Agent-based Modeling of COVID-19 to Support Public Health Decision Making

Jonathan Ozik, Ph.D.
The University of Chicago

Anna Hotton, Ph.D., MPH, BS
The University of Chicago

ABSTRACT:
The COVID-19 pandemic has highlighted the need for detailed modeling approaches that can capture the myriad complexities of emerging infectious diseases. In response, our group has developed CityCOVID, an agent-based model capable of tracking COVID-19 transmission in large, urban areas. Through partnerships between Argonne National Laboratory, the University of Chicago, the Chicago Department of Public Health, and the Illinois COVID-19 Modeling Task Force we combined multiple data sources to develop a locally informed, realistic, and statistically representative synthetic agent population, with attributes and processes that reflect real-world social and biomedical aspects of transmission. We model all 2.7 million individual residents of Chicago, as they go to and from 1.2 million different places according to their individual hourly schedules. The places include locations such as households, workplaces, schools, and hospitals, and, as individuals congregate with other individuals in these places over the course of their daily routines, they are exposed to potential infection from other infectious people who are also at those places. Transitions between disease states depend on agent attributes and exposure to infected individuals, placed-based risks, and protective behaviors. This detailed modeling approach allows us to implement very specific and realistic mitigation strategies that are being considered by stakeholders, and which have been evolving over the course of the pandemic. We continue to apply CityCOVID to examine the impact of non-pharmaceutical interventions, SARS-CoV-2 variants of concern, vaccination deployment strategies, and to understand the impacts of social determinants of health on disease outcomes. In this presentation we will describe CityCOVID, including how the synthetic population was developed, what agent-based modeling and high-performance computing technologies were required, and our efforts in supporting local public health stakeholders in understanding, responding to and planning for the current and future population health emergencies.

Causal interaction in high dimension

Kosuke Imai, Ph.D.
Princeton University

ABSTRACT:
Estimating causal interaction effects is essential for the exploration of heterogeneous treatment effects. In the presence of multiple treatment variables with each having several levels, researchers are often interested in identifying the combinations of treatments that induce large additional causal effects beyond the sum of separate effects attributable to each treatment. We show, however, the standard definition of causal interaction effect, typically estimated with the standard linear regression or ANOVA, suffers from the lack of invariance to the choice of baseline condition and the difficulty of interpretation beyond two-way interaction. We propose an alternative definition of causal interaction effect, called the marginal treatment interaction effect, whose relative magnitude does not depend on the choice of baseline condition while maintaining an intuitive interpretation even for higher-order interaction. The proposed approach enables researchers to effectively summarize the structure of causal interaction in high-dimension by decomposing the total effect of any treatment combination into the marginal effects and the interaction effects. We also establish the identification condition and develop an estimation strategy for the proposed marginal treatment interaction effects. Our motivating example is conjoint analysis where the existing literature largely assumes the absence of causal interaction. Given a large number of interaction effects, we apply a variable selection method to identify significant causal interaction. Our exploratory analysis of a survey experiment on immigration preferences reveals substantive insights the standard conjoint analysis fails to discover.

Big data for small minds: With great promise comes great responsibility

Robert Gibbons, Ph.D.
University of Chicago

ABSTRACT:
With the promise of big data comes the responsibility of analyzing them wisely.  Statisticians have struggled with the analysis of observational data for decades and too often this work has been ignored by data scientists.  I present a series of examples of analytic work involving big data to illustrate good and bad analytic approaches.