I think there's a lot of recognition now that coronavirus arrived here earlier than many thought based on the testing. I think I have some strong evidence on just how early that I haven't seen presented elsewhere. Although our detection of coronavirus in America has been pretty poor early on, we do have a vast surveillance network for influenza like illness (ILI). Healthcare providers report data to the CDC on the number of patients with 1. A fever over 100 and 2. a cough/sore throat. Looking at this data, some have noticed a spike in flu like illness that could really be a spike in coronavirus before the testing ramped up. I have a different way of looking at this data.

We know that coronavirus disproportionately impacts the older people. We also know that, unlike the flu, it has very little impact on younger people. I decided to look at the percentage of cases in the ILI data that are 25+ going all the way back to 1998 using CDC data. This is calculated as the percentage of ILI reported for those 25 and older divided by all cases reported. I looked at the data by week. for 2020 compared to all other years. What you see is that this year is an extreme outlier. Starting around week 6-7 (mid-February), the portion of 25+ as a percentage of the total was the highest it has been in the last 20+ years. The number starts to grow and then completely takes off. It seems very clear that this year is like no other. See for yourself below.

So what does this all mean? For coronavirus to start showing up in these numbers, you would probably need hundreds if not thousands of cases. Think about what was happening in the world when these numbers started to rise. There were 15 confirmed cases in the US, all traced back to foreign travel or imported. Barr had just publicly rebuked Trump's Twitter habit. Coronavirus was still a distant threat many weren't even thinking about. But at that point it was already here, likely in large numbers. There were almost certainly many cases well before we shut down any borders. It could also mean there are millions more cases in the US than we think today.

I'm not sure how many conclusions to draw from this. Do not take this post as me arguing that we're already at herd immunity levels. I think that's very unlikely, and this analysis wouldn't prove it. It does imply that there were many more cases than we knew back then and many more cases today than we now know. I don't know the exact numbers, but likely a lot. I don't want to overstate the case because this is a very rough approximation. But it's a piece of evidence that is worth pointing out.

EDIT: raw numbers for those who care. Feel free to check me. This was done quickly.