Serial interval label:question

[vleotti]

Hi, thank you for the app. I have a theoretical question actually. Can I use the serial interval distribution for incident cases (like the one from Li et al) as the serial interval distribution for the hospitalization series? I suppose the interval between two infections is smaller than the interval between two hospitalizations in the same chain.

Best, Vanessa

[thibautjombart]

Hi Vanessa, that is good point. In using this we implicitly assume that admissions are shifted from the onset of symptoms by a constant number of days, which in practice is not the case. In practice the admission interval (if we call it that) may be more dispersed than the serial interval. This said, I would not necessarily expect there to be a bias / different average time. Why would you expect the interval between admissions to be shorter than onset of symptoms (or infections, but that would be generation time, not serial interval)?

[vleotti]

Hi,

Thank you very much for the answer and sorry for the delay. I actually was thinking about the interval between an infection and the admission of the secondary case, but now I realize this does not make sense. I think you are correct, it is reasonable to say that the "admission interval" has the same average as the serial interval, but it is interesting to consider greater variability.

Best, Vanessa

[image: Mailtrack] https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5& Remetente notificado por Mailtrack https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5& 30/06/20 19:11:47

Em qua., 17 de jun. de 2020 às 05:21, Thibaut Jombart < notifications@github.com> escreveu:

Hi Vanessa, that is good point. In using this we implicitly assume that admissions are shifted from the onset of symptoms by a constant number of days, which in practice is not the case. In practice the admission interval (if we call it that) may be more dispersed than the serial interval. This said, I would not necessarily expect there to be a bias / different average time. Why would you expect the interval between admissions to be shorter than onset of symptoms (or infections, but that would be generation time, not serial interval)?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/thibautjombart/covid19_bed_occupancy/issues/85#issuecomment-645229691, or unsubscribe https://github.com/notifications/unsubscribe-auth/APDBGZVZHWACLAA53KZMQQ3RXB4IZANCNFSM4N73DCOA .