Awards

Vinicius de Araujo Oliveira of Fiocruz in Brazil will develop a framework for the re-use of large clinical and administrative datasets to enable comparative analysis of COVID-19 vaccine safety and effectiveness in Brazil and in Pakistan, with colleagues at Shaukat Khanum Memorial Cancer Hospital and Research Centre there, to improve pandemic responses and promote data-driven evidence generation in the Global South. Monitoring vaccinations across different settings is crucial for containing pandemics. However, comparative analysis of large health datasets in different scenarios is challenging due to concerns around safety and reproducibility and the loss of the context in which the data was collected, which can affect research results. They will adapt data science standards and tools to different local health system scenarios and run individual and joint vaccine effectiveness analyses for the two countries to assess compatibility and reproducibility of the findings. They will also build a public data visualization dashboard for health managers and policymakers to monitor the pandemic, particularly in vulnerable populations.

Vincent Cubaka of Partners In Health in the U.S. will build robust data governance structures to enable the utilization of electronic medical records from multiple countries for research purposes to improve health. So-called FAIR (Findable, Accessible, Interoperable, Reusable) data principles enhance the value of personal medical records for research, and CARE principles were developed to protect the owners of these data. However, the rigidity of these principles can create conflicts, which can make it difficult, for example, to open access to datasets across different countries. To address this, for their current project studying the impact of COVID-19 on chronic care patients across four low- and middle-income countries, they will develop data governance structures and set-up a multi-country community oversight committee to enable full access by researchers to appropriately de-identified individual-level data on a suitable platform.

Haroon Hafeez of Shaukat Khanum Memorial Cancer Hospital and Research Centre in Pakistan will develop a framework for the re-use of large clinical and administrative datasets to enable comparative analysis of COVID-19 vaccine safety and effectiveness in Pakistan and in Brazil, with colleagues at Fiocruz there, to improve pandemic responses and promote data-driven evidence generation in the Global South. Monitoring vaccinations across different settings is crucial for containing pandemics. However, comparative analysis of large health datasets in different scenarios is challenging due to concerns around safety and reproducibility and the loss of the context in which the data was collected, which can affect research results. They will adapt data science standards and tools to different local health system scenarios and run individual and joint vaccine effectiveness analyses for the two countries to assess compatibility and reproducibility of the findings. They will also build a public data visualization dashboard for health managers and policymakers to monitor the pandemic, particularly in vulnerable populations.

Luc Samison of Centre d'Infectiologie Charles Mérieux - University of Antananarivo in Madagascar will support more responsive and resilient antimicrobial resistance (AMR) surveillance systems in Madagascar and Burkina Faso by building a data science center for the electronic collection, analysis and dissemination of data. They will develop and refine data collection tools and sharing processes to promote multi-disciplinary collaborations and strengthen data governance and standards. These will be applied to detecting multi-drug resistant Escherichia coli and Enterobacteriaceae in several settings including pregnant women, hospitalized patients, chickens and surface water. They will also develop new tools and processes to provide stakeholders with strategic AMR indicators in real-time to support decision-making. This project will also support data-centered health research on AMR surveillance and can be applied to a range of pathogen surveillance settings in other low- and middle-income countries.

Damazo Kadengye of the African Population and Health Research Center in Kenya will establish a functional Learning Health System to promote the exploration of population health data from multiple sources to improve public health responses to infectious diseases in sub-Saharan Africa. Utilizing the data revolution to generate new knowledge is crucial for achieving global health targets, but there is a lack of suitable tools and limited access to data from different sources. They will integrate multiple large HIV/AIDS datasets from 11 longitudinal population cohorts in East Africa and develop an organizational architecture that enhances data discoverability and promotes responsible open data sharing, supporting collaborations between healthcare professionals, policy makers and researchers. They will also train scientists to produce data-based evidence using data science tools and predictive statistical models and to work with policy makers at local and national levels.

