Supervisors info:
Βασιλική - Αναστασία Σύψα (Επιβλέπουσα), Αναπληρώτρια Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Νικόλαος Δεμίρης, Επίκουρος Καθηγητής, Τμήμα Στατιστικής, ΟΠΑ
Δημήτριος Παρασκευής, Αναπληρωτής Καθηγητής, Ιατρική Σχολή , ΕΚΠΑ
Summary:
Epidemics have plagued mankind since ancient times, with the most recent example being that of the COVID-19 pandemic caused by the new SARS-CoV-2 coronavirus. This thesis concentrates on the use of mathematical models for the study of infectious diseases spread through direct contact. Mathematical models utilize data on the epidemiological characteristics of the pathogen and the disease and allow to assess the course of the epidemic and the effectiveness of various strategies to reduce transmission. Aim of the thesis is the review of mathematical models, the detailed presentation of models for influenza and COVID-19, the application to data from Greece and the evaluation of the effectiveness of various strategies
Initially, we provide a brief historical review of the epidemics that have affected humanity as well as a reference to modern epidemics of infectious diseases transmitted through direct contact. Furthermore, we present basic epidemiological concepts, such as the basic reproduction number R0 and the generation time Tg. Then, we perform a review of the models that can be used to describe the course of epidemics of pathogens spread through direct contact, and in particular models for COVID-19 and influenza. We analyze the SEIR (Susceptible-Exposed-Infectious-Recovered) model in which we take into account the presence of symptomatic and asymptomatic patients and measures such as case isolation, vaccination of susceptible individuals and other interventions.
We apply extensions of the SEIR model on data from SARS-CoV-2 pandemic and influenza H1N1/2009 in Greece. More specifically, we estimate the effective reproduction number for SARS-CoV-2, from the beginning of August to November 21, 2020 and we investigate the effect of social distancing and isolation of symptomatic patients under various scenarios. We also estimate the effective reproduction number for influenza, from the end of August 2009 to mid-February 2010 and we examine the effect of treatment, separately and in combination with vaccination. Finally, we compare the effectiveness of the above interventions.
According to the results of the present analysis for SARS-CoV-2, Rt in the country was around 1 from the end of August until about mid-October 2020. However, before the enforcement of the nationwide lockdown in November, Rt increased to around 1.5 to 1.75, while during its implementation - that is from November 7 onwards - it decreased and towards the end of the month it fell below 1. According to the simulations performed, interventions such as social distancing leading to the reduction of social contacts by 80% as well as the isolation on the day of onset of symptoms, although unrealistic, leads to control of the epidemic. If the measures are milder, the infections are certainly reduced, but the epidemic cannot be brought under control. Therefore, we conclude that in the absence of pharmaceutical interventions very strict measures of social distancing or isolation are required to achieve control of the COVID-19 epidemic.
We also estimated that the basic reproduction number for influenza H1N1/2009 in Greece was 1.5. The effective reproduction number was close to 1.2 shortly after the start of local transmission and then decreased below 1. If 10% of the symptomatic patients received antiviral treatment, 52% of the total population would be infected until the end of the epidemic. If treatment was combined with vaccination 2 months after the start of the epidemic, with a daily vaccination rate of 0.2%, 30% - 40% of the total population would be infected by the end of the epidemic depending on the effectiveness of the vaccine (versus the estimated 58% in the absence of measures).
In conclusion, mathematical models utilize the available epidemiological data and the characteristics of the population under study and allow to assess the healthcare burden as well as to evaluate and compare the effectiveness of alternative strategies - pharmaceutical and non-pharmaceutical - to reduce the epidemic.
Keywords:
Mathematical models, Epidemics, SEIR, COVID-19, Influenza
