Hong Kong contact tracing study identifies case clusters and superspreading events

NIAID CEIRS | Research Publication Commentary

Adam DC et al. (2020) Clustering and superspreading potential of SARS-CoV-2 infections in Hong Kong. Nat Med.

A team of World Health Organization (WHO) researchers representing Hong Kong, Australia, and France recently published a study in Nature Medicine investigating SARS-CoV-2 “superspreading events” (SSEs). The study was led by Dr. Ben Cowling, from the University of Hong Kong and member of the St. Jude Center of Excellence for Influenza Research and Surveillance (SJCEIRS), and identified and traced cases in Hong Kong over a 3-month period to identify patterns of spread and links to potential SSEs. Because previous coronavirus epidemics of SARS-CoV and MERS-CoV were characterized by superspreading transmission events, the authors used contract tracing information to help determine if SSEs are a component of SARS-CoV-2 transmission as well.

The data for the study were collected between January 23rd and April 28th, 2020 in Hong Kong, with the majority of cases occurring between early March and mid-April. Over the course of the study, Hong Kong instituted progressively more restrictive control measures in response to increasing numbers of cases. Cases included in the study were separated by local or imported primary case and by the locale of transmission (social setting, at work, at home, traveling).

The authors found that that SSEs are the cause of a significant number of SARS-CoV-2 cases. While most case clusters only involved two people (one infected person passing the virus to one other), the scope of the SSEs cause a large increase in infections. When considering all the clusters together, researchers determined that 80% of locally transmitted cases could be traced back to just 19% of primary cases. From this study data collected in Hong Kong, the authors found that 69% of cases were not linked to any additional infections. This is demonstrated especially effectively when considering the R value calculated in this study. A critical epidemiological concept is the basic reproductive rate, or R, which considers the infectivity of a pathogen as well as the environment it exists in and the behavior of its hosts. An R value of 1 indicates that each person who is infected with a disease will pass it to one other person. When R is less than 1, each infected person causes less than one new infection on average. This study took place in the context of social distancing and government-mandated quarantine in Hong Kong. Overall, the R was 0.58, and when a cluster size model was implemented to account for possible bias, R was 0.74. This is notable, given that there were several SSEs identified where a few infected individuals propagated the disease to many others. The low overall R suggests that despite SSEs, mitigation measures like social distancing and quarantine of SARS-CoV-2 positive individuals may be effective in slowing spread of the disease.

This study investigated SSEs during the early stages of the pandemic before testing was more widely available. To evaluate the impact of increased testing on these events, the authors adjusted their model to decease the time between a person developing symptoms and when they received a positive test result (faster turnaround). The authors did not find a decrease in disease transmission in their computational model, consistent with claims that individuals may be infectious while asymptomatically infected. Indeed, 19% of the positive cases identified in this study were from individuals who were asymptomatic at the time of testing. This is consistent with the concepts of social spread at social events like restaurants and bars, which are more often frequented by younger individuals. The majority of cases recorded for this study were individuals in their 30s, and there was a higher rate of transmission observed between individuals of similar age. Therefore, before Hong Kong instituted restrictions on bars in early April, it is possible that people frequenting indoor social spaces were asymptomatically infected and contributed to these SSEs.

This study has important implications in policy recommendations for social gatherings and surveillance testing. While the authors studied a limited dataset of only seven SSEs, it effectively illustrates the potential risk posed by large gatherings and spread by asymptomatic individuals, as well as the effectiveness of quarantine and social distancing in containingSARS-CoV-2. Policy makers should continue to consider the data available, as well as the population and environment, to make the best management recommendations to avoid SSEs and prevent continued spread of SARS-CoV-2.