Model selection and parameter estimation for dynamic epidemic models via iterated filtering: application to rotavirus in Germany.

Despite the wide application of dynamic models in infectious disease epidemiology, the particular modeling of variability in the different model components is often subjective rather than the result of a thorough model selection process. This is in part because inference for a stochastic transmission model can be difficult since the likelihood is often intractable due to partial observability. In this work, we address the question of adequate inclusion of variability by demonstrating a systematic approach for model selection and parameter inference for dynamic epidemic models. For this, we perform inference for six partially observed Markov process models, which assume the same underlying transmission dynamics, but differ with respect to the amount of variability they allow for. The inference framework for the stochastic transmission models is provided by iterated filtering methods, which are readily implemented in the R package pomp by King and others (2016, Statistical inference for partially observed Markov processes via the R package pomp. Journal of Statistical Software69, 1-43). We illustrate our approach on German rotavirus surveillance data from 2001 to 2008, discuss practical difficulties of the methods used and calculate a model based estimate for the basic reproduction number $R_0$ using these data.

Dynamic modeling of hepatitis C transmission among people who inject drugs.

To reach the WHO goal of hepatitis C elimination, it is essential to identify the number of people unaware of their hepatitis C virus (HCV) infection and to investigate the effect of interventions on the disease transmission dynamics. In many high-income countries, one of the primary routes of HCV transmission is via contaminated needles shared by people who inject drugs (PWIDs). However, substantial underreporting combined with high uncertainty regarding the size of this difficult to reach population, makes it challenging to estimate the core indicators recommended by the WHO. To support progress toward the elimination goal, we present a novel multi-layered dynamic transmission model for HCV transmission within a PWID population. The model explicitly accounts for disease stage (acute and chronic), injection drug use status (active and former PWIDs), status of diagnosis (diagnosed and undiagnosed) and country of disease acquisition (domestic or abroad). First, based on this model, and using routine surveillance data, we estimate the number of undiagnosed PWIDs, the true incidence, the average time until diagnosis, the reproduction numbers and associated uncertainties. Second, we examine the impact of two interventions on disease dynamics: (1) direct-acting antiviral drug treatment, and (2) needle exchange programs. As a proof of concept, we illustrate our results for a specific data set. In addition, we develop a web application to allow our model to be explored interactively and with different parameter values.

A stochastic model for the normal tissue complication probability (NTCP) and applicationss.

The normal tissue complication probability (NTCP) is a measure for the estimated side effects of a given radiation treatment schedule. Here we use a stochastic logistic birth-death process to define an organ-specific and patient-specific NTCP. We emphasize an asymptotic simplification which relates the NTCP to the solution of a logistic differential equation. This framework is based on simple modelling assumptions and it prepares a framework for the use of the NTCP model in clinical practice. As example, we consider side effects of prostate cancer brachytherapy such as increase in urinal frequency, urinal retention and acute rectal dysfunction.