When does contact tracing end up looking like a local lockdown? New analysis with @JoshAFirth @LewisSpurgin @HellewellJoel @petrakle @StephenKissler and others looks at the dynamics of different COVID-19 control measures in a high-resolution local network... 1/
We used data from a 2017 citizen science project, in collaboration with Cambridge colleagues, @FryRsquared and @BBCFOUR, which tracked hundreds of volunteers in the UK town of Haslemere over a three day period (with their consent) biorxiv.org/content/10.110… 2/ 
(By coincidence, this was where the first locally acquired UK case would later be reported in 2020: bbc.co.uk/news/uk-516834…) 3/
As well as standard contact tracing and physical distancing measures, we were interested in what effect tracing more widely (e.g. tracing contacts-of-contacts) might have, as has been done in place like Vietnam (ourworldindata.org/covid-exemplar…) 4/
Without control measures, we found simulated outbreaks could spread quickly in network. Isolation & contact tracing slowed transmission, but tracing contacts-of-contacts was more likely to bring it under control – albeit by putting large proportion of community in quarantine. 5/
It illustrates that contact tracing and local lockdowns/quarantines aren't a simple dichotomy - depending on how widely they are targeted, one can end up looking like the other, even if the initial approach is different. 6/
We also looked at the effect of some background physical distancing, by removing a proportion of the 'weak links' in the network (i.e. pairs of people who interacted rarely). This reduced the proportion in quarantine, but still enabled control. 7/
The dense, dynamic nature of real-world contact networks means that local outbreaks can spread rapidly, and in hard-to-predict ways. There won't be one simple solution to COVID – we'll have to balance a combination of measures, each with different effectiveness and disruption. 8/
Model code is available online (github.com/biouea/covidhm) and builds on earlier contact tracing model code by @HellewellJoel et al (github.com/cmmid/ringbp & thelancet.com/journals/langl…) 9/
Thanks to @JoshAFirth @LewisSpurgin for all their work on this (and for volunteering via the RAMP scheme in the first place – blogs.royalsociety.org/in-verba/2020/…), and of course all the participants in the original BBC citizen science project. 10/10
