It’s being suggested that Columbus, Ohio, is a new hot spot in the US based on reported cases. But diving deeper into the data shows why that is not accurate.

The reported cases curve is not a representation of the current trend because it reports on old cases.

Since there is a time lag from when a test is administered until the test is reported, the reported cases curve represents the rearview mirror of the cases trend; it’s what already happened and not what is happening.

The Columbus epi curve, in contrast, is a real-time representation of cases because it shows the date of symptom onset - the date each person first experienced a symptom. This is a more accurate representation of the trend of cases because it tells when symptoms are happening.

The data is showing that the peak of symptom onset was in late June and has been on the decline for July, showing Columbus could very well be past the peak.

@wyattsheepie
@jackwindsor