuberc Lung Dis. 2017; 21: 790-796Crossref PubMed Scopus (37) Google Scholar, 5Durovni B Saraceni V Puppin MS et al.The impact of the Brazilian Family Health Strategy and the conditional cash transfer on tuberculosis treatment outcomes in Rio de Janeiro: an individual-level analysis of secondary data.J Public Health. 2018; 40: 359-366Crossref Scopus (14) Google Scholar, 6Torrens AW Rasella D Boccia D et al.Effectiveness of a conditional cash transfer programme on TB cure rate: a retrospective cohort study in Brazil.Trans R Soc Trop Med Hyg. 2016; 110: 199-206Crossref PubMed Scopus (50) Google Scholar Bearing in mind that directed acyclic graphs allow us to assess different aspects of an association between covariates in a study, we are only just starting to explore this association, and there is a broad field to be explored. The next steps should be to look for the total effect of BFP by estimating the indirect effects of all causal paths and trying to explain the causal mechanisms. In summary, observational epidemiology has an unavoidable uncertainty that means we cannot be certain that the causal structure under consideration includes the true cause,7Herman MA Robins JM Estimating causal effects from epidemiological data.BMJ. 2006; 60: 553-653Google Scholar but, at the same time, it has a strong theoretical basis that is capable of dealing with causes rooted in social structure. Discussion on this conceptual field is welcomed and contributes to strengthening the scientific evidence, especially at this time, with the emergence of a counter-scientific wave in Brazil. We declare no competing interests. Effect of the Bolsa Familia Programme on the outcome of tuberculosis treatment: a prospective cohort studyBFP alone had a direct effect on tuberculosis treatment outcome and could greatly contribute to the goals of the WHO End TB Strategy. Full-Text PDF Open AccessEffect of the Bolsa Familia Programme on tuberculosis treatment outcomesTuberculosis is strongly influenced by social determinants and has a direct relationship with poverty and social exclusion. Janaina Oliosi and colleagues1 concluded that being a beneficiary of the Bolsa Familia Programme (BFP) was an independent influencing factor on the positive health outcome among people having pharmacological treatment for tuberculosis in Brazil, in terms of cure and lower treatment dropout numbers. Full-Text PDF Open Access