The impact of household physical distancing and its timing on the transmission of SARS-CoV-2: Insights from a household transmission evaluation study

Abstract

Background Studies on SARS-CoV-2 household transmission often assume random mixing, overlooking detailed contact patterns and the timing of physical distancing. Methods To address this, we examined interactions within 280 households, including 280 index cases and 544 members, enrolled from April 2020 to April 2021 in Nashville, Tennessee, and central Wisconsin. Eligible households were enrolled within 7 days of index case symptom onset if at least one member was initially asymptomatic. Participants were monitored for 14 days, with symptoms and respiratory specimens collected daily, and contact data retrospectively assessed at three time points: the day before index case symptom onset, the day before enrollment, and 14 days post-enrollment. We fitted Exponential Random Graph Models to the contact pattern to identify drivers of household contact. We used the fitted household models to inform a two-level mixing model to account for community infection risk, and we calibrated it to the infection data. We then used the calibrated model to study different implementation of physical distancing. Results Contact patterns showed a significant reduction in physical interactions after infection awareness, particularly avoidance of index cases, with a 77% reduction in contact density (95% CI [65%-84%], p<0.001). Simulations from the two-level mixing model indicated that initiating contact reductions at symptom onset could lower secondary infections by over 25% in households of 4-5 members. Conclusions These results demonstrate how behavior changes following infection awareness reduce transmission. Implementing physical distancing earlier, at symptom onset, could further limit secondary infections and enhance household transmission control.