An Open-Data Analysis of Heterogeneities in Lung Cancer Premature Mortality Rate and Associated Factors among Toronto Neighborhoods

TL;DR

This study analyzes open public health data to identify key environmental and non-environmental factors influencing lung cancer mortality across Toronto neighborhoods, revealing heterogeneity and potential policy targets.

Contribution

Developed a flexible nonlinear analysis framework to identify and explain factors associated with lung cancer mortality in heterogeneous urban neighborhoods.

Findings

Green space and industrial pollution are linked to lung cancer mortality.

Neighborhood heterogeneity affects the influence of certain risk factors.

Additional dwelling-related factors are significant in the most severe neighborhoods.

Abstract

In public health, various data are rigorously collected and published with open access. These data reflect the environmental and non-environmental characteristics of heterogeneous neighborhoods in cities. In the present study, we aimed to study the relations between these data and disease risks in heterogeneous neighborhoods. A flexible framework was developed to determine the key factors correlated with diseases and find the most relevant combination of factors to explain observations of diseases through nonlinear analyses. Taking Lung Cancer Premature Mortality Rate (LCPMR) in Toronto as an example, two environmental factors (green space, and industrial pollution) and two non-environmental factors (immigrants, and mental health visits) were identified in the relational analysis of all of the target neighborhoods. To determine the influence of the heterogeneity of the neighborhoods, they were clustered into three different classes. In the most severe class, two additional factors related to dwellings were determined to be involved, which increased the observation's deviance from 48.1% to 80%. The factors determined in this study may assist governments in improving public health policies.