Incidence and risk factors of tuberculosis among 420 854 household contacts of patients with tuberculosis in the 100 Million Brazilian Cohort (2004-18): a cohort study.

BACKGROUND: Although household contacts of patients with tuberculosis are known to be particularly vulnerable to tuberculosis, the published evidence focused on this group at high risk within the low-income and middle-income country context remains sparse. Using nationwide data from Brazil, we aimed to estimate the incidence and investigate the socioeconomic and clinical determinants of tuberculosis in a cohort of contacts of tuberculosis patients. METHODS: In this cohort study, we linked individual socioeconomic and demographic data from the 100 Million Brazilian Cohort to mortality data and tuberculosis registries, identified contacts of tuberculosis index patients diagnosed from Jan 1, 2004 to Dec 31, 2018, and followed up the contacts until the contact's subsequent tuberculosis diagnosis, the contact's death, or Dec 31, 2018. We investigated factors associated with active tuberculosis using multilevel Poisson regressions, allowing for municipality-level and household-level random effects. FINDINGS: We studied 420 854 household contacts of 137 131 tuberculosis index patients. During the 15 years of follow-up (median 4·4 years [IQR 1·9-7·6]), we detected 8953 contacts with tuberculosis. The tuberculosis incidence among contacts was 427·8 per 100 000 person-years at risk (95% CI 419·1-436·8), 16-times higher than the incidence in the general population (26·2 [26·1-26·3]) and the risk was prolonged. Tuberculosis incidence was associated with the index patient being preschool aged (<5 years; adjusted risk ratio 4·15 [95% CI 3·26-5·28]) or having pulmonary tuberculosis (2·84 [2·55-3·17]). INTERPRETATION: The high and sustained risk of tuberculosis among contacts reinforces the need to systematically expand and strengthen contact tracing and preventive treatment policies in Brazil in order to achieve national and international targets for tuberculosis elimination. FUNDING: Wellcome Trust and Brazilian Ministry of Health.

Ethnoracial inequalities and child mortality in Brazil: a nationwide longitudinal study of 19 million newborn babies.

BACKGROUND: Racism is a social determinant of health inequities. In Brazil, racial injustices lead to poor outcomes in maternal and child health for Black and Indigenous populations, including greater risks of pregnancy-related complications; decreased access to antenatal, delivery, and postnatal care; and higher childhood mortality rates. In this study, we aimed to estimate inequalities in childhood mortality rates by maternal race and skin colour in a cohort of more than 19 million newborns in Brazil. METHODS: We did a nationwide population-based, retrospective cohort study using linked data on all births and deaths in Brazil between Jan 1, 2012, and Dec 31, 2018. The data consisted of livebirths followed up to age 5 years, death, or Dec 31, 2018. Data for livebirths were extracted from the National Information System for livebirths, SINASC, and for deaths from the Mortality Information System, SIM. The final sample consisted of complete data for all cases regarding maternal race and skin colour, and no inconsistencies were present between date of birth and death after linkage. We fitted Cox proportional hazard regression models to calculate the crude and adjusted hazard ratios (HRs) and 95% CIs for the association between maternal race and skin colour and all-cause and cause-specific younger than age 5 mortality rates, by age subgroups. We calculated the trend of HRs (and 95% CI) by time of observation (calendar year) to indicate trends in inequalities. FINDINGS: From the 20 526 714 livebirths registered in SINASC between Jan 1, 2012, and Dec 31, 2018, 238 436 were linked to death records identified from SIM. After linkage, 1 010 871 records were excluded due to missing data on maternal race or skin colour or inconsistent date of death. 19 515 843 livebirths were classified by mother's race, of which 224 213 died. Compared with children of White mothers, mortality risk for children younger than age 5 years was higher among children of Indigenous (HR 1·98 [95% CI 1·92-2·06]), Black (HR 1·39 [1·36-1·41]), and Brown or Mixed race (HR 1·19 [1·18-1·20]) mothers. The highest hazard ratios were observed during the post-neonatal period (Indigenous, HR 2·78 [95% CI 2·64-2·95], Black, HR 1·54 [1·48-1·59]), and Brown or Mixed race, HR 1·25 [1·23-1·27]) and between the ages of 1 year and 4 years (Indigenous, HR 3·82 [95% CI 3·52-4·15]), Black, HR 1·51 [1·42-1·60], and Brown or Mixed race, HR 1·30 [1·26-1·35]). Children of Indigenous (HR 16·39 [95% CI 12·88-20·85]), Black (HR 2·34 [1·78-3·06]), and Brown or Mixed race mothers (HR 2·05 [1·71-2·45]) had a higher risk of death from malnutrition than did children of White mothers. Similar patterns were observed for death from diarrhoea (Indigenous, HR 14·28 [95% CI 12·25-16·65]; Black, HR 1·72 [1·44-2·05]; and Brown or Mixed race mothers, HR 1·78 [1·61-1·98]) and influenza and pneumonia (Indigenous, HR 6·49 [95% CI 5·78-7·27]; Black, HR 1·78 [1·62-1·96]; and Brown or Mixed race mothers, HR 1·60 [1·51-1·69]). INTERPRETATION: Substantial ethnoracial inequalities were observed in child mortality in Brazil, especially among the Indigenous and Black populations. These findings demonstrate the importance of regular racial inequality assessments and monitoring. We suggest implementing policies to promote ethnoracial equity to reduce the impact of racism on child health. FUNDING: MCTI/CNPq/MS/SCTIE/Decit/Bill & Melinda Gates Foundation's Grandes Desafios Brasil, Desenvolvimento Saudável para Todas as Crianças, and Wellcome Trust core support grant awarded to CIDACS-Center for Data and Knowledge Integration for Health.

