Well, the media have shown little interest in this. I guess empirical evidence on #COVID19 is less interesting than the antics of SAGE or the #RussiaReport. So here's a #thread about what we did and what we found. #epitwitter 1/11 https://twitter.com/martinwhite33/status/1283537075019460608

We treated the #COVID19 epidemic as a natural experiment in 149 countries with data, conducting interrupted time series and meta-analysis of the impact of 5 physical distancing policy measures on Incidence rate ratios (IRRs) 2/11

Data were from the @BlavatnikSchool @UniofOxford Covid response tracker, as well as the @ecdc_eu and @worldbank data portals. Fantastic work by @_Nazrul_Islam and @_ShShabnam bringing this all together and conducting the analyses 3/11

We looked at the effects of: closures of schools, workplaces, and public transport, restrictions on mass gatherings and public events, and restrictions on movement (lockdowns) between 1 January and 30 May 2020. 4/11

We conducted a range of sensitivity analyses, such as using different outcome variables, different time lags and excluding some very large countries and have gone to great lengths to discuss the limitations of our methods. 5/11