Andrew Boulle and colleagues at the Western Cape Government Health Department and the University of Cape Town in South Africa will use a data science approach applied to anonymized COVID-19 health data from the government health department including over one million tests and 60,000 hospital admissions, to study the clinical epidemiology and evolution of a new variant of SARS-CoV-2 that emerged in South Africa and the impact on patients with existing health conditions. They will conduct a case-control study to determine the clinical severity of the variant and use a cross-sectional design to explore the evolution of viral load. They will also analyze the impact of COVID-19 on pregnancy by evaluating birth weight and other birth outcomes, such as still births, and use death registries to determine mortality rates in patients with HIV, TB, and diabetes.

Xiaofan Liu at the City University of Hong Kong in China and colleagues will reconstruct COVID-19 transmission chains between individuals in communities and households using statistical methods applied to existing datasets to more reliably estimate COVID-19 transmission characteristics, such as reproduction rates, that are critical for planning effective control measures. Currently, transmission characteristics are estimated using aggregated-level data, which leads to inaccuracies. Ideally, data on how COVID-19 is transmitted between individuals are needed. They will curate an existing collection of datasets containing over 40,000 COVID-19 cases in five Asian countries with person-to-person transmission evidence to reconstruct transmission chains. They will then apply statistical tests and an analytical methodology called regression analysis to identify the most important transmission risk factors, which may include virus strain, transmission media, population density, and climate conditions.

Luis Felipe Reyes at the Universidad de La Sabana in Colombia and colleagues will develop a standardized strategy for researchers to better utilize the ISARIC-COVID-19 dataset, which consists of over 520,000 hospitalized patients from more than 62 countries, and identify the causes and health impacts of severe complications. The dataset is particularly valuable because it covers varying standards-of-care around the world and could be used to study the geographic and time-based variability of the disease. The team will develop a standardized strategy to reformat and clean the ISARIC-COVID-19 dataset by producing data descriptors and reference codes and use this strategy to identify the risk factors and clinical characteristics of COVID-19 complications, such as cardiovascular complications, which are a major contributor to long-term morbidity and mortality, in order that vulnerable patients can be better treated.

Fernando Bozza at Fiocruz in Brazil and colleagues will quantify the real-world value of COVID-19 vaccines in Brazil for protecting individuals from severe disease and for protecting the entire population from being infected. Knowing how effective vaccination is, and how durable the response in the real world is, particularly in low- and middle-income countries, it is critical for ending the pandemic. They will determine the effectiveness of the vaccine for protecting individuals using an approach called test-negative design together with statistical and machine learning approaches to compare the severity of respiratory disease in COVID-19 patients from 43 hospitals. At the population level, they will perform an ecological study, and use regression analysis accounting for inequities to vaccine access, to measure the effect of vaccinations on COVID-19 cases, hospitalizations, and deaths.

Maria Yury Ichihara and colleagues at the Centre for Data and Knowledge Integration for Health (Cidacs) at Fiocruz in Brazil will create a social disparities index to measure inequalities relevant to the COVID-19 pandemic, such as unequal access to healthcare, to identify regions that are more vulnerable to infection and to better focus prevention efforts. In Brazil, markers of inequality are associated with COVID-19 morbidity and mortality. They will develop the index of available COVID-19 surveillance data, hosted on the Cidacs platform, and build a public data visualization dashboard to share the index and patterns of COVID-19 incidence and mortality with the broader community. This will enable health managers and policymakers to monitor the pandemic situation in the most vulnerable populations and target social and health interventions.

Kirsty Le Doare and colleagues at the MRC/UVRI & LSHTM Uganda Research Unit and Makarere University John's Hopkins University in Uganda will develop a model using data collected in real-time to identify the risk factors for adverse pregnancy and infant outcomes caused by the COVID-19 pandemic that can be used to rapidly inform interventions. Lockdowns can severely impact women giving birth and access to maternal, neonatal, and child healthcare. They will apply a Bayesian multivariate network meta-analysis, (a methodology that simultaneously analyses multiple outcomes and multiple treatments, allowing more studies to contribute towards each outcome and treatment comparison) to electronic medical records, leveraging existing data on the effect of the lockdown on antenatal and delivery services for over 30,000 pregnancies, vaccination data, and information on COVID-19 infection in pregnancy and infancy. They will also build a user-friendly data dashboard to support decision-making on infection prevention and control at the Ministry of Health.

