Stochastic epidemic models can be a useful tool for understanding and managing disease progression as well as assessing different disease-control measures in public health. However, unless a sufficiently accurate epidemic model is used, such advantages are of minimal utility. It is feasible to interpret parameter estimates, compare disease outbreaks, and execute control techniques if the model gives a sufficient fit. This paper presents a new method for determining stochastic Susceptible-Exposed-Infected-Removed (SEIR) epidemic models with varying infection rates. The technique investigates how SEIR models with various infection mechanisms can be assessed and differentiated given a set of removal times. The concept is built on employing a posterior predictive model checking technique with notion of predictive residuals to evaluate the discrepancy between observed and predicted removal times. The predictive distributions of removal trajectories and epidemic duration are both investigated. Simulation studies suggest that our method can successfully distinguish the infection rate from the assumption of the SEIR stochastic epidemic models.

stochastic epidemic models, the posterior predictive distribution, model determination.

Received: March 2, 2022; Accepted: April 11, 2022; Published: April 27, 2022

How to cite this article: Muteb Alharthi, Model discrimination for epidemiological SEIR-type models with different transmission mechanisms, JP Journal of Biostatistics 20 (2022), 27-50. http://dx.doi.org/10.17654/0973514322012

This Open Access Article is Licensed under Creative Commons Attribution 4.0 International License

References:

[1] M. Alharthi, Bayesian model assessment for stochastic epidemic models, PhD thesis, University of Nottingham, 2016.
[2] M. Alharthi, T. Kypraios and P. D. O’Neill, Bayes factors for partially observed stochastic epidemic models, Bayesian Anal. 14(3) (2019), 907-936.
[3] O. N. Bjørnstad, B. F. Finkenstädt and B. T. Grenfell, Dynamics of measles epidemics: estimating scaling of transmission rates using a time series SIR model, Ecological Monographs 72(2) (2002), 169-184.
[4] R. J. Boys and P. R. Giles, Bayesian inference for stochastic epidemic models with time-inhomogeneous removal rates, J. Math. Biol. 55(2) (2007), 223-247.
[5] R. Deardon, S. P. Brooks, B. T. Grenfell, M. J. Keeling, M. J. Tildesley, N. J. Savill, D . J. Shaw and M. E. J. Woolhouse, Inference for individual-level models of infectious diseases in large populations, Statist. Sinica 20(1) (2010), 239-261.
[6] A. P. Dempster, N. M. Laird and D. B. Rubin, Maximum likelihood from incomplete data via the EM algorithm, J. Roy. Statist. Soc. Ser. B 39 (1977), 1-38.
[7] A. Gardner, R. Deardon and G. Darlington, Goodness-of-fit measures for individual-level models of infectious disease in a Bayesian framework, Spatial and Spatio-Temporal Epidemiology 2(4) (2011), 273-281.
[8] A. E. Gelfand and A. F. Smith, Sampling-based approaches to calculating marginal densities, J. Amer. Statist. Assoc. 85(410) (1990), 398-409.
[9] A. Gelman, J. B. Carlin, H. S. Stern, D. B. Dunson, A. Vehtari and D. B. Rubin, Bayesian Data Analysis, CRC Press, 2013.
[10] A. Gelman, X.-L. Meng and H. Stern, Posterior predictive assessment of model fitness via realized discrepancies, Statist. Sinica 6(4) (1996), 733-760.
[11] G. J. Gibson, G. Streftaris and D. Thong, Comparison and assessment of epidemic models, Statist. Sci. 33(1) (2018), 19-33.
[12] I. Guttman, The use of the concept of a future observation in goodness-of-fit problems, J. Roy. Statist. Soc. Ser. B 29 (1967), 83-100.
[13] E. S. Knock and P. D. O’Neill, Bayesian model choice for epidemic models with two levels of mixing, Biostatistics 15(1) (2014), 46-59.
[14] T. Kypraios, Efficient Bayesian inference for partially observed stochastic epidemics and a new class of semi-parametric time series models, PhD thesis, Lancaster University, 2007.
[15] P. E. Lekone and B. F. Finkenstädt, Statistical inference in a stochastic epidemic SEIR model with control intervention: Ebola as a case study, Biometrics 62(4) (2006), 1170-1177.
[16] P. Neal and G. O. Roberts, A case study in non-centering for data augmentation: stochastic epidemics, Stat. Comput. 15(4) (2005), 315-327.
[17] P. J. Neal and G. O. Roberts, Statistical inference and model selection for the 1861 Hagelloch measles epidemic, Biostatistics 5(2) (2004), 249-261.
[18] P. D. O’Neill, Introduction and snapshot review: relating infectious disease transmission models to data, Stat. Med. 29(20) (2010), 2069-2077.
[19] P. D. O’Neill and N. G. Becker, Inference for an epidemic when susceptibility varies, Biostatistics 2(1) (2001), 99-108.
[20] P. D. O’Neill and P. J. Marks, Bayesian model choice and infection route modeling in an outbreak of norovirus, Stat. Med. 24(13) (2005), 2011-2024.
[21] P. D. O’Neill and G. O. Roberts, Bayesian inference for partially observed stochastic epidemics, J. Roy. Statist. Soc. Ser. A 162(1) (1999), 121-129.
[22] P. D. O’Neill and C. Wen, Modelling and inference for epidemic models featuring non-linear infection pressure, Math. Biosci. 238(1) (2012), 38-48.
[23] D. B. Rubin, Estimation in parallel randomized experiments, Journal of Educational and Behavioral Statistics 6(4) (1981), 377-401.
[24] N. C. Severo, Generalizations of some stochastic epidemic models, Math. Biosci. 4(3) (1969), 395-402.
[25] G. Streftaris and G. J. Gibson, Bayesian analysis of experimental epidemics of foot-and-mouth disease, Proceedings of the Royal Society of London B 271(1544) (2004), 1111-1118.
[26] G. Streftaris and G. J. Gibson, Non-exponential tolerance to infection in epidemic systems-modeling, inference and assessment, Biostatistics 13(4) (2012), 580-593.
[27] S. Watanabe, Mathematical Theory of Bayesian Statistics, CRC Press, 2018.