Read on Twitter
A friend asked me about why results from our recent paper on aerosol and surface stability disagreed with those from another study that got press suggesting the virus was "viable on plastic for up to nine days" This gave me a chance to explain a common misconception
(THREAD)
(THREAD)
Here's our study: https://www.nejm.org/doi/10.1056/NEJMc2004973
People quoted a "9 days of viability on plastic" from that paper and a "3 days of viability from ours. Both are misleading. Why?
Absolute times to virus undetectability depend on the initial quantity of virus. If you start with more virus, you'll have at least some infectious virus around for longer. If you start with less, it'll be undetectable sooner.
This is why we estimated decay rates/half-lives. The virus decays exponentially: every hour makes you safer, but the biggest changes happen in the first few hours. @TomAvril1 did a great job summarizing these points in the @PhillyInquirer https://www.inquirer.com/health/coronavirus/coronavirus-covid19-how-long-lives-on-surfaces-20200319.html
If this all seems a bit academic, it isn't: we really don't want people to think there's a binary threshold between when things are dangerous and when they're safe. Every minute makes you safer, but whether "you're safe" depends on many things...
e.g. how much virus was initially deposited? How susceptible are you? How much virus in general is needed to infect a human being?
Even so, how to reconcile our 3 days with their 9? First of all, that paper has no SARS-CoV-2 data. It reviews studies of more benign human-native coronaviruses, as well as SARS-CoV-1 and MERS. The 9 days figure is for SARS-CoV-1, and comes from the following table:
2 things to notice:
1) 9 days is by far the largest number in the table, and it's from a paper that found 6–9 days
2) Look at the inoculum! 10^7