Juliane Foseca de Oliveira and colleagues at Fiocruz in Brazil will develop mathematical and statistical methods to model COVID-19 infection transmission, prevention and control across populations in Brazil to better inform local intervention efforts. Social and economic inequalities are known to shape the spread of diseases, therefore the team will integrate existing health data together with social and economic determinants for 5,570 Brazilian cities, as well as assessing data on the effects of the mitigation strategies and social mobility patterns. These data will be used to develop and apply statistical analyses and nonlinear mathematical modelling to forecast disease evolution and outcomes that consider the specific socio-economic conditions, which influence transmission rates. The results will be presented on a user-friendly surveillance platform that can be used by local governments and communities to identify the most effective control methods for their region.

Dale Barnhart and colleagues at Harvard Medical School in the U.S. and Partners in Health of Haiti, Malawi, Mexico, and Rwanda will determine how the COVID-19 pandemic has impacted health care provision and utilization for patients with HIV, heart disease, and diabetes, and the health outcomes of these patients, in all four countries. They will pool existing electronic medical data on chronic care patients collected from up to 30 health facilities in each country and create a harmonized database to identify the impacts of COVID-19 and any successful strategies used to improve care. They will also develop a predictive model to identify which patient populations are most at risk from care disruption during the pandemic, which can help prioritize clinical and geographic areas that need interventions. Finally, they will develop data visualization tools to facilitate the communication and interpretation of the data by chronic care managers across the four different countries.

Carl Marincowitz and colleagues at the University of Sheffield in the United Kingdom and the University of Cape Town in South Africa will develop a risk assessment tool to help emergency clinicians quickly decide whether a patient with suspected COVID-19 needs emergency care or can be safely treated at home to avoid overburdening hospitals particularly in low- and middle- income countries (LMICs). They will use existing data to which they have access on 50,000 patients with suspected COVID-19 infection who sought emergency care in the United Kingdom, South Africa, and Sudan to develop prediction models for specific COVID-19 related outcomes in all income settings. These prediction models will be used to develop risk stratification tools, which enable providers to identify the right level of care and services for distinct subgroups of patients. These will be developed with input from patient and clinical stakeholders. The team will test the performance of their risk assessment tools for identifying high-risk patients with existing triage methods.

Catherine Arsenault at the Harvard T.H. Chan School of Public Health in the U.S. and colleagues will measure the effect of the COVID-19 pandemic and associated containment policies such as curfews on the quality of health care in seven countries and the rates of mortality from non-COVID conditions. They have extracted data from health management information systems spanning two years from Ethiopia, Ghana, Haiti, Laos, Mexico, Nepal, and South Africa. They will first clean the data and then apply an analytical tool called segmented regression analysis to assess the effect of the pandemic on health service delivery, such as the provision of certain preventive and curative services, and use a statistical technique called difference-in-differences estimations to assess the effect of containment policies on healthcare demand, such as patient appointments. This will help countries to address gaps in their health care systems and plan recovery strategies for missed health care.

This project aims to understand and analyze the determinants of vaccination coverage in the Brazilian territory by assessing its association with socio-economic factors, public health spending, coverage of primary health care and the Bolsa Família Program and the influence of patterns of content dissemination on immunization on social media. The results will be disseminated through virtual games, podcasts, interactive panels, infographics, an e-book for municipal managers, a webinar for undergraduate students in the health field and a seminar on World Immunization Day.

This project aims to demonstrate the importance of adequate health care for pregnant women in health facilities and its effects on maternal and child mortality. The project will also measure the causal effect of the distances taken for pregnant women to access health facilities and the health care provided, which will be useful to identify the gaps in maternal health care and to drive decisions on resources allocation and prioritized actions. A new database will be created for monitoring of access to public health facilities, considering distances and real-world accessibility conditions.

