Transmission of Severe Acute Respiratory Syndrome Coronavirus 2 in Households with Children, Southwest Germany, May-August 2020

Abstract

Resolving the role of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission in households with members from different generations is crucial for containing the current pandemic. We conducted a large-scale, multicenter, cross-sectional seroepidemiologic household transmission study in southwest Germany during May 11-August 1, 2020. We included 1,625 study participants from 405 households that each had ≥1 child and 1 reverse transcription PCR-confirmed SARS-CoV-2-infected index case-patient. The overall secondary attack rate was 31.6% and was significantly higher in exposed adults (37.5%) than in children (24.6%-29.2%; p = <0.015); the rate was also significantly higher when the index case-patient was >60 years of age (72.9%; p = 0.039). Other risk factors for infectiousness of the index case-patient were SARS-CoV-2-seropositivity (odds ratio [OR] 27.8, 95% CI 8.26-93.5), fever (OR 1.93, 95% CI 1.14-3.31), and cough (OR 2.07, 95% CI 1.21-3.53). Secondary infections in household contacts generate a substantial disease burden.

Keywords: COVID-19; SARS-CoV-2; antibodies; children; coronavirus disease; households; respiratory infections; serology; severe acute respiratory syndrome coronavirus 2; transmission; viruses; zoonoses.

Figure 1

Flowchart of participant enrollment in study of transmission of severe acute respiratory syndrome coronavirus 2 in households with children, southwest Germany, May–August 2020. RT-PCR, reverse transcription PCR; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Figure 2

Observed and predicted SARs in household members exposed to severe acute respiratory syndrome coronavirus 2, southwest Germany, May–August 2020. SARs shown are associated with age of index case-patient (A), age of exposed household member (B), household size (C), and SARS-CoV-2 seropositivity of the index case-patient (D). The mean observed SAR is shown as a black dot. The mean (black triangles), interquartile range (white bars), maximum and minimum (ends of vertical black line), and distribution (gray shading) of the predicted SAR are shown in the violin plots. The predicted SARs were calculated from the generalized linear mixed-effects logistic regression model. SAR, secondary attack rate.

Figure 3

Generalized linear mixed model binary decision trees in study of transmission of severe acute respiratory syndrome coronavirus 2 in households with children, southwest Germany, May–August 2020. A) Model incorporating the 2 most dominant effects (p<0.001) on the SAR of exposed household members, SARS-CoV-2 seropositivity of the index case-patient and age of exposed household members with a seronegative or a seropositive index case-patient. B) Model incorporating only age of the index case-patient as a risk factor; SAR was modeled by age of exposed household member within each node. In both panels, the observed SAR as a proportion of seropositive (black) and seronegative (gray) exposed household members with these characteristics are shown within final nodes and as a percentage with the total number of seropositive/total exposed household members in parentheses above each node. In panel B, the predicted SARs are indicated within each final node as a red dot and red straight line. SAR, secondary attack rate.