Mathematical models for dengue

In the framework of the XIV edition of the dengue international course

Three days Course on construction and use of mathematical models for dengue transmission and control

17 to 19 August, 2015

¨Pedro Kouri¨ Tropical Medicine Institute
and
The University of Florida, USA

Instructors:  Tom Hladish, Diana Rojas, Ira Longini (Florida University) and Martha Castro (IPK)

Language: English

Capacity: 20 participants maximum

Those interested contact Dr. Martha Castro to martac@ipk.sld.cu

Fees: 250.00 CUC (meals not included)

In this course, participants will learn to construct and analyze mathematical models for dengue transmission and control. They will learn about deterministic models, but we will concentrate on individual-level stochastic models.  The course will be taught via nine, one-hour models, with three taught per day.  There will be extra time to help participants with modeling exercises on the computer.

1.  Introduction:
–    What is a model, and why are they important in epidemiology?
–    Brief history of epidemiological models and their successful application.
–    Overview of types of models and their trade-offs.
2.  Natural history of dengue & Aedes spp.:
–    Mosquito-borne, symptoms, incidence and range.
–    Serotypes and primary/secondary infection dynamics.
–    Stages & durations
3.  Ross-Macdonald model:
–    Explanation
–    Implementation
–    Explore dynamics
4.  Determinism & stochasticity:
–    Introduce stochastic version of R-M
–    Consequences of deterministic vs. stochastic models
–    Choice of random variables for model mechanisms
–    Pseudo random numbers and seeding
5.  Agent-based modeling:
–    Natural history to model structure
–    Computational and storage complexity concerns
–    Language and hardware choices
–    Brief discussion of parallelizeable approaches
6.  Object-oriented programming:
–    People, mosquitoes and locations as variables
–    Advantages of literate programming
–    Pseudocode and real-code examples
–    Add functionality to a simple, working agent-based model
7.  Data sources and limitations:
–    Epidemiological
–    Demographic
–    Entomological
–    Geographic
8.  Model fitting and validation:
–    Models must be validated for a particular purpose
–    Bayesian parameter estimation
–    Priors, posteriors, and metrics
–    ABC tutorial
–   Brief discussion of more advanced topics like convergence, identifiability and correlations
9.  Interventions:
–    Available and emerging options
–    Availability and limitations of efficacy/effectiveness data
–    Model representations
–    Discussion of effect on model dynamics