Predicting patient flow congestion using extreme value theory

This project constitutes a thematic follow-up to the on-going project 'Predicting patient flow congestion using extreme value theory' (hereafter called the on-going project). The project is part of the new context of a world panic-stricken by the pandemic due to the SARS-CoV-2 and its disease COVID-19. At this major turning point for humanity, the goals of the project naturally extend the aims of the on-going project and the particular focus on virus-relatedpandemic congestion will be addressed. More specifically, the aims of this project are:-- to develop models for 'time to recovery' after a period of congestion by analyzing therelation between the duration of a congestion and its magnitude,-- to develop branching processes to explore the extremes of loads (staff and/or bed utilization) ow creating congestion at different units in the hospital,-- based on new risk management developments, to improve the predictability power ofthe 'Congestiometer' that has recently been implemented at the University Hospital ofLausanne (CHUV) and that serves as a measure of the congestion.To achieve these goals in a real hospital context, we will benefit from the new and valuabledatabase (with confidentiality assured according to the strictest standards) of patients testedfor COVID-19 from the CHUV in addition to the database provided by the CHUV emergencyunit.