Estimation of introduction and transmission rates of SARS-CoV-2 in a prospective household study

by Michiel van Boven, Christiaan H. van Dorp, Ilse Westerhof, Vincent Jaddoe, Valerie Heuvelman, Liesbeth Duijts, Elandri Fourie, Judith Sluiter-Post, Marlies A. van Houten, Paul Badoux, Sjoerd Euser, Bjorn Herpers, Dirk Eggink, Marieke de Hoog, Trisja Boom, Joanne Wildenbeest, Louis Bont, Ganna Rozhnova, Marc J. Bonten, Mirjam E. Kretzschmar, Patricia Bruijning-Verhagen

Household studies provide an efficient means to study transmission of infectious diseases, enabling estimation of susceptibility and infectivity by person-type. A main inclusion criterion in such studies is usually the presence of an infected person. This precludes estimation of the hazards of pathogen introduction into the household. Here we estimate age- and time-dependent household introduction hazards together with within household transmission rates using data from a prospective household-based study in the Netherlands. A total of 307 households containing 1, 209 persons were included from August 2020 until March 2021. Follow-up of households took place between August 2020 and August 2021 with maximal follow-up per household mostly limited to 161 days. Almost 1 out of 5 households (59/307) had evidence of an introduction of SARS-CoV-2. We estimate introduction hazards and within-household transmission rates in our study population with penalized splines and stochastic epidemic models, respectively. The estimated hazard of introduction of SARS-CoV-2 in the households was lower for children (0-12 years) than for adults (relative hazard: 0.62; 95%CrI: 0.34-1.0). Estimated introduction hazards peaked in mid October 2020, mid December 2020, and mid April 2021, preceding peaks in hospital admissions by 1-2 weeks. Best fitting transmission models included increased infectivity of children relative to adults and adolescents, such that the estimated child-to-child transmission probability (0.62; 95%CrI: 0.40-0.81) was considerably higher than the adult-to-adult transmission probability (0.12; 95%CrI: 0.057-0.19). Scenario analyses indicate that vaccination of adults can strongly reduce household infection attack rates and that adding adolescent vaccination offers limited added benefit.