Ask HN: Simple formula to combine new cases and test positivity?

1 point

4 years ago

As the next wave hits, many of us are in regions where new daily cases are reported along with test positivity rate, but those are difficult numbers to consider separately if you wish to set objective thresholds for masking or going out at all.

microCOVID has an adjusted prevalence percent that uses Youyang Gu's simple formula https://covid19-projections.com/estimating-true-infections-revisited/ which multiplies cases by the square root of test positivity and another factor, 2.

I contacted him about whether this was still a good model for a simple dashboard I made https://knoxville.day/ and he responded, "...that equation is from 2020. I don't think it's relevant anymore, so I wouldn't use it. The true prevalence multiplier is much higher now due to the availability of at-home testing."

Is there a better, still simple formula out there for combining these two numbers into one? Without it, I'm lost as to whether 100 cases a day at 40% positive is worse than 200 at 10% and can't make evidence-based decisions on what to do. Or, if much more advanced models exist, is there a dashboard that takes historical cases, positivity, and population into account to estimate the percent in my county with a currently active infection?

1 comment

1 comment