A Hybrid Epidemic Model to Explore Stochasticity in COVID-19 Dynamics

Predicting Covid-19 pandemic waves with biologically and behaviorally informed universal differential equations

During the COVID-19 pandemic, it became clear that pandemic waves and population responses were locked in a mutual feedback loop in a classic example of a coupled behavior-disease system. We demonstrate for the first time that universal differential equation (UDE) models are able to extract this interplay from data. We develop a UDE model for COVID-19 and test its ability to make predictions of second pandemic waves. We find that UDEs are capable of learning coupled behavior-disease dynamics and predicting second waves in a variety of populations, provided they are supplied with learning biases describing simple assumptions about disease transmission and population response. Though not yet suitable for deployment as a policy-guiding tool, our results demonstrate potential benefits, drawbacks, and useful techniques when applying universal differential equations to coupled systems.

Exploring the potential of learning methods and recurrent dynamic model with vaccination: A comparative case study of COVID-19 in Austria, Brazil, and China

In order to effectively manage infectious diseases, it is crucial to understand the interplay between disease dynamics and human conduct. Various factors can impact the control of an epidemic, including social interventions, adherence to health protocols, mask-wearing, and vaccination. This article presents the development of an innovative hybrid model, known as the Combined Dynamic-Learning Model, that integrates classical recurrent dynamic models with four different learning methods. The model is composed of two approaches: The first approach introduces a traditional dynamic model that focuses on analyzing the impact of vaccination on the occurrence of an epidemic, and the second approach employs various learning methods to forecast the potential outcomes of an epidemic. Furthermore, our numerical results offer an interesting comparison between the traditional approach and modern learning techniques. Our classic dynamic model is a compartmental model that aims to analyze and forecast the diffusion of epidemics. The model we propose has a recurrent structure with piecewise constant parameters and includes compartments for susceptible, exposed, vaccinated, infected, and recovered individuals. This model can accurately mirror the dynamics of infectious diseases, which enables us to evaluate the impact of restrictive measures on the spread of diseases. We conduct a comprehensive dynamic analysis of our model. Additionally, we suggest an optimal numerical design to determine the parameters of the system. Also, we use regression tree learning, bidirectional long short-term memory, gated recurrent unit, and a combined deep learning method for training and evaluation of an epidemic. In the final section of our paper, we apply these methods to recently published data on COVID-19 in Austria, Brazil, and China from 26 February 2021 to 4 August 2021, which is when vaccination efforts began. To evaluate the numerical results, we utilized various metrics such as RMSE and R-squared. Our findings suggest that the dynamic model is ideal for long-term analysis, data fitting, and identifying parameters that impact epidemics. However, it is not as effective as the supervised learning method for making long-term forecasts. On the other hand, supervised learning techniques, compared to dynamic models, are more effective for predicting the spread of diseases, but not for analyzing the behavior of epidemics.

Modeling impact of vaccination on COVID-19 dynamics in St. Louis

The region of St. Louis, Missouri, has displayed a high level of heterogeneity in COVID-19 cases, hospitalization, and vaccination coverage. We investigate how human mobility, vaccination, and time-varying transmission rates influenced SARS-CoV-2 transmission in five counties in the St. Louis area. A COVID-19 model with a system of ordinary differential equations was developed to illustrate the dynamics with a fully vaccinated class. Using the weekly number of vaccinations, cases, and hospitalization data from five counties in the greater St. Louis area in 2021, parameter estimation for the model was completed. The transmission coefficients for each county changed four times in that year to fit the model and the changing behaviour. We predicted the changes in disease spread under scenarios with increased vaccination coverage. SafeGraph local movement data were used to connect the forces of infection across various counties.

SPARSEMODr: Rapidly simulate spatially explicit and stochastic models of COVID-19 and other infectious diseases

Building realistically complex models of infectious disease transmission that are relevant for informing public health is conceptually challenging and requires knowledge of coding architecture that can implement key modeling conventions. For example, many of the models built to understand COVID-19 dynamics have included stochasticity, transmission dynamics that change throughout the epidemic due to changes in host behavior or public health interventions, and spatial structures that account for important spatio-temporal heterogeneities. Here we introduce an R package, SPARSEMODr, that allows users to simulate disease models that are stochastic and spatially explicit, including a model for COVID-19 that was useful in the early phases of the epidemic. SPARSEMOD stands for SPAtial Resolution-SEnsitive Models of Outbreak Dynamics, and our goal is to demonstrate particular conventions for rapidly simulating the dynamics of more complex, spatial models of infectious disease. In this report, we outline the features and workflows of our software package that allow for user-customized simulations. We believe the example models provided in our package will be useful in educational settings, as the coding conventions are adaptable, and will help new modelers to better understand important assumptions that were built into sophisticated COVID-19 models.

A Hybrid Epidemic Model to Explore Stochasticity in COVID-19 Dynamics

The dynamic nature of the COVID-19 pandemic has demanded a public health response that is constantly evolving due to the novelty of the virus. Many jurisdictions in the USA, Canada, and across the world have adopted social distancing and recommended the use of face masks. Considering these measures, it is prudent to understand the contributions of subpopulations—such as “silent spreaders”—to disease transmission dynamics in order to inform public health strategies in a jurisdiction-dependent manner. Additionally, we and others have shown that demographic and environmental stochasticity in transmission rates can play an important role in shaping disease dynamics. Here, we create a model for the COVID-19 pandemic by including two classes of individuals: silent spreaders, who either never experience a symptomatic phase or remain undetected throughout their disease course; and symptomatic spreaders, who experience symptoms and are detected. We fit the model to real-time COVID-19 confirmed cases and deaths to derive the transmission rates, death rates, and other relevant parameters for multiple phases of outbreaks in British Columbia (BC), Canada. We determine the extent to which SilS contributed to BC’s early wave of disease transmission as well as the impact of public health interventions on reducing transmission from both SilS and SymS. To do this, we validate our model against an existing COVID-19 parameterized framework and then fit our model to clinical data to estimate key parameter values for different stages of BC’s disease dynamics. We then use these parameters to construct a hybrid stochastic model that leverages the strengths of both a time-nonhomogeneous discrete process and a stochastic differential equation model. By combining these previously established approaches, we explore the impact of demographic and environmental variability on disease dynamics by simulating various scenarios in which a COVID-19 outbreak is initiated. Our results demonstrate that variability in disease transmission rate impacts the probability and severity of COVID-19 outbreaks differently in high- versus low-transmission scenarios.