Maximizing the Efficiency of Active Case Finding for SARS-CoV-2 Using Bandit Algorithms
Published: June 2021
Researchers identify an efficient way to identify high-impact testing areas for COVID-19, operating under the assumption that vaccination rates will continue to fall short of reaching national herd immunity and detecting asymptomatic cases will continue to be a priority.
Exploring the classic decision science dilemma of explore vs. exploit, researchers identify bandit algorithms, which are used in various scenarios where sequential decisions must be made with partial information, as an ideal candidate for assisting policymakers and healthcare workers in detecting high-impact testing locations. They created a web app that can be easily used to achieve these goals, and note that their prototype made for Columbus, Ohio, is generalizable across the country.
Abstract and Citation
Gonsalves GS, Copple JT, Paltiel AD, et al. Maximizing the Efficiency of Active Case Finding for SARS-CoV-2 Using Bandit Algorithms. Medical Decision Making. 2021;41(8):970-977. doi:10.1177/0272989X211021603