Age-structured Stochastic SEIR Modelling Covid-19 Pandemic in Kenya - Shiny Contest Submission

Age-structured Stochastic SEIR Modelling Covid-19 Pandemic in Kenya

Authors: Lucy Njoki Njuki

Abstract: The app describes the Covid-19 pandemic in Kenya: the trend of reported daily confirmed cases, deaths, recovered. Additionally, the app shows plots from simulated death data using an age-structured stochastic SEIR model using the squire package.

Full Description: The age-structured stochastic model using the squire package allows for the model to explicitly pass through the health care setting and different stages of the severity of the disease. The model also allows us to calibrate it with different epidemic start dates based on available death data.
Such interventions included reducing the contact matrix by 20% and the assumed number of bed in ICU and hospitals were 38 and 1893 respectively. Additionally, the change date for the basic reproduction number in the model was June 6, 2020 - dusk-to-dawn extension.
Above all, the app achieved to simulate Covid-19 deaths among other parameters in two scenarios(with and without interventions). However, the model did poorly in simulating the data because the specific must model parameters may not be the case in Kenya. Therefore, further investigations are needed.

Keywords: Covid-19, SEIR, Age-structured model, Squire, tidycovid19, tidyverse, future, lubridate, viridis, ggdark, plotly, shinyWidgets
Shiny app:
Repo: GitHub - LucyNjoki/SEIR_Model_Covid-19_KE-Squire
RStudio Cloud: RStudio Cloud


Full image: