Spatial survival analysis: an application to lung cancer data in Manitoba

Loading...
Thumbnail Image
Date
2023-12-13
Authors
Sobhan, Shamsia
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Survival data are often collected in cluster such as geographic regions. Incorporating the cluster effect (between cluster dependence and within cluster dispersion) in survival model not only improves the accuracy and efficiency of parameter estimation, but also investigates spatial pattern and identify high-risk areas. The commonly used spatial-survival models are mostly restricted to single-event or competing risks settings, with a few extensions of semi-competing risks setting which only incorporate between cluster variation. This thesis proposed a spatial semi-competing risk model that allows for spatial dependence while estimating the risks of terminal (e.g., death) and non-terminal (e.g., lung cancer) events. A real data application of our model on a merge dataset of Manitoba lung cancer registry and vital statistics was provided to investigate the pattern of events (lung cancer and death) in Manitoba and evaluate the effect of demographic and socio-economic factors on the risk of events. Socio-economic status score, high percentage of visible minority, and high percentage of Indigenous population were found to be risk factors of both lung cancer and death. Male population were at higher risk of both lung cancer and death in comparison to female population. The performance of our proposed model was also evaluated through simulation studies.
Description
Keywords
Cluster effect, Semi-competing risks, Spatial-survival model
Citation