A crash course on modeling infection inequities
This talk is an effort to distill some of the challenges associated with making infectious disease models more useful for accounting for infection inequity. The goal is to briefly highlight why I think we didn’t quite get it right during the COVID-19 pandemic, discuss important progress towards these goals in the wake of the pandemic, and highlight remaining challenges.
A brief (and incomplete!) annotated bibliography of resources on modeling acute infection inequities
Structural Racism and Infection
(Adkins-Jackson et al. 2021) provides a very useful guide for thinking about how epidemiological models and modelers should account for the impacts of racism on disease outcomes. This piece has specific recommendations for analysts around model construction and accounting for multiple, intersecting identities.
(Noppert, Hegde, and Kubale 2022) articulate a framework based on fundamental cause theory for accounting for and addressing inequities in infection risk. (Zelner, Naraharisetti, and Zelner 2023) is an invited commentary that builds on some of the ideas in the original piece.
(Chowkwanyun and Reed 2020) is an excellent piece from early in the COVID-19 pandemic which highlights many of the challenges and pitfalls associated with a focus on racial health disparities without adequate attention to their material causes.
Modeling for Equity
(Zelner et al. 2022) and (Abuelezam et al. 2023) attempt to articulate frameworks for doing a better job of using epidemic models to capture the dynamics of inequity in infection and disease outcomes.
(Ma et al. 2021) is an excellent data-driven exercise in modeling the linkages between race/ethnicity, essential work and inequity in disease outcomes.
(Nande et al. 2021) is a data-informed simulation model of the potential impact of ending the CDC eviction moratorium on population-wide infection risks and group-specific disparities.
This is a hands-on example we put together that translates the framework in (Acevedo-Garcia 2000) into a simple dynamic model of the relationship between residential segregation and infection.
Methodological challenges
(Zelner et al. 2020) is an analysis examining the relative impacts of case-fatality vs. incidence of infection in inequities in COVID-19 mortality in Michigan.
(Trangucci, Chen, and Zelner 2022) presents a computationally efficient Bayesian method for accounting for non-random missingness of race-ethnicity data without making restrictive assumptions and/or using surname coding.
This is a hands-on comparison of different metrics of spatial segregation applied to the same simulated dataset.
(If you’d like a bibtex file w/these resources, you can find it here)
References
Citation
@online{zelner2023,
author = {Zelner, Jon},
title = {A Crash Course on Modeling Infection Inequities},
date = {2023-04-07},
url = {https://zelnotes.io/posts/scenario-modeling-hub-talk},
langid = {en}
}