In many countries where enhanced public health measures against COVID-19 have been implemented, a surge of COVID-19 cases associated with the delta variant was reported in early August 2021 [17]. Our study provides evidence of this surge by demonstrating the transmission dynamics of the delta variant in South Korea, where strict social distancing measures and active case finding for eliminating COVID-19 have been implemented [18].
In our study, all clustered outbreaks were associated with indoor facilities, where activities in close proximity had taken place. Our result is consistent with the finding of other studies suggesting that the delta variant is highly transmissible in indoor facilities [19, 20]. Our mean estimate of the serial interval of the delta variant of 3.3 days was shorter than that in the mean serial interval of the wild type from a previous Korean study (4.0 days) [21, 22]. A Chinese study also reported a shorter serial interval of the delta variant than that of the wild type [23]. We also identified that 12% of the transmission occurred during the asymptomatic period of the infector, which warrants the current Korean strategy of test-trace-isolation of asymptomatic contacts with the patient.
The overall estimated R0 of 0.99 in the study was larger than that in a previous Korean study (R0 of 0.7–0.8) during the COVID-19 pandemic in 2020 [14]. Previous studies demonstrated that a larger R0, which indicates higher transmissibility, results from higher viral replication [3] and a longer duration of viral shedding in the human respiratory tract [24]. We also identified that the expected proportion of cases responsible for 80% of secondary transmission in our study (15%) was smaller than that reported in a previous study in which 26% of the cases were of the wild type infection, which suggests the increased transmission potential of the delta variant [14].
The household secondary attacks rate (63%) in our study was larger than that in previous studies conducted in Korea in 2020 (19%) [8] and pooled estimates of 17% for the wild type infection [25]. However, our finding is similar to the finding (53%) from a study of the delta variant in the United States [19]. As the household transmission of SARS-CoV-2 is one of the main drivers of the pandemic, a high household secondary attack rate is likely to contribute to the rapid surge in cases during the fourth community epidemic wave in South Korea.
Previous studies demonstrated a positive correlation between lower Ct value (higher nucleic acid level) and greater ability to culture SARS-CoV-2, which in turn, correlated with higher infectivity [26, 27]. Based on the viral shedding pattern, we identified that the delta variant has transmission potential before symptom onset, similar to the finding of a previous study of the wild type infection [28].
We observed no significant difference in the viral load across the different groups based on sex and age, similar to previous studies of the wild type infection [28, 29]. In our study, about half of the patients were under 19 years old. An increased proportion of COVID-19 cases among children was also observed in the United States [30]. This change in age distribution is likely due to the increased immunity among older age groups and increased social interactions as a result of reopening of schools.
Because the risk-benefit ratio for children remains unclear, unimmunized children are likely to be a potential transmission driver of SARS-CoV-2 in the community. Therefore, strict adherence to public health measures and personal preventive measures including universal use of masks in schools by children is needed to reduce the SARS-CoV-2 transmission risk in the population [30].
Our study has some limitations. First, the presence of some individuals with SARS-CoV-2 immunity in the community may bias our results of the superspreading potential. Overall, as of early August 2021, 17% of the South Korean population (52 million) had received two doses of the vaccination [31]; however, the vaccination program has not yet been extended to the public below 55 years of age [32], which comprised 98% of patient population in the present study. Second, some cases may have been incorrectly attributed to the clusters while the true source of infection was elsewhere. Third, in our study, the mean reporting delay of 1.7 days, the period between symptom onset and case confirmation, was shorter than the previous estimate (mean of 3.3 days) [33]. The reduced reporting delay with rapid case isolation could shorten the serial interval by truncating the infectious period of an infector [34]. Fourth, we did not account for potential confounding factors such as comorbidities of the patients, which could modify the viral shedding dynamics [35]. Lastly, we were unable to examine the potential deterministic factor for the household secondary attack rate including the characteristics of the index case, households, and household contacts.
In conclusion, the asymptomatic transmission and superspreading potential and the higher transmissibility of the delta variant likely contribute to shorter serial interval and increased secondary attack rate. Strict adherence to personal preventive measures during indoor activity and rapid case finding under the test-trace-isolation strategy remain essential to reduce the spread of the delta variant in the community.