In this discussion, we will draw on the field of HIV modelling, whose early history was fraught with stigmatizing language, but which has advanced to where best practice rests now: modelling in partnership with people living with and/or most affected by HIV.

We will discuss examples of common analytic biases when interpreting data in the context of communities disproportionately affected by HIV and by COVID-19, and the role of simplifying assumptions and misinterpretation of epidemic models, with examples from modelling and data related to HIV and of COVID-19. We will focus on terminology and interpretation, review best practices in de-stigmatizing language in health and the work of experts in social inequalities in health; and highlight discussions our team has had with communities about our analytic or modelling results related to HIV and COVID-19.

In doing so, we will challenge and ask ourselves: did our models directly or indirectly support an inequitable public health response to COVID-19? We will conclude with some considerations for potential ways forward.

Sharmistha Mishra is an Infectious disease epidemiologist and physician at University of Toronto and St. Michael’s Hospital. Her lab’s work is centered around the causes and consequences of heterogeneity in onward transmission risks, with a focus on HIV and other sexually transmitted infections. Her lab’s work on COVID-19 has focused on social inequalities and networks as mechanistic determinants of transmission risks, and how mechanistic and data-driven clarity around local transmission dynamics could be harnessed for a more specific public health response. She holds a Tier 2 Canada Research Chair in Mathematical Modelling and Program Science.