SimAEN -- Simulated Automated Exposure Notification

TL;DR

This paper investigates the effectiveness of automated contact tracing using simulated disease spread models to inform public health decisions, balancing intervention benefits and costs.

Contribution

It introduces a simulation framework to evaluate automated exposure notification strategies and identifies critical parameters influencing their effectiveness.

Findings

Automated contact tracing can significantly reduce infection spread under certain conditions.

Model parameters such as testing delay and contact logging accuracy are critical for intervention success.

The study provides insights into optimal intervention strategies balancing efficacy and resource constraints.

Abstract

Mitigation strategies that remove infectious individuals from the greater population have to balance their efficacy with the economic effects associated with quarantine and have to contend with the limited resources available to the public health authorities. Prior strategies have relied on testing and contact tracing to find individuals before they become infectious and in order to limit their interactions with others until after their infectious period has passed. Manual contact tracing is a public health intervention where individuals testing positive are interviewed to identify other members of the community who they may have come into contact with. These interviews can take a significant amount of time that has to be tallied in the overall accounting of the outbreak cost. The concept of contact tracing has been expanded recently into Automated Exposure Notification whereby cellphones can be used as sensor platforms to log close contacts and notify the owner in the event that one of their close contacts tests positive. The intention is that this notification will prompt the person to be tested and then restrict their interactions with others until their status is determined. In this paper we describe our efforts to investigate the effectiveness of contact tracing interventions on controlling an outbreak. This is accomplished by creating a model of disease spread and then observing the impact that simulated tracing and testing have on the number of infected individuals. Model parameters are explored to identify critical transition points where interventions become effective. We estimate the benefits as well as costs in order to offer insight to public health officials as they select courses of action.