Spatial Mapping Health Equity in Toronto

In this project, we delve into the spatial analysis of COVID-19 data from Toronto, examining the relationships between the virus’s impact, household income, and ethnicity. Utilizing R—a powerful language for statistical computing and graphics—we will process and visualize this vital data to uncover patterns that may inform public health decisions.

You’ll quickly realize how useful spatial mapping is for public health decision making. Be sure to check out this other project if you want to learn more.

Access the Source Code

For a deeper dive into the methodologies used in this analysis, all the relevant scripts and code are available on my GitHub repository. You can access the repository here to explore and download the scripts used in this project.

Setting Up the Workspace

Initially, we ensure that all necessary data and R libraries are prepared for analysis. We employ libraries like tidyverse for data manipulation and ggplot2 for advanced plotting. Additional libraries such as sf and raster support spatial data operations, crucial for mapping the spread of COVID-19.

library(data.table)
library(tidyverse)
library(readr)
library(sf)
library(raster)
library(rgdal)
library(rgeos)
library(ggspatial)
library(readxl)
library(scales)

Data Acquisition and Preparation

Our analysis incorporates multiple datasets including COVID-19 case data by Forward Sortation Area (FSA), alongside demographic data from the census. Each dataset is meticulously loaded and preprocessed in R to align with our analysis objectives.

sex <- read_excel("Sex by FSA.csv.xlsx") # census data
income <- read_excel("CRA Income.csv.xlsx") # census data
tocases <- read_csv("COVID19 cases.csv")

We prepare each dataset for analysis by selecting relevant columns and filtering out unwanted data. For example, we focus on Toronto by selecting records where the first letter of the geo name is ‘M’, which corresponds to Toronto FSAs.

sex <- sex[!(sex$GEO_NAME == "Canada"),]
sex$first_letterfsa <- substr(sex$GEO_NAME, 1, 1)
sex <- sex[which(sex$first_letterfsa == "M"),]

Visualizing the Data

Our visualizations focus on depicting the spread of COVID-19 cases across Toronto’s FSAs, correlating it with demographic and economic variables.

COVID-19 Cases by FSA

Here, we create a choropleth map to visualize total COVID-19 cases by FSA, employing a color scale that enhances clarity and visual impact.

ncases <- merge(case_info[,c(1,4)], fsaboundary, by.x="FSA", by.y="CFSAUID", all=TRUE)
ncases <- st_as_sf(ncases)
ggplot(ncases) + geom_sf(aes(fill=total_cases)) + theme_minimal()

Average Income and Demographics by FSA: Further analyses include mapping average income and demographic data to observe potential correlations with COVID-19 case rates.

Below are the resulting plots, with some post-R circles added on for easier comparisons.

I’ve circled the FSA’s in Toronto with the most extreme COVID rates (either very low or very high).

We can easily see how the rates correlate with the percentage of minorities are in each of the circled areas.

Similarly, it’s clear how household income correlates with COVID rates.

Insights and Conclusions

Our analysis highlights significant correlations between COVID-19 impacts and socio-economic factors such as income and ethnicity. By identifying trends and patterns, public health authorities can tailor interventions more effectively, ensuring resources are allocated where they are most needed.

Importance in Health Equity

This project underscores the importance of considering health equity in data analysis, enabling us to identify and address disparities in health impacts across different communities. Such insights are pivotal for guiding equitable public health interventions and policies.

Implications for Public Health Prioritization

The insights derived from this analysis serve as a crucial tool for public health prioritization, helping to pinpoint areas requiring urgent intervention during the ongoing pandemic. This approach not only aids in immediate response strategies but also informs long-term public health planning.