This project will link the SINASC, SIM, SIGA and SIH databases to assess the role of interventions in childbirth and will also analyze the avoidability of neonatal and maternal outcomes. It will develop guidelines and training for the classification of “onset of childbirth”. The group will use the Robson classification, map the disruptions of the pandemic to perinatal care, and will explore the role of maternal nutritional status on intervention rates in childbirth and on maternal and neonatal results, making the databases available for research and for training purposes.

In Brazil, the only information system that provides data on maternal morbidity is the Hospital Information System (SIH), but there are difficulties in implementing the criteria recommended by WHO and doubts about the quality of information. The project will validate the SIH against the Maternal Near Miss criteria and will build and validate an algorithm to identify severe maternal morbidity. As a product, it will develop an online indicator panel for the surveillance of maternal health to be used by SUS managers.

The objective of this project is to map the implementation and evolution of breastfeeding initiatives in the scope of primary health care (PHC), assessing their spatial and temporal distribution patterns and their correlation with the evolution of the indicators. The team will map the successful cases of pro-breastfeeding programs and identify their impact. A single dataset containing information on PHC infrastructure and breastfeeding rates and programs will be generated, as well as the qualification and validation of SISVAN breastfeeding information.

The project will identify the poverty profiles of families with children under the age of five enrolled in the Single Registry considering data from the Family Development Index, the Brazilian Deprivation Index and the Municipal Human Development Index. It will assess the amount needed to overcome the families' poverty gap and also estimate the mortality rates in each of these profiles. The results will be disseminated through dashboards, forecasting scenarios for stakeholders and a workshop with stakeholders from the Ministry of Citizenship and stakeholders. The project will also provide forecast scenarios for COVID-19 associated crises with a high level of granularity.

This project will identify clinical, sociodemographic, psychosocial, neurocognitive and epigenomic factors to assist in the identification of the most effective response to the treatment offered by SUS to detoxify the use of crack and cocaine by women. The project will use the Random Forest algorithm in a database developed by the research group itself in order to predict the factors that impact adherence and maintenance of abstinence among users.

This project will produce a consolidated database, which aggregates available data on emerging diseases (Zika and COVID-19); external climatic conditions (droughts and floods) and environmental problems (disasters, fires, pollution) as risk factors for unfavorable obstetric and neonatal outcomes. The database will enable the production of information on maternal and early neonatal morbidity and mortality in an accessible way.

The objective of this project is to create an obstetric observatory through an interactive platform for monitoring, analyzing public data and disseminating information in the area of ​​Obstetrics in Brazil. It will provide exploratory data analysis with the purpose of assessing the impacts of the H1N1 (2009) and COVID-19 (2020) pandemics on maternal, fetal and neonatal health. A book entitled on the subject will also be produced and made available free of charge.

This project will investigate the joint effects of exposure to particulate matter, carbon monoxide, ozone, nitrogen dioxide and sulfur dioxide on maternal and child health. The joint and non-linear effects of these pollutants on maternal and child health will be evaluated, followed by those from wild fires and fires in the northern region. As a product, it will develop a map of multiple air pollution exposure levels by municipality, and it intends to create a definition for critical air pollution exposure levels for child health to inform pollution guidelines in Brazil and thus improve the environmental conditions for the development of children in the country.

Since the gestational weight curves (GWC) adopted by the Ministry of Health are based only on neonatal outcomes and on maternal and child outcomes, this project aims to fill these gaps and determine the GWC ranges for the four categories of pre-gestational corporal mass index. As products, it will develop guidelines for the use of the curves and recommendations, packages to increase the capacity-building of healthcare professionals, policymakers and students and an interactive dashboard using SISVAN data for continuous surveillance.

The purpose of this proposal is to develop a system for the identification, monitoring and forecasting of maternal near-miss cases. The project will seek to create a classifier that identifies near-miss cases based on historical data and a predictive model to infer the number of annual near misses in a region. In addition, it will also update near-miss historical number data to forecast the annual number by location.

