Statistical inference in a stochastic epidemic SEIR model with control intervention: Ebola as a case study.
read more
Citations
Mathematical models of infectious disease transmission
Mathematical modeling of the spread of the coronavirus disease 2019 (COVID-19) taking into account the undetected infections. The case of China.
Transmission dynamics and control of Ebola virus disease (EVD): a review
Estimating the future number of cases in the Ebola epidemic--Liberia and Sierra Leone, 2014-2015.
Estimating the reproduction number of Ebola virus (EBOV) during the 2014 outbreak in West Africa
References
Infectious Diseases of Humans: Dynamics and Control
The Mathematics of Infectious Diseases
On the definition and the computation of the basic reproduction ratio R0 in models for infectious diseases in heterogeneous populations
On the definition and the computation of the basic reproduction ratio : $R_ 0$ in models for infectious diseases in heterogeneous populations
Related Papers (5)
Frequently Asked Questions (17)
Q2. What is the common theoretical framework for the study of Ebola?
The theoretical framework most commonly used is based on the division of the human host population into categories containing susceptible, infected but not yet infectious (exposed), infectious, and recovered individuals.
Q3. What is the scope for the estimation framework developed here?
There is scope for the estimation framework developed here to be extended to other more complex scenarios such as, for example, heterogeneously mixing populations, other distributional assumptions for the transition density (see Riley et al., 2003 for a description of a model that incorporates gamma-distributed waiting times), and other types of incomplete data.
Q4. What is the basic idea of the stochastic continuous-time model?
The basic idea is that transitions of individuals from the previous to the nextstage of the disease are seen as stochastic movements between the corresponding population compartments.
Q5. What is the effect of imputation of B on the gamma priors?
Preliminary analysis showed that imputation of B introduces a negative correlation between the chains for q and and the authors find that convergence is improved if the authors sample them jointly and independently of their previous values using a bivariate normal proposal centered at the mode of the bivariate conditional distribution (see, e.g., Chib and Greenberg, 1994).
Q6. How many epidemics did Chowell and his team simulate?
The authors simulated 500 epidemics for the case where the system was intervened at time point t∗ = 130 and another 500 epidemics where the authors allowed unimpeded spread of the disease by setting β(t) = β.
Q7. How many people would have been affected by the Ebola epidemic?
The authors estimate that an unimpeded spread of the epidemic would have lasted about five times longer and would have affected two thirds of the population.
Q8. What is the simplest way to describe the binomial distributions?
The binomial distributions (5) result from summation over the individual Bernoulli trials assuming that they are independent and identical for all members of a compartment.
Q9. What is the appropriate procedure for a model with latent variables?
The appropriate procedure for a model with latent variables is to integrate over their probability distribution and MCMC provides a feasible algorithm for doing this numerically.
Q10. How can the authors assess the fit of the model?
The model fit can be assessed, for example, by using the posterior predictive model checking technique based on a discrepancy measure as suggested by Gelman, Meng, and Stern (1996).
Q11. How long did the posterior mean of the incubation period be?
The authors also estimate that the posterior density of the mean incubation period has a mean of 10 days and a posterior standard deviation of about 1 day.
Q12. What is the simplest way to describe the exponential distribution of the incubation and the infectious?
It follows that the exponential distribution of the incubation and the infectious period is approximated by the corresponding geometric distribution with means 1/pC and 1/pR, respectively.
Q13. What can be done to improve the estimation of parameters?
In such cases, methods that impute the unobserved process in between (regularly or irregularly spaced) discrete time points (Elerian, Chib, and Shephard, 2001; Durham and Gallant, 2002) can substantially improve parameter estimation.
Q14. How much higher is the estimated transmission parameter?
Their estimate of the transmission parameter β is 0.21, which is smaller than the value of 0.33 obtained by Chowell et al. (2004) while their posterior standard deviation for this parameter is about three times higher than their reported standard error.
Q15. How much of the posterior mean is in agreement with the ML estimates?
As can be expected, the posterior standard deviations for β, , and q are larger demonstrating that the algorithm has incorporated the increased level of uncertainty, as effectively about 30% of the complete data are not available.
Q16. What is the importance of a probabilistic model for estimating the parameters of unobserve?
This is important in achieving parameter estimates with standard errors that realistically reflect increased uncertainty when variables are unobserved.
Q17. What is the first step in constructing a probabilistic model for the disease to be studied?
The first step is to construct a probability model for the disease to be studied and to investigate parameter identifiability under the scenario of the available data.