Conditional cash transfer program and child mortality: A cross-sectional analysis nested within the 100 Million Brazilian Cohort.

BACKGROUND: Brazil has made great progress in reducing child mortality over the past decades, and a parcel of this achievement has been credited to the Bolsa Família program (BFP). We examined the association between being a BFP beneficiary and child mortality (1-4 years of age), also examining how this association differs by maternal race/skin color, gestational age at birth (term versus preterm), municipality income level, and index of quality of BFP management. METHODS AND FINDINGS: This is a cross-sectional analysis nested within the 100 Million Brazilian Cohort, a population-based cohort primarily built from Brazil's Unified Registry for Social Programs (Cadastro Único). We analyzed data from 6,309,366 children under 5 years of age whose families enrolled between 2006 and 2015. Through deterministic linkage with the BFP payroll datasets, and similarity linkage with the Brazilian Mortality Information System, 4,858,253 children were identified as beneficiaries (77%) and 1,451,113 (23%) were not. Our analysis consisted of a combination of kernel matching and weighted logistic regressions. After kernel matching, 5,308,989 (84.1%) children were included in the final weighted logistic analysis, with 4,107,920 (77.4%) of those being beneficiaries and 1,201,069 (22.6%) not, with a total of 14,897 linked deaths. Overall, BFP participation was associated with a reduction in child mortality (weighted odds ratio [OR] = 0.83; 95% CI: 0.79 to 0.88; p < 0.001). This association was stronger for preterm children (weighted OR = 0.78; 95% CI: 0.68 to 0.90; p < 0.001), children of Black mothers (weighted OR = 0.74; 95% CI: 0.57 to 0.97; p < 0.001), children living in municipalities in the lowest income quintile (first quintile of municipal income: weighted OR = 0.72; 95% CI: 0.62 to 0.82; p < 0.001), and municipalities with better index of BFP management (5th quintile of the Decentralized Management Index: weighted OR = 0.76; 95% CI: 0.66 to 0.88; p < 0.001). The main limitation of our methodology is that our propensity score approach does not account for possible unmeasured confounders. Furthermore, sensitivity analysis showed that loss of nameless death records before linkage may have resulted in overestimation of the associations between BFP participation and mortality, with loss of statistical significance in municipalities with greater losses of data and change in the direction of the association in municipalities with no losses. CONCLUSIONS: In this study, we observed a significant association between BFP participation and child mortality in children aged 1-4 years and found that this association was stronger for children living in municipalities in the lowest quintile of wealth, in municipalities with better index of program management, and also in preterm children and children of Black mothers. These findings reinforce the evidence that programs like BFP, already proven effective in poverty reduction, have a great potential to improve child health and survival. Subgroup analysis revealed heterogeneous results, useful for policy improvement and better targeting of BFP.

CIDACS-RL: a novel indexing search and scoring-based record linkage system for huge datasets with high accuracy and scalability.

