"Can we estimate epidemiological parameters of respiratory viruses from prospective household studies ? - Impact of study design"Chervet, SophieProspective longitudinal studies in households are a great opportunity to investigate the transmission of respiratory viruses. During the SARS-CoV-2 pandemic, household prospective studies have been set up worldwide to better understand the dynamics of transmission in households and quantify its risk. Here, we use a simulation study to explore how study design (e.g., recruitment of households or sample size) can impact estimates and generate potential interpretation biases. A stochastic agent-based model was developed to formalize transmission of a respiratory virus in households including members of different ages. While accounting for heterogeneities in susceptibility or infectiousness across age and symptoms, the model uses symptoms onset dates and positive tests dates to estimate the person-to-person risk of transmission within the household. Based on a review of household studies published between 2020 and 2024, a list of recruitment designs was established and modelled. Household outbreaks were simulated using the model and databases were formed by selecting households following the different designs. In each scenario, the model parameters were estimated by Bayesian inference using Markov Chain Monte Carlo sampling with data augmentation. By comparing the estimates of the model parameters to the theoretical values used in the simulation, we quantified the bias introduced by specific study designs or sample sizes. We show that design has a significant impact on the estimates of characteristics of respiratory viruses transmission in households and we propose new insights for the implementation of future household studies to minimize these biases. |
« back