Study population and setting
This retrospective population-based serological analysis leveraged serological data from 4,534 individuals older than five from 2,267 households in Geneva, Switzerland between April and June 2020. Serology-based analyses were conducted to confirm previous SARS-CoV-2 infections of household members, which were combined with demographic data, reported contacts, reported symptoms, and behavior to estimate the risk of SARS-CoV-2 infection from household members and the risk of infection from extra-household exposures.
Summary of Main Findings
Overall, 6.6% of individuals included within the study has evidence of prior SARS-CoV-2 infection, of whom 70.6% were symptomatic. While 9.8% of households had at least one seropositive household member, this proportion increased with household size (4.8% of single-person households to 17% of three-person households). Overall, the probability of being infected from a single infected household member was 17.2 % (95% Credible Interval [CrI]: 13.3%-21.5%), with the risk of infection increasing with age (Aged 5-9 years: 7.5% (95% CrI: 1.3%-20.3%), Aged >65 years: 30.2% (95% CrI: 14.3%-48.2%)). Those without any defined symptoms (cough, fever, shortness of breath, loss of smell/taste) had 0.25 times the odds (95% CrI: 0.10-0.56) of infecting other household members, reflecting an estimated 19.6% (95% CrI: 12.9%-24.5%) of all within-household infections, compared to symptomatic individuals. The cumulative risk of infection from extra-household sources was 5.1% (95% CrI: 4.5%-5.8%), with men being more likely to be infected outside the household (Odds Ratio 1.4, 95% CrI: 1.1-1.8). While those 20-49 years old had the highest risk (7.4%, 95% CrI: 5.9%-9.0%) of extra-household infection, those who reduced extra-household contacts in this age-group had a reduced risk of extra-household infection (OR 0.66 , 95% CrI: 0.39-1.2). The authors estimate that 18.8% (95% CrI: 16.7%-20.4%) of all infections were due to household transmission, with the attributable proportion of intra-household infection increasing with household size (16.1% of two-person households (95% CrI: 13.7%-20.4%) to 41.2% (95% CrI: 35.3%-46.5%) of five-person households).
This study builds off a well-described population-based survey study, SEROCoV-POP, which is well suited to explore household-associated transmission risks. In addition to using appropriate transmission models , all code and related notes are provided.
As noted by the authors, while symptomatic infection was defined using multiple COVID-19-assocated symptoms, additional known symptoms such as muscle aches or pain, chills, tiredness, and gastro-intestinal symptoms were not used to identify symptomatic infections and thus individuals with symptomatic disease could have been misclassified as asymptomatic. It is unclear how the final analytical sample was selected (2,627 households had available data and 2,267 households were included), which potentially introduces selection biases if the included households had more (or less) household transmission compared to excluded households. It is also unclear how extra-household transmission results from younger individuals were obtained (extra-household contact data was not collected for those <14, yet extra-household transmission results for these children are provided). The authors point out Geneva is an urban high-income area with a small average household size (37.9% of households with available data were single-person households), which affects the ability to generalize these results outside high-income urban settings. Additionally, as the initial surge of the SARS-CoV-2 pandemic in Geneva may not reflect the burden of disease elsewhere, extrapolating any estimated extra-household infection risks to other regions should be made with caution.
This study provides important findings related to the risks of household and extra-household SARS-CoV-2 infection, and confirms the important role asymptomatic infection plays in transmission. In addition to reaffirming the importance of age and sex in the risk of infection, this study quantified the risk attributable to asymptomatic infections within the household. Additional study in non-wealthy or urban areas with small household sizes, and in cohorts with a greater number of young children, is needed.
This review was posted on: 11 January 2021