Genomics-enhanced contact tracing enabled the characterization of SARS-CoV-2 transmission chains and their associated infection contexts in the general population during community transmission

Understanding how pathogens spread in human populations is key to effective infection prevention, but for most pathogens in community transmission, population transmission chains and their associated infection contexts remain elusive. From February to December 2021, we carried out one of the highest-intensity SARS-CoV-2 genomic surveillance programmes in Europe and implemented genomics-enhanced contact tracing for the characterization of SARS-CoV-2 community transmission. Here, we report on the first retrospective analysis of the complete dataset, integrating data 32,380 cases, 49,906 forward and backward contact tracing records, 162 recorded institutional outbreaks, and 8,028 viral genome sequences (case sequencing coverage 24.5%). 84% of putative transmission detected by forward contact tracing were associated with private households (estimated attack rate 30%), followed by “recreational context” contacts (9% of putative transmissions, 13% transmission probability). Lower transmission probabilities in the 3% – 6.5% range observed for many contexts (e.g. workplaces, schools, residential care facilities) suggested effective infection prevention in these. All institutional outbreaks combined were associated with 5% of total cases; schools and kindergartens exhibited the highest (1.3 and 1.2, respectively), and care homes the lowest (0.2), number of outbreak-linked community cases per outbreak case. A sequencing-led search showed that routine contact tracing underestimated the number of cases associated with school outbreaks (18% increase in associated sequenced cases) and it detected previously unrecognized links between different outbreaks. Combining all available epidemiological data, we identified a putative infection source for 23.7% of sequenced cases; even when ignoring genetic constraints, however, 55.6% of all registered cases remained without a putative transmission source. The available genetic data indicated that a sequencing-led search for epidemiological connections across all sequenced cases could substantially contribute to the detection of transmission events not captured by routine contact tracing (upper bound 60.5% of sequenced cases with a putative infection source, assuming that that all genetic links imply a discoverable epidemiological connection). Different indicators of the ability to characterize SARS-CoV-2 transmission showed no evidence of saturation towards the achieved case sequencing proportion of 24.5%, suggesting that higher sequencing rates would have been desirable. In conclusion, we demonstrated the comprehensive characterization of SARS-CoV-2 transmission in different contexts. Our study informs the implementation and application of genomics-enhanced contact tracing, which could become an instrument for more efficient infection prevention during future pandemics and help navigate difficult societal trade-offs.