Analysing group testing algorithms for COVID-19 surveillance and case identification in UK schools, universities, and health and social care settings

In the context of a pandemic disease such as Covid-19, group testing (also known as 'pooled testing' or 'batch testing') provides a way of identifying infected individuals in a large group, by pooling samples together in different ways and testing the pooled samples. Group-testing algorithms can yield large efficiency gains over individual testing, when the number of infected individuals (i.e., the prevalence) is relatively low. This project will analyse whether and how strategies based on group-testing algorithms may be effectively used for monitoring Covid-19 prevalence, and for case-identification (i.e., screening of asymptomatics to identify infected individuals), in UK schools, universities, care homes and health care settings. Close attention will be paid to all the practically relevant issues: the impacts on public health and education-loss, resource requirements/costs, the practical ease or difficulty of implementing different strategies in situations of changing (and geographically varying) prevalence, the feasibility of incorporating a pooling step into existing UK testing systems, and ethical/anonymity requirements. As well as producing technical preprints and papers, we will produce non-technical reports aimed at policymakers and written in a way accessible to those without mathematical or scientific training, to aid decision-makers in determining which protocols to implement. Our initial focus will be on schools and universities, and later we will extend our study to health and social care settings.