What was the timing and geographical pattern of social and structural inequities in COVID-19 cases in Toronto, Canada? Patterns revealed using Gini coefficients and Lorenz curves

This paper has been published online at Annals of Epidemiology.

What did we do?

We quantified the disproportionate risk (i.e., heterogeneity) of SARS-CoV-2 cases in Toronto, Ontario (population 2.7 million) using surveillance data from January-November 2020. We examined neighbourhood characteristics including socio-demographic (income, % visible minority, % recent immigration); dwelling-related (% suitable housing, % multi-generational households); and occupation-related (% working in essential services).

We generated Lorenz curves by neighbourhoods to visualize inequalities in the distribution of SARS-CoV-2 cases. Lorenz curves show the relationship between cumulative proportion of cases and the cumulative proportion of the population. We then quantified the magnitude of disproportionate risk using the Gini coefficient. A coefficient of zero represents complete equality and one represents complete inequality.

What did we find?

  1. There was considerable overlap in the neighbourhood characteristics. That is, neighbourhoods with the highest proportion of essential workers tended to be the neighbourhoods with lowest household income and highest proportion of people living in multi-generational households.

2. Travel-related cases early in the epidemic were concentrated in higher income neighbourhoods, with a pattern of early spillover of local transmission within higher income neighbourhoods until a rapid transition to cases concentrating in lower income neighbourhoods by early April 2020. Epidemic curves by each social determinant revealed that the epidemic in higher income neighbourhoods versus lower income neighbourhoods were almost as different as epidemic curves between provinces or countries.

3. We found that 53.7% (95% CI: 53.2 – 54.3%) of cumulative cases were diagnosed in 25% of the population, with the largest concentration of cases in the north-west part of the city (a region with the largest proportion of essential workers for example).

4. The magnitude of the Gini coefficient and inequality visualized with the Lorenz curved varied across the neighbourhood characteristics, but with a consistent overall pattern of flipping from one side of the line of equality to the other (for example, representing the cross-over from higher income to lower income neighbourhoods) after early April 2020. The largest inequality measure was by % essential workers.

5. Throughout each period of shelter-in-place mandates, SARS-CoV-2 cases remained concentrated in the same neighbourhoods where they concentrated shortly after travel-related importation.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s