Characterizing superspreading events and age-specific infectiousness of SARS-CoV-2 transmission in Georgia, USA

Abstract

It is imperative to advance our understanding of heterogeneities in the transmission of SARS-CoV-2 such as age-specific infectiousness and superspreading. To this end, it is important to exploit multiple data streams that are becoming abundantly available during the pandemic. In this paper, we formulate an individual-level spatiotemporal mechanistic framework to integrate individual surveillance data with geolocation data and aggregate mobility data, enabling a more granular understanding of the transmission dynamics of SARS-CoV-2. We analyze reported cases, between March and early May 2020, in five (urban and rural) counties in the state of Georgia. First, our results show that the reproductive number reduced to below one in about 2 wk after the shelter-in-place order. Superspreading appears to be widespread across space and time, and it may have a particularly important role in driving the outbreak in rural areas and an increasing importance toward later stages of outbreaks in both urban and rural settings. Overall, about 2% of cases were directly responsible for 20% of all infections. We estimate that the infected nonelderly cases (<60 y) may be 2.78 [2.10, 4.22] times more infectious than the elderly, and the former tend to be the main driver of superspreading. Our results improve our understanding of the natural history and transmission dynamics of SARS-CoV-2. More importantly, we reveal the roles of age-specific infectiousness and characterize systematic variations and associated risk factors of superspreading. These have important implications for the planning of relaxing social distancing and, more generally, designing optimal control measures.

Awareness-driven behavior changes can shift the shape of epidemics away from peaks and toward plateaus, shoulders, and oscillations

“…First, epidemic outbreaks include stochasticity of multiple kinds. Fluctuations could arise endogenously via process noise (especially at low levels of disease) or exogenously via time-varying parameters, Moreover, given the evidence for clustered transmission and superspreading events ( 18 – 21 ), extensions of the present model framework should explicitly account for awareness-driven behavior associated with risky gatherings ( 22 , 23 ). Next, the link between severity and behavior change depends on reporting of disease outcomes.…”

A Cross-Regional Analysis of the COVID-19 Spread during the 2020 Italian Vacation Period: Results from Three Computational Models Are Compared

“…In this case, we judged the use of a negative binomial distribution as a desirable choice, owing to the necessity of counting data [ 15 , 23 , 24 ]. A possible alternative would be the employment of a Poisson distribution, which assumes equal values of mean and variance, making it unusable in the context of COVID-19 due to the well-recognized characteristics of over-dispersion manifested by the spread of this kind of disease, often referred to as a super spreading scheme [ 25 , 26 ]. In conclusion, the opportunity of using a negative binomial distribution is well recognized in the specialized epidemiological literature in the presence of viral spread phenomena, where the distribution of individual infectiousness is often highly skewed much like, for example, in the previous cases of SARS, MERS, and Ebola, in addition to COVID-19 itself.…”

Quantifying superspreading for COVID-19 using Poisson mixture distributions

“…Based on detailed contact tracing data from Hunan, China, Sun et al [16] found that 15% of cases were responsible for 80% of transmission, and a negative binomial dispersion parameter k of 0.3. Lau et al [17] found that superspreading was widespread across space and time, with an increasing presence towards later stages of the investigated outbreaks, highlighting the importance of maintaining social distance measures. They also found that about 2% of the most infectious cases were directly responsible for 20% of all infections.…”