Speaker
Description
Historically epidemiological contact-tracing data has often reported much higher rates of transmission between household contacts compared to non-household contacts. SARS-CoV-2 has a proof-reading mechanism resulting in a relatively slow mutation rate (e.g. compared to influenza). In high transmission periods we expect true transmission pairs to have, at most, 1 single-nucleotide polymorphism (SNP) difference between them. Reported transmission pairs with >1 SNP difference are considered discordant transmission pairs, where transmission likely occurred from an alternate source.
We use NHS test and trace (CTAS, >3,000,000 cases/contacts) data, linked to UKHSA patient metadata and COG-UK viral sequence data, to identify reported transmission pairs and “true” transmission pairs (<=1 SNP difference) during the SARS-CoV-2 pandemic outbreak in the UK. CTAS data range from 01/09/2020-22/02/22, providing a wide period for observing changes in discordance. We use a strict infection period window (7 days following identification of a case) to identify potential contacts which could be infected by a primary case, as well as gaps in transmission chains. We develop multivariate models to compare concordance variance for transmission pairs adjusting for multiple covariates of interest, including age-groups, household size, lineage (e.g. BA.1, BA.2), non-pharmaceutical interventions, and vaccination.
We estimated relatively low rates of concordance between reported transmission pairs from households (e.g. <60% for BA.1 transmission pairs during non-lockdown period) suggesting non-household transmission may be higher than estimated from traditional contact-tracing or epidemiological estimates. However, adjusting for age, vaccination status, lineage, non-pharmaceutical interventions, and household size heavily impact these estimates. In particular, lower age households have significantly lower concordance in household transmission compared to older age households.
Findings provide invaluable insight into the “true” level of concordance between transmission pairs over time. Although contact-tracing data comes with self-reporting limitations, particularly for non-household contacts, our primary interest is in concordance of household contacts. These results have implications for future outbreaks, where traditional methods estimated household transmission to be much higher than non-household transmission.
