Open Collections

Simkin, Jonathan David

University of British Columbia

Background: In British Columbia (BC), geographic variation in cancer risk was reported across large geographic areas and without spatial statistics. More local-based analyses, incorporating spatial statistics are warranted to support regional cancer prevention efforts. Overall, this thesis demonstrates the utility of geospatial approaches to better understand cancer risk in the context of place and geography. After reviewing the literature on Canadian population oncology research that applied geospatial methods (Chapter 2), this thesis demonstrates several geospatial methodologies to explore different research questions, cancer types and settings. Specifically, geospatial methods were used to highlight geographic inequities for a cancer with strong geographic variation (Ch. 3), examine rural-urban differences for a highly preventable cancer (Ch. 4), estimate small area risk (Ch. 5), and explore relationships between environmental exposures and area-level cancer incidence (Ch. 6). Methods: Incident cancers were obtained from the BC Cancer Registry (Ch. 4-6) and cancer registries across eight countries (Ch. 3). Age-standardization was used to derive age-standardized rates and incidence ratios (Ch. 3-6). Spatial data linkage provided levels of urbanization and environmental exposures (Ch 4-6). The Modified Besag, York, and Mollie model was used to estimate small area risk and examine ecologic associations (Ch. 5-6). An R-package was developed for small area risk estimation (Ch. 5). Results: Circumpolar and Indigenous populations experienced elevated stomach cancer incidence rates versus regional counterparts (Ch. 3). Cervical cancer incidence was associated with ethnicity/race and urbanization-level (Ch. 4). Lung cancer risk was estimated among 218 BC Community Health Service Areas and 2357 excess cases were estimated among areas with elevated risk between 2011 and 2018 (Ch. 5). Fine particulate air pollution was associated with an increased lung cancer risk and was modified by area greenness (Ch. 6). Conclusions: This thesis utilizes accessible geospatial approaches that make full use of cancer registry data. Area-level studies alone are not well-suited for causal inference but, combined with other methodologies, offer additional perspectives to better understand cancer risk in the context of place and geography and support localized cancer prevention efforts. These approaches are advance the adoption of geospatial methods in Canadian population oncology research and surveillance.

Text

2022-08-24

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/82518

Population and Public Health

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/