This study aims at formulating more effective strategies for vaccine trials in infectious disease emergencies.
DOI: https://doi.org/10.6084/m9.figshare.11777832
There are many components in this project. One of them is to construct transmission dynamic models to simulate a disease spread: While most of the transmission model is only applable to one type of pathogen in one setting, this transmission model is written with an intend that it is versatile enough so that it can be applied to three types of pathogens: a) A pathogen that spreads sustainably between people in the community (e.g., influenza); b) A pathogen that spreads between people largely in hospitals or classrooms (e.g., SARS-CoV); and c) A pathogen that spreads in animal populations, but that can spillover and spread between humans unsustainably or otherwise (e.g., COVID-19, avian influenza, monkeypox).
This transmission model is to simulate a disease transmission that can be described in two parts: 1) a stochastic susceptible-exposed-infectious-removed-vaccinated (SEIRV) model that describes the epidemiological dynamics of the disease within a population; and 2) a network model that describes the spatial dynamics. The network model has three levels of model structure of clusters of varying sizes: i) small clusters, which may represent households or hospital wards; ii) communities of small clusters; and iii) a region of these communities. The rest of the document will assume these small clusters represent households, and each node within each household represent an individual.
A potential infection occurs within the population if and only if there is a contact edge between two nodes, and there may be a contact edge between two nodes whether or not the nodes are from the same household or community. Rather, whether or not there is a contact edge between two nodes are determined by stochastic block network model (SBM), in which the probability of having contacts between individuals within the same cluster may be different from those that are between clusters. Subsequently, if there is a contact edge between two nodes, and one of them is an infectious (I) individual while the other is a susceptible (S), another stochastic block network model will determine whether there is a transmission edge between these two nodes.
Infections are also introduced into the population randomly, that is, case importations occur to random clusters at different time points, and the disease importation rate, m, is defined as the number of cases per year arising completely independently from the population being studied. The per-timestep probability of infection for an individual in the population is proportional to the weighted sum of infectious cases in each cluster. Transmission through import case does not require a contact edge between two nodes.
Assuming infection status (yes/ no) as the endpoint, in the individual randomised controlled trials (iRCTs), the entire population was recruited; whereas in clustered randomized controlled trials (cRCTs), it was the population within a cluster(s), whether it is a district, a hospital, or a school, where there have had infectious case(s) was recruited.
During infectious disease emergency (PHE), vaccine or drug trial can be difficult to be implemented because of the fear induced from the PHE and limited resources available from tight budget. Adaptive designs, which use data until the time to review the design during an ongoing trial and ‘adapt’ accordingly in accordance with pre-specified rules, may be able to shorten the duration of the trial with fewer resources.
Results indicate that an iRCT is similar to cRCT at lowering the number of infected cases with having adaptive designs being a better design. It is suscepted that is because individuals are connected to others in other districts, so that vaccinating people in one district does not cut the transmission chain in other districts. It is also noted that not everyone is recruited to participate into the trial. Hence, the transmission chain may not be cut off by vaccination in the district where the trial is carried out.
Accurately assessing the ability for a pathogen to spread from one person to another and the average time between these transmissions of each infector-infectee pair is a public health priority because the information guides control strategies. Based on the transmission of pandemic and seasonal influenza in 2009-2013 in Hong Kong, this project studies the epidemiological characteristics of the diseases, assess the risks of transmission and infection, as well as proposes and evaluates methods to improve these assessments.
When comparing microneutralization (MN) and hemagglutination inhibition (HI) assays for epidemiological analysis of longitudinal serological studies of influenza, results suggest that the MN assay should be considered as a more specific endpoint for epidemiological studies of influenza transmission based on paired data, even when a lower-resource HI assay is available.
When comparing cross-sectional and longitudual designs to estimate cumulative incidence of influenza infection, results suggest that simulated longitudual studies had lower biases when background immunity was high or level of antibody boosting was low.
Also, the basic reproductive number and mean serial interval can be estimated simultaneously in real time during an outbreak of an emerging pathogen in a small, well-observed population.
The COVID-19 pandemic has been widely benefited by mathematical modelling. These models range from estimating incidence, reproductive number, serial interval, excessive mortality and mobility. They also deliberate in-hospital stay, travel patterns, effectiveness of pharmaceutical and non-pharmaceutical interventions (NPI), vaccine allocation, and impact of co-infections with other infectious diseases. The use of modeling in infectious disease dates back to the work of Daniel Bernoulli in 1760 and 1766 on smallpox. This article discusses the history of the use of mathematical modelling.
Copyright © 2018-2021 Kendra M. Wu. All rights reserved. Last Update: 15 January 2021