Thread by @adamjkucharski

I often see the misconception that control measures directly scale COVID case numbers (e.g. “hospitalisations are low so measures should be relaxed”). But in reality, measures scale *transmission* and transmission in turn influences cases. Why is this distinction important? 1/

If discussions are framed around the assumption of a simple inverse relationship between control and cases, it can lead to erroneous claims that if cases/hospitalisations are low, control measures can be relaxed and case counts will simply plateau at some higher level. 2/

But of course, this isn’t how infectious diseases work. If control measures are relaxed so that R is above 1, we’d expect cases - and hospitalisations - to continue to grow and grow until something changes (e.g. control reintroduced, behaviour shifts, immunity accumulated). 3/

If control measures are keeping cases flat at 10k per day (for example), those same measures would also keep things flat if cases were at lower level. In fact, given a choice of R=1 and a high or low infection level, there are two benefits to going for the low option... 4/

First, it means less COVID burden in terms of hospitalisations and deaths. And second, it means more capacity to use targeted measures (e.g. test & trace) to keep transmission down, which in turn could allow other types of measures to be relaxed. 5/

We need discussions about what measures should look like, and what is feasible/sustainable. But we also need to frame any discussions around the actual dynamics of SARS-CoV-2 as a contagious disease, not under simplistic assumptions about control vs cases. 6/6