An interaction Neyman–Scott point process model for coronavirus disease-19

Abstract

With rapid transmission, the coronavirus disease 2019 (COVID-19) has led to over three million deaths worldwide, posing significant societal challenges. Understanding the spatial patterns of patient visits and detecting local cluster centers are crucial to controlling disease outbreaks. We analyze COVID-19 contact tracing data collected from Seoul, which provide a unique opportunity to understand the mechanism of patient visit occurrence. Analyzing contact tracing data is challenging because patient visits show strong clustering patterns, while cluster centers may have complex interaction behavior. Cluster centers attract each other at mid-range distances because other cluster centers are likely to appear in nearby regions. At the same time, they repel each other at too small distances to avoid merging. To account for such behaviors, we develop a novel interaction Neyman–Scott process that regards the observed patient visit events as offsprings generated from a parent cluster center. Inference for such models is challenging since the likelihood involves intractable normalizing functions. To address this issue, we embed an auxiliary variable algorithm into our Markov chain Monte Carlo. We fit our model to several simulated and real data examples under different outbreak scenarios and show that our method can describe the spatial patterns of patient visits well. We also provide useful visualizations that can inform public health interventions for infectious diseases, such as social distancing.

A spatio‐temporal Dirichlet process mixture model for coronavirus disease‐19

“…As shown in Figure 9, the number of kernels got larger and their sizes became smaller as the number of cases increased. This pattern coincides with what has been observed by Park et al 12 in their similar but simpler purely spatial model. Individual clusters became smaller as the spatial distribution of events became denser, which led to a smaller value of the range parameter for spatial kernels.…”

“…The ideas and method proposed in this paper can generally be applicable to model other clustering processes such as distribution of rat sightings 37 and dynamics of crime. 38 From a computational perspective, the use of Dirichlet process allows us to detect disease spreading centers quickly with much lower computational cost, compared to the Neyman-Scott process approaches; 12,21 fitting the Neyman-Scott process model is computationally prohibitive for complex spatio-temporal processes due to the slow mixing of birth-death MCMC. In terms of modeling and application aspects, our method provides insights on disease hot spots, which are useful for planning public health policy.…”

COVID-19 transmission risk in Surabaya and Sidoarjo: an inhomogeneous marked Poisson point process approach

“…2022 ; Park et al. 2022 ). In this context, point processes are a standard tool to analyze such data (Baddeley et al.…”

A mechanistic spatio‐temporal modeling of COVID‐19 data

“…For instance, the spatio-temporal distribution of COVID-19 cases and hospitalizations in United Kingdom during the first 6 months of the pandemic have been analyzed through kernel density estimation methods (Elson et al, 2021) and log-Gaussian Cox processes (Watson et al, 2021). Furthermore, an interaction Neyman-Scott model has been applied to a contact tracing data set collected from Seoul, South Korea, in order to capture local spreading events (Park et al, 2021), while nonstationary spatio-temporal point process models have also been used to analyze the cases recorded in Cali, Colombia (Dong et al, 2021). In other studies, Hawkes point processes have been adapted to model the self-exciting properties of a contagious disease such as COVID-19 at an areal level (Chiang et al, 2021).…”