Seminar Series: Ngianga-Bakwin Kandala

Bayesian Diseases mapping of Global Public Health issues: what can we learn from complex household surveys from Africa

Ngianga-Bakwin Kandala

Department of Epidemiology and Biostatistics
Schulich School of Medicine & Dentistry
Western University

Distinguished Professor
Division of Epidemiology and Biostatistics
School of Public Health
University of Witwatersrand (South Africa)

Short Biography:
Kandala is Professor of Biostatistics at the Department of Epidemiology and Biostatistics, University of Western Ontario, Canada and a Distinguished Professor of Biostatistics at the School of Public Health, University of Witwatersrand, South Africa.
Kandala pioneered a Bayesian spatial modelling of maternal and child health outcomes using large scale household data in Africa the approach now widely used in spatial demography. He is the current Editor-in-Chief of the Journal of African Population Studies. Over 20 years, his main research interests are in capacity building in Biostatistics in Africa and Bayesian statistical methods and their application to epidemiology and population health including maternal and child health both in the developing countries and command economies, using large scale household data. His four recent books are titled (1) Advance Techniques in modelling Maternal and child health in Africa, Springer (2) Female Mutilation around the World: Analysis of Medical Aspects, Law and Practice (2018) with Springer Nature, (3) Statistical Modelling of Complex correlated and clustered data using Household surveys in Africa (2019) with Nova Science Publishers, and (4) Adolescents and Youth sexual and reproductive health in Central Africa. L’ Harmattan, Paris, France (2021).

Hierarchical spatial modelling is a common and useful approach for modelling complex spatially correlated data in many settings in epidemiology, public health and development studies. Most of the data collected by many African governments through household surveys and sentinel surveillance are geo-referenced by districts, counties, provinces or other administrative units.
Because of the complexity of factors associated with survival and health in Africa, traditional measures such as household socioeconomic and education may require supplementation with types of data that are both novel and less conventional. Statistical techniques that incorporate spatial analysis using a combination of data sources and spatial covariates offer such possibility, though broadening the view of environment at both the macro level and the micro level , may be required to understand fully the scope of such influences.
Some applications of spatial Bayesian models based on complex surveys data will be presented in this talk. The modelling relies on Markov Chain Monte Carlo (MCMC) simulation techniques. The models cover a number of well-known model classes as special cases, including Generalized additive models (Hastie & Tibshirani 1990), Generalized additive mixed models (Lin & Zhang 1999), Geoadditive models (Kammann & Wand 2003), varying coefficient models (Hastie & Tibshirani 1993), and Geographically weighted regression (Fotheringham, Brunsdon & Charlton (2002).

Bayesian statistical methods and their application to epidemiology and global health, Spatial statistics/Tropical Diseases mapping, Cluster randomization trials, Prospective epidemiological studies, Evidence Synthesis, Capacity building in Biostatistics in Africa

Date: Friday, March 10th
Time: 1:30 pm - 2:30 pm
Location: PHFM 3015 (Western Centre for Public Health and Family Medicine)