Combining multiple data sets from HBGDKi using ML tools for prediction, classification and topic discovery may yield new insights for adverse birth outcomes and intermediate outcomes of interest. The study is based on a set of epidemiological, clinical and biochemical variables risk stratification algorithms for various adverse outcomes with practical applicability in health programme, and clinical settings may be feasible to develop using ML tools.ML can be used to suitably impute/bin missing values within datasets and merge variables from multiple datasets using robust data triangulation algorithms.

Angela Dramowski of Stellenbosch University in South Africa will determine the most effective antibiotics to use and when to use them for treating bloodstream infections in neonates to reduce mortality rates. In Sub-Saharan Africa, a quarter of a million deaths in children under five are caused by bacterial infections during the first 28 days of life. The causal bacteria and their susceptibility to antibiotics change over time and across different regions, thus standard treatment guidelines are likely outdated. They will assemble infection datasets collected between 2016 and 2018 from eight Sub-Saharan African neonatal units that include the bacterial types and antibiotic susceptibility, the treatment given, and the outcome. They will then develop a probabilistic decision-tree model to estimate the impact of using alternative antibiotics on neonatal mortality compared to the standard treatments, as well as to determine the best timing for recommending different treatment strategies based on age at infection onset.

The study proposes on pregnant women (Garbh-ini) cohort, a multidimensional longitudinal dataset purposely designed to study preterm birth. The study will apply data-driven machine learning approaches to develop an accurate and clinically useful model to predict the risk of preterm births. It will use multiple models for classification, with better objective functions and misclassification penalties that will aid in a higher rate of accurate predictions, and resampling of the data to avert biases arising from class imbalance. The primary deliverable will be dynamic prediction models that can predict, at different periods of gestation, the PTB risk using the clinical, epidemiological and imaging data.

The study aims to calculate cut-offs using data provided by HBGDki and datasets with SAS, SJRI where weight, height, and age are available for children below five years in combination with other outcomes such as death, morbidity or hospitalization. Using WHO standards, weight for height, height for age and weight for age will be calculated, and these metrics will be used as determinants for risk of death, morbidity/hospitalization. A finer categorization of malnutrition based on the risk of mortality or significant morbidity can be used to develop and then deliver tailored optimized therapeutic options for what is essentially a far more eclectic group than what is captured by a three-category classification MAM, SAM and others.

Geoffrey Arunga of BroadReach in Kenya will develop a digital assessment tool to identify women with the highest risk of maternal morbidity and adverse pregnancy outcomes, and their causes, and to inform clinicians and health policies to improve maternal health and survival. They will apply advanced statistical analyses and machine learning techniques to clinical, social, and economic data from an existing longitudinal study of pregnant women in West Africa to identify the data that can best predict risk. This will be used to derive a minimum set of questions that can be incorporated into a digital tool for health workers to assess a woman’s risk at any given timepoint during pregnancy. The tool will be pilot tested for feasibility and predictive performance in rural and urban-based health facilities in Ghana.

Joseph Akuze Waiswa of Makerere University College of Health Sciences, School of Public Health in Uganda will leverage their dataset on pregnancy outcomes collected by household survey in four African sites to improve the quality of data on stillbirth and neonatal death rates to optimize interventions and investments. Around half the global burden of stillbirths and neonatal deaths occurs in Africa. However, the total numbers are derived from household surveys, and the quality of the data is relatively poor, with missing events and misclassifications of the cause of death. To address this, they recently performed a large-scale multi-site study in Africa to collect relevant household survey data on over 68,000 pregnancies using an Android-based platform. They will further analyze this dataset to identify factors that influence the quality of the data collected, such as the type and structure of the questions, to develop more effective survey questions and inform training for interviewers.