BACKGROUND: Record linkage is the process of identifying and combining records about the same individual from two or more different datasets. While there are many open source and commercial data linkage tools, the volume and complexity of currently available datasets for linkage pose a huge challenge; hence, designing an efficient linkage tool with reasonable accuracy and scalability is required. METHODS: We developed CIDACS-RL (Centre for Data and Knowledge Integration for Health - Record Linkage), a novel iterative deterministic record linkage algorithm based on a combination of indexing search and scoring algorithms (provided by Apache Lucene). We described how the algorithm works and compared its performance with four open source linkage tools (AtyImo, Febrl, FRIL and RecLink) in terms of sensitivity and positive predictive value using gold standard dataset. We also evaluated its accuracy and scalability using a case-study and its scalability and execution time using a simulated cohort in serial (single core) and multi-core (eight core) computation settings. RESULTS: Overall, CIDACS-RL algorithm had a superior performance: positive predictive value (99.93% versus AtyImo 99.30%, RecLink 99.5%, Febrl 98.86%, and FRIL 96.17%) and sensitivity (99.87% versus AtyImo 98.91%, RecLink 73.75%, Febrl 90.58%, and FRIL 74.66%). In the case study, using a ROC curve to choose the most appropriate cut-off value (0.896), the obtained metrics were: sensitivity = 92.5% (95% CI 92.07-92.99), specificity = 93.5% (95% CI 93.08-93.8) and area under the curve (AUC) = 97% (95% CI 96.97-97.35). The multi-core computation was about four times faster (150 seconds) than the serial setting (550 seconds) when using a dataset of 20 million records. CONCLUSION: CIDACS-RL algorithm is an innovative linkage tool for huge datasets, with higher accuracy, improved scalability, and substantially shorter execution time compared to other existing linkage tools. In addition, CIDACS-RL can be deployed on standard computers without the need for high-speed processors and distributed infrastructures.

Examining the quality of record linkage process using nationwide Brazilian administrative databases to build a large birth cohort.

BACKGROUND: Research using linked routine population-based data collected for non-research purposes has increased in recent years because they are a rich and detailed source of information. The objective of this study is to present an approach to prepare and link data from administrative sources in a middle-income country, to estimate its quality and to identify potential sources of bias by comparing linked and non-linked individuals. METHODS: We linked two administrative datasets with data covering the period 2001 to 2015, using maternal attributes (name, age, date of birth, and municipally of residence) from Brazil: live birth information system and the 100 Million Brazilian Cohort (created using administrative records from over 114 million individuals whose families applied for social assistance via the Unified Register for Social Programmes) implementing an in house developed linkage tool CIDACS-RL. We then estimated the proportion of highly probably link and examined the characteristics of missed-matches to identify any potential source of bias. RESULTS: A total of 27,699,891 live births were submited to linkage with maternal information recorded in the baseline of the 100 Million Brazilian Cohort dataset of those, 16,447,414 (59.4%) children were found registered in the 100 Million Brazilian Cohort dataset. The proportion of highly probably link ranged from 39.3% in 2001 to 82.1% in 2014. A substantial improvement in the linkage after the introduction of maternal date of birth attribute, in 2011, was observed. Our analyses indicated a slightly higher proportion of missing data among missed matches and a higher proportion of people living in an urban area and self-declared as Caucasian among linked pairs when compared with non-linked sets. DISCUSSION: We demonstrated that CIDACS-RL is capable of performing high quality linkage even with a limited number of common attributes, using indexation as a blocking strategy in larg e routine databases from a middle-income country. However, residual records occurred more among people under worse living conditions. The results presented in this study reinforce the need of evaluating linkage quality and when necessary to take linkage error into account for the analyses of any generated dataset.

On the Accuracy and Scalability of Probabilistic Data Linkage Over the Brazilian 114 Million Cohort.

Data linkage refers to the process of identifying and linking records that refer to the same entity across multiple heterogeneous data sources. This method has been widely utilized across scientific domains, including public health where records from clinical, administrative, and other surveillance databases are aggregated and used for research, decision making, and assessment of public policies. When a common set of unique identifiers does not exist across sources, probabilistic linkage approaches are used to link records using a combination of attributes. These methods require a careful choice of comparison attributes as well as similarity metrics and cutoff values to decide if a given pair of records matches or not and for assessing the accuracy of the results. In large, complex datasets, linking and assessing accuracy can be challenging due to the volume and complexity of the data, the absence of a gold standard, and the challenges associated with manually reviewing a very large number of record matches. In this paper, we present AtyImo, a hybrid probabilistic linkage tool optimized for high accuracy and scalability in massive data sets. We describe the implementation details around anonymization, blocking, deterministic and probabilistic linkage, and accuracy assessment. We present results from linking a large population-based cohort of 114 million individuals in Brazil to public health and administrative databases for research. In controlled and real scenarios, we observed high accuracy of results: 93%-97% true matches. In terms of scalability, we present AtyImo's ability to link the entire cohort in less than nine days using Spark and scaling up to 20 million records in less than 12s over heterogeneous (CPU+GPU) architectures.