MATHEMATICAL MODELLING TO INVESTIGATE INFECTIOUS DISEASE DYNAMICS AND CONTROL STRATEGIES

Infectious diseases still represent one of the major threats for human health due to both their direct and indirect effects on public health and worldwide economies. Despite the current possibility to eradicate or control certain infections like smallpox, polio and measles, the increase in incidence of new infections (so called emerging diseases), or the increase in incidence or geographic range of ones that have existed previously (re-emerging diseases), poses a new threat to public health. To further complicate things, the role of animals in the insurgence and spread of new diseases is central. Of all emerging diseases indeed, the 60.3% originate by, or involve into their cycles, animals, and represent the so called zoonoses. The increase in number of emerging and re-emerging infections and their potential to fast spread into animal and human populations make central the development of tools to reduce human infection risk. Epidemiological studies become then central to understand the relationship among events, investigate their causal effects and understand risk factors. Despite of that, classical epidemiology, centred on the study of the relationships between events, show limits in the investigation of mechanisms underlying infection spread, and in considering the interactions among populations, thus possibly leading to simplistic and spurious conclusions. Mathematical modelling instead, and the development of a “dynamical epidemiology”, allows the investigation of dynamics of infections, thus providing us a mechanistic point of view to understand infection spread. The strengths of mathematical modelling applied to epidemiological studies are several. At first, they can investigate the extent to which an event can mechanistically influence another consequential event. This characteristic of mathematical modelling has a great application in public health, as it allows to prioritise interventions or studies on those events that have a major impact on disease outbreak. Another strength of mathematical modelling is its ability to describe the dynamics of an infectious disease by accounting for interactions among populations and sub-groups of populations within the same population. At least, mathematical modelling permits for theoretical investigations of mechanisms of transmission and to answer to “what if?” questions, allowing to explore theoretical scenarios which have not yet occurred or which needs to be preventively tested, like the application of an intervention strategy. With the present work we then provide four applications of mathematical modelling to infectious disease. We focused on two wildlife-originating infections: West Nile virus (WNV) and baylisascariasis. Both infections are emerging or re-emerging in Italy and can represent a threat for human beings due to their possible severe outcomes. Due to the potential harm they are for human beings, a thorough surveillance and a wide intervention and control plans are ongoing both to promptly identify the presence and circulation of their causative agents and to reduce human infection risk. A full understanding of WNV cycle is fundamental to reduce human infection risk, but several knowledge gaps still exist, especially on the role played by different bird species involved in its spread. For both infections moreover, despite several are the control strategies proposed a quantitative analysis of their performance has never been performed. Aimed at filling these gaps, we developed a mathematical model to simulate WNV spread, and used it to explore mechanisms driving infection spread. We found birds recovery rate and mosquito biting rate having the major influence on disease spread and thus being the most urgent mechanisms to be investigated via field and laboratory experiments. Birds’ susceptibility and their competence to infection have a negligible influence on disease spread, thus making investigations to understand them of secondary importance. These results might be of aid also in defining the characteristics of a bird species to be a good WNV spreader, by focusing the attention on species that have a small recovery rate or are frequently bitten by mosquitoes. Moreover, we found a negative effect of birds’ abundance in affecting WNV prevalence in mosquitoes, further helping us in distinguishing among species that are suitable to have a role in WNV spread. We then exploited the built model to explore intervention strategies against WNV. We showed that a reduction of the vector population is more effective than a reduction of birds’ abundance in an area. In particular, the best efficacy is shown by the reduction of mosquito breeding sites, followed by the active elimination of their eggs and larvae. On the contrary, reducing the abundance of competent birds or their reproductive sites can obtain an increase in human infection risk. Similarly, we also studied the effectiveness of different intervention strategies to reduce the number of Baylisascaris procyonis eggs in the environment. The ingestion of B. procyonis eggs indeed is the cause of baylisascariasis, an infection that can have severe health consequences in human beings. With our work we explore the effects both in terms of efficacy (i.e., potential to eliminate eggs from environment) and efficiency (i.e. the timing needed) of three different intervention strategies. The interventions tested are: the active culling of raccoons, raccoons’ anthelmintic treatment and faeces removal. We found that raccoon culling might have the best and faster results, highlighting the importance of assessing the intervention on the base of an objective prove on its efficacy. With the proposed work then, we highlighted the role of mathematical modelling in epidemiological studies, by, at first, exploiting their potential to investigate the extent to which an event can influence another consequential event. Secondly, we used them to describe the dynamics of an infectious disease, by accounting for interactions among populations, and focusing on mechanisms underlying infection spread. Moreover, we also exploited them for theoretical investigations, like the simulation of the application of an intervention strategy to reduce human infection risk is. In conclusion, mathematical modelling can widely help our understanding and management of infectious diseases through a new and different point of view from that provided by classical epidemiology. Mathematical modelling indeed includes the investigation of spreading mechanisms and non-linearity of interactions among individuals and subgroups of populations, thus allowing a more complete comprehension of diseases spread. The cooperation of diverse health professionals is fundamental to fully exploit both classical epidemiological studies and dynamics ones, and the effects of their cooperation can lead to a better knowledge of infections and a consequent reduction of human infection risk.