Intend to adopt unconventional data analytical techniques to explore the multiple dimensions of child undernutrition in India, utilizing existing national surveys and the HBGDKi database. The novel and sophisticated analytical methods that will be used include geospatial analysis, data triangulation through statistical matching, and multilevel modeling. The idea is unconventional because India does not have a single granular and multidimensional Health and Nutrition survey that can be analyzed at the district and sub-district level to provide precision insights to public health policy.

The study aims to explore regional trends and variations in vaccine uptake, uncover relationships to other socioeconomic, demographic, and public health indicators, and develop a predictive model of the state of vaccine confidence in different parts of India. This will infer local-level confidence in vaccines by identifying areas with good access to healthcare infrastructure. The main goal of the proposal to develop a prototype coverage monitoring and forecasting system across districts by using Gaussian process methods.

Naeemah Abrahams of the South African Medical Research Council will study the impact of gender-based violence on the health of mothers and children in South Africa to better inform prevention and health strategies. Gender-based violence affects one in three women globally, and rates are high in Africa, particularly for women of reproductive age. However, the effects on their health remain largely unknown, in part because of the lack of follow-up studies. They will analyze data from an ongoing longitudinal study, covering over 36 months, of over 1,600 women in South Africa, and they will perform statistical analyses to identify associations between the types of violence that they suffered, such as from intimate partners or during pregnancy, and the effects on HIV acquisition, pregnancy outcomes and their mental health. They will also identify factors that can influence these outcomes, such as physical injuries or lack of social support.

Adeladza Kofi Amegah of the University of Cape Coast in Ghana will investigate how diet, the environment and low birth weight lead to child undernutrition in socially-disadvantaged communities in West Africa. Studies have shown strong associations between socio-economic status and child undernutrition, but they have not identified the actual causes, which is critical for developing effective interventions. They will evaluate potentially causative dietary factors such as meal diversity and frequency, and environmental factors including water, sanitation, and hygiene practices, along with birth weight. They will analyze four demographic and health survey (DHS) datasets from the last twenty years that include birth weight, infant height and weight, wealth status, and education attainment. They will use various modeling approaches and statistical analyses to identify the associations between the different parameters, and a sequential causal mediation analysis to establish the role of specific factors, and combinations of those factors, on undernutrition.

Said Mohammed Ali from the Ministry of Health in Tanzania will use machine learning to develop a model for predicting gestational age based on routinely collected data and use it to better define when a birth should be classified as premature in Africa and thereby at higher risk of neonatal death. Premature births in Africa are currently defined as those occurring before 37 weeks. However, data suggests that births occurring just after this cut-off are also at higher risk of neonatal death, suggesting that the definition needs reevaluating. They will do this by developing an artificial neural network-based model using existing data, such as the date of the last menstrual period and pregnancy history, from 5,000 pregnancies with an ultrasound-verified gestational age to identify accurate predictors of gestational age. This will then be used to estimate gestational age on an additional 20,000 pregnancies, and further to identify the age cut-off that best predicts neonatal mortality, using logistic regression models.

Christopher Seebregts of Jembi Health Systems NPC in South Africa will use a data science approach to improve maternal, newborn, and child health by developing algorithms that integrate diverse personal and clinical data taken from disparate sources to make them more informative. To test their approach, they will apply it to existing data taken in the Tshwane health district in South Africa for a program looking to prevent mother-to-child transmission of HIV. These data include quantitative diagnostic readouts on HIV status for mother and child, therapy initiation reports, household and socio-economic data, and district health information, all produced by different groups using different formats. They will use data science methodology and algorithms to integrate all the data to produce longitudinal and cohort data and apply machine learning techniques to identify predictors of failure to prevent mother-to-child HIV transmission, and of mother or infant mortality.

Sikolia Wanyonyi of Aga Khan University in Kenya will analyze datasets on maternal mortality during hospital deliveries to determine the causes and to develop prediction models to help identify effective interventions in specific settings. Maternal mortality rates in Kenya are only slowly reducing, despite the increase in hospital deliveries, which may be due to a combination of different factors such as the quality of care and clinical characteristics. They have assembled available data from the Kenyan Ministry of Health for different counties over the last five years and will apply a hierarchical Bayesian model to identify the trends and their causes. They will also fit so-called generalizable estimating equations to the data to determine whether the risk of maternal mortality can be predicted from combinations of specific types of data, such as socio-demographic, which could be used to identify high-risk patients for more timely treatments.

