SF case count now starting to look a little more clearly sigmoidal. I've been somewhat surprised that this hasn't flattened out more aggressively, now 15 days into shelter-in-place. Do we think that's due to testing backlog?
Conversation
This delayed flattening seems about what we should expect, given that symptoms appear up to 14 days after initial exposure? (Or maybe I'm misinterpreting what data the graph is showing?)
2
Replying to
My reading was that incubation period median is around 5 days. ncbi.nlm.nih.gov/pubmed/32150748
I always assumed 14 days was to see the results in hospitalization numbers (incubation + progress to severe disease). When it's driven by widespread testing it should be far less... If only severe cases get tested then it makes sense though.
4
Replying to
Ah, I see now—yeah, if that estimate is still accurate, that curve does seem delayed a bit. I agree, it's weird, maybe delayed testing as other suggest? (It would be interesting to see the distribution to see if incubation is skewed somehow.)