RESEARCH

 

Spatial Data Mining
Exploring Co-Location of Adverse Birth Outcomes and Environmental Variables Project

(Authoured by Osnat Wine) This interdisciplinary research project was jointly developed by researchers and knowledge users in response to the need to understand how environmental factors impact health. The project will investigate chemical and socioeconomic environmental influences on maternal/infant health, a major knowledge gap to date. The objective of the project, using large publicly funded databases, is to investigate association patterns among adverse birth outcomes (ABO) (i.e. low birth weight, preterm births, stillbirths and perinatal mortality), multiple pollutants released by industry into the air, and socioeconomic variables (e.g. income, education). We anticipate that findings will help determine the co-location/spatial association of ABO with multiple combinations of environmental and other known risk factors and serve as foundation for future research.

Data used for this research will be obtained from Environment Canada (industrial pollution, wind data), CAREX Canada (national PM2.5 and NO2 air pollution models), Statistics Canada (birth rates and ABO, socioeconomic variables), the Canadian Perinatal Programs Coalition and the CIHR funded Maternal Infant Care (MiCare) program (birth rates, ABO and maternal related variables).

Newly developed Data mining methodologies will be employed to explore multipollutant interactions, as well as potential interactions of these pollutants with adverse birth outcomes and socioeconomic and biological variables. Data mining is the process of discovering previously unknown patterns from large data sets involving methods from artificial intelligence, statistics and database management. Maps to portray potential exposures will illustrate geo--?spatial distribution of pollutants, ABO, wind patterns and socioeconomic variables. Specialists in computing sciences, geographical information systems (GIS), exposure assessment, adverse birth outcomes, epidemiology, bio?statistics and environmental health will assess potential interactions between variables Our knowledge users include Health Canada, the Canadian Perinatal Programs Coalition, MiCare, and the Canadian Partnership for Children's Health & Environment as integral partners. Additional analysis will incorporate other maternal variables and models of urban air pollution to test for other known participating factors in ABO. Dissemination areas will serve as the smallest geographic unit used to aggregate and analyze the data.

The complexity of the study requires collaboration, between academic disciplines and between researchers and knowledge users. To promote this collaboration, the project design reflects an integrated knowledge translation (iKT) strategy.

Two main outcomes relevant to knowledge users are anticipated: 1) Evidence on the usefulness of this newly developed approach to data mining as a tool for exploring multi--?factorial and multi--? pollutant relationships to environmental health outcomes, and 2) assessment of the knowledge that can be obtained by using Canada's National Pollutant Research Inventory data in studying exposures from industrial emissions and health outcomes. In addition, we will provide an initial assessment of the relative contribution of environmental, maternal, and SE factors in ABO in Canada. We will develop a resource (methodology) to assess complex interactions of coexisting determinants of health, serving as a foundation for future research that will focus on identifying causal relationships. We anticipate that participating universities will benefit from the enhanced research and teaching capacity that results from this interdisciplinary collaboration.

 

Researchers:
Prof Alvaro Osornio-Vargas (PI), Prof Osmar Zaiane (Co-PI), Dr. Laura Arbour, Dr. Khalid Aziz, Dr. Sarah Bowen, Dr. Irena Buka, Dr. Sue Chandra, Dr. Paul Demers, Dr. Perry Hystad, Dr. Manoj Kumar, Dr. Eleanor Setton, Dr. Prakeshkumar Shah, Dr. Paul Villeneuve, Prof Yutaka Yasui, Dr.Yan Yuan
Knowledge Users:
Dr. Dave Stieb, Ms. Nancy Aelicks, Ms. Erica Phipps
Students:
Emily Chan, Anders Erickson, Jundong Li, Charlene Nielsen, Jesus Serrano, and Osnat Wine
Research Coordinator:
Osnat Wine