Transmission of SARS-CoV-2 within households: a prospective cohort study in the Netherlands and Belgium - Interim results

Background: Household transmission studies are useful to obtain granular data on SARS-CoV-2 transmission dynamics and to gain insight into the main determinants. In this interim report we investigated secondary attack rates (SAR) by household and subject characteristics in the Netherlands and Belgium. Methods: Households with a real-time reverse transcription polymerase chain reaction (RT-PCR) confirmed SARS-CoV-2 index case were enrolled <48 hours following report of the positive test result. Daily symptom follow-up, standardized nose-throat sampling at enrollment and at new-onset acute respiratory illness (ARI) and paired dried blood spots (DBS) were collected from each participant. Children 0-2 years of age were additionally requested to collect a stool sample 7 days after enrollment and at new-onset of ARI. Swabs and stool samples were tested by RT-PCR for virus detection and DBS by multiplex protein microarray for detection of SARS-CoV-2 antibodies. The SAR was calculated 1) per-household as the proportion of households with ≥1 secondary SARS-CoV-2 case and 2) per-person as the probability of infection in household members at risk. We explored differences in SARs by household and subject characteristics. Results: This analysis includes 117 households that completed follow-up between April-December 2020. Among 382 subjects, 74 secondary infections were detected, of which 13 (17.6%) were asymptomatic and 20 (27.0%) infections were detected by seroconversion only. Of cases detected by RT-PCR, 50 (67.6%) were found at enrollment. The household SAR was 44.4% (95%-CI: 35.4-53.9%) and was higher for index cases meeting the ARI case definition (52.3%; 95%-CI 41.4-62.9%) compared to mildly symptomatic (22.2%; 95%-CI: 9.4-42.7%) and asymptomatic index cases (0.0%; 95%-CI: 0.0-80.2%). The per-person SAR was 27.9% (95%-CI: 22.7-33.8%). Transmission was lowest from child to parent (9.1%; 95%-CI: 2.4-25.5%) and highest from parent to child (28.1%; 95%-CI: 19.7-38.4%) and in children 6-12 years (34.2%; 95%-CI: 20.1-51.4%). Among 141 subjects with RT-PCR confirmed SARS-CoV-2 infections, seroconversion was detected in 111 (78.7%). Conclusion: We found a high household SAR, with the large majority of transmissions detected early after identification of the index case. Our findings confirm differential SAR by symptom status of the index. In almost a quarter of RT-PCR positive cases, no antibodies were detected. Other factors influencing transmission will be further explored as more data accumulate.