Anthony Ngugi of Aga Khan University in Kenya will use a modeling approach to determine the optimal allocation of limited child nutrition budgets that will most effectively reduce mortality and morbidities, like stunting and anemia, caused by malnutrition. They will use the Optima Nutrition modeling tool, which combines cost functions with an epidemiological model, to make predictions about the cost-efficacy of different funding allocations, for example on food fortification or education. They will focus on the 24 counties in Kenya with the highest burdens of malnutrition and assemble local academics and collaborators at the National Ministry of Health to help collect and harmonize data for the modelling analysis, including existing nutrition-related datasets and health budgets. They will run and validate the model, and then test different optimization algorithms to identify the most effective funding allocations.

Lucas Malla of the Kemri-Wellcome Trust Research Program in Kenya will apply a novel statistical method to determine how the timing and rates of gestational weight gain during pregnancy affect maternal and child health in Africa to identify risk factors and those at highest risk. Excessive or insufficient gestational weight gain can predict adverse maternal and child health outcomes such as gestational diabetes, preterm birth, and infant mortality. They will use six existing longitudinal datasets with over 40,000 data-points on gestational weight gain and apply a univariate and multivariate SITAR (super-imposition by translation and rotation) growth curve analysis approach to model the gestational weight trajectories. They will then use regression techniques to identify specific parameters (size, timing, and velocity of the weight gain) that can predict specific health outcomes.

The project aims to solve the problem of a large number of low-weight births in the Indian population and the inability to predict the risk reliably in the antenatal period. The study plan goes beyond finding a single metric or rule aiming to describe the risk of Small-for-Gestational Age (SGA) in all women plan to use clustering with factor-selection on the various available data-sets. They will use algorithm development and in Bayesian modelling. At the individual level, an identification of possible causal factors behind the prediction of SGA, which will aid the clinician in determining the appropriate intervention

Datasets within the purview of HBGDKi with data on duration of breastfeeding, anthropometric measurements, and socio-demographic characteristics will be extracted, compiled and harmonized. WAZ, WHZ and HAZ scores will be calculated for each child. The study may describe the breastfeeding patterns in individual studies using a survival analysis approach and overall through a meta-analytic approach.

Aims to access all 68.3 million living births certificates from Brazil, from 1994 to 2016, and compare them with breastfeeding policies in all Brazilian hospitals to assess the impact of the initiatives on infant health. The study also plans to estimate the number of avoidable deaths during this time period, if those initiatives were adopted in Brazil.

The study is aimed at evaluating the effectiveness of Mãe Coruja intervention in reducing low birthweight and preterm birth. By using appropriate statistical methods, the study will use the Cidacs dataset combined with the data from Mãe Coruja program to carry out the quasi-experimental study. With the support of machine learning techniques, the project will also Identify social, economic, geographic and environmental conditions that are associated with the outcomes. The researchers will also build an index of perinatal health risk to inform improvements in targeting populations and the deployment of similar strategies and programs elsewhere in Brazil.

Identifying the preventable causes and performing early risk stratification of pregnant women are instrumental to develop strategies to prevent and reduce preterm birth (PTB). The ability to identify at-risk pregnancies and to enroll women in prevention strategies has been difficult due to complexity of associated risk factors. The study aims to combine different national level data sources to understand the main predictors of PTB and develop a machine-learning-based predictive model to conduct automated risk stratification at the point of care level, integrated with advanced data visualization for clinical decision support.

The project will develop a platform to provide services for decision-making support for neonatal death preventive actions by using data from CIDACS cohort. The platform will offer three services: cohort data visualization for decision-making support by comparative human visual analysis, prediction of risk of neonatal death based on machine learning models, and simulator of public policies impact influencing on the risk of neonatal death.

