Assessing Transmissibility and Associated Risk Modifiers of Emerging Infectious Diseases from Contact Tracing Data

• Activity Number: 532 - Statistical Innovations Driven by the COVID-19 Pandemic
• Type: Invited
• Date/Time: Thursday, August 11, 2022 : 10:30 AM to 12:20 PM
• Sponsor: Committee on Applied Statisticians
• Author(s): Yang Yang* and Mingjin Liu and Neda Jalali
• Companies: University of Florida and University of Florida and University of Florida
• Keywords: Contact-tracing; infectious diseases; surveillance bias; transmissibility

Abstract

Contact-tracing data provide crucial information for understanding transmissibility, risk drivers and intervention efficacies for newly emerging infectious diseases. A few challenges have been long existing in statistical analysis of such data, and COVID-19 raised even more. We examine several of these challenges: (1) ascertainment bias, which includes both biased diagnosis towards symptomatic infections and imperfect diagnostic tools; (2) presymptomatic infectivity, i.e., the latent period is shorter than the incubation period; and (3) reporting bias, where only cases or transmission networks are reported but uninfected close contacts remain unknown. These issues, if left unaddressed, can lead to erroneous estimation of key epidemiological parameters including but to limited to distributions of natural history of disease (e.g., generation interval), transmission risk (e.g., secondary attack rate), and effects of risk modifiers (e.g., age and vaccination). I will discuss several strategies we have been developing to counter surveillance biases, such as sampling of asymptomatic infections and conditioning transmission likelihood on final size of case clusters.