Dynamic Models for Public Health

Dynamic models are important tools for public health purposes; they allow esimaing the impact of diferent approaches and can be used for a populaion health improvement planning or to organize a response to an emerging public health issue. The aims of my thesis are to design, build, test, parameterized, and illustrated two dynamic models. One is a populaion-based model of Human Papilloma Virus (HPV) infecion natural history and vaccinaion; the other is an individual based model of tobacco control. Both models have been adapted to account for public health measure of prevenion and control. Also, both models can incorporate enough demographic data to adapt their outputs to speciic populaion, in which realisic intervenions are simulated. The model of HPV transmission and control has been adapted and parameterized to represent diferent populaions from both high- and low-middle-income countries, i.e. Sweden, Italy, US, and India, and diferent HPV types, i.e. 16,18, and 45. By contrast, the model of tobacco control has been designed to reproduce smoking behaviors and demography of the Italian populaion between year 2000 and 2013 and used to evaluate the efect of public health policy, e.g. smoking prevalence on the Italian populaion in year 2030. Both models are coded in C compuing language, according to high-performance programming standards, eicient data structures and algorithms, and opimizaion techniques to maximize the compuing eiciency. In my thesis, I have used the two models to predict the expected impact of selected public health intervenions both for HPV and cervical cancer control and for tobacco control. The outputs of each set of simulations have been analysed using advanced staisical methods/libraries, e.g. GNU Scieniic Library and sotware such as R (CRAN) and STATA. These predicions illustrate the potenial of using mathemaical model for assessing the efeciveness of selected prevenion and control measures.