The study aims to develop an Early Childhood Development friendly index (ECD-FI) based on a core set of evidence-based nurturing care indicators to assess the factors contributing to enabling environments and promote ECD at the municipal level by monitoring and identifying opportunities to scale up ECD programs. The index will be created through machine learning and will run analytical models considering demographic information and risk factors at the municipal level. This disaggregated data is not available in Brazil.

Seeks to understand the impacts of the Bolsa Família conditional cash transfer on birth outcomes (e.g., birth weight, gestational weeks, etc). The proposed design will disentangle the measured effects into two components: one that is associated to the cash transfer; and another related to prenatal care assistance. Moreover, this strategy will allow the researchers to determine the window of opportunity where CCT interventions exhibit highest impacts on birth outcomes, recognizing heterogeneous impacts according to how early in the pregnancy the CCT intervention starts.

Studies show that seasonal influenza in Ceará, in the Northeast region of Brazil, occurs 2 to 3 months earlier than in the South and Southeast, which guides the national calendar of vaccination. By using data science approaches, the study will test if Brazil's current national policy targeting vaccination only during the months of April and May inadequately protects against the harmful maternal-fetal effects of influenza in the Semi-Arid and northern regions of Brazil. If the hypothesis confirms, the study has the potential to change policy and modify the vaccination calendar.

Aims to validate the International Fetal and Newborn Growth Consortium for the 21st century (Intergrowth-21st) standards for gestational weight gain (GWG) and create new recommendations of GWG based on those standards for first trimester normal and overweight women to be used in the Brazilian Unified Health System (SUS). GWG recommendations currently used in SUS have not been properly tested or validated, thus the project might improve prenatal nutritional care and reduce post gestational weight retention.

The main goal of the project is to develop and explore an innovative measure of gestational age - "potential pregnancy days lost" (PPDL) - to produce evidence of its association with maternal and child health, morbidity and mortality in the short, medium and long term. The indicator also aims to convince women and policy makers about the need to promote less interventions and "harm-free care" during pregnancy.

By analyzing national children vaccination coverage from spatial perspectives, the study aims to uncover insights into the traditional surveillance. This will help to identify coverage rates, regions of greater vulnerability by providing a differentiated look at the logic of equity in health. Understanding the low childhood vaccination coverage will help to guide public policies for the purpose of interventions.

This research aims to analyze the relationship between a conditional cash transfer program and the child's health, considering two generations of the families and using two different approaches: econometric analysis and data mining algorithms. By analyzing the long term impacts of Bolsa Familia program on future generations' health performance, the project will investigate if a child who was born in a family whose grandparents received the cash transfer is in better health conditions than a similar child born in a family whose grandparents did not receive the same benefit.

The proposal will develop a platform for the analysis and visualization of data that will allow managers, public servants and other stakeholders involved in the Mãe Coruja Program at Pernambuco state (PMCP) to extract strategic information to improve the intervention. The focus will be on the implementation and actual enforcement of public policies, considering the high gestational risk and sexually transmitted infections (STI). Currently, health databases are for consultations only. The innovation of this proposal is to create an intelligent cloud platform for the analysis and distribution of health information to improve health care of women enrolled in PMCP.

Infectious diseases may have only transitory impacts on pregnant mothers, but they can have lasting impacts on children. Can public interventions mitigate these impacts? This project aims to identify how exposure to localized epidemiological risk factors in the fetal period influences developmental outcomes for children through the early years of life. The researchers propose to evaluate in what extent the access to primary health care and social welfare programs mitigate negative impacts in child development.

Does air pollution affect the rates of stillbirths, congenital malformations and neonatal mortality? This study aims to answer this question by merging the child health data collected within the 100 Million Brazilian Cohort from Cidacs with high-resolved satellite-derived data on air pollution to establish critical ambient air pollution thresholds for preventing adverse birth outcomes and malformations based on concentrations of fine particles, PM 2.5.