During the COVID-19 pandemic, the urgent need to rapidly understand the situation and implement effective public health measures led public institutions to collect an unprecedented volume of data. In Geneva, this effort was implemented through the ARGOS database, a centralized system initially developed for operational purposes such as case follow-up and contact tracing, but which soon provided opportunities for addressing broader scientific questions.

This thesis builds on this database to demonstrate how statistical and epidemiological approaches can provide insights into diverse dimensions of an unfolding epidemic. Following an extensive introduction that contextualizes the pandemic and presents original epidemiological observations from Geneva, the manuscript examines: the incidence of COVID-19 acquired within hospitals and long-term care institutions; the interplay between testing practices, socio-economic inequalities, and gender; the performance of contact tracing in capturing secondary infections; and the impact of vaccination and natural infection on viral transmission.

Taken together, these investigations illustrate how a single database can be mobilized through a wide variety of statistical approaches to address distinct yet complementary questions. They underline the critical role of data science and applied epidemiology in transforming raw data into knowledge, and in informing effective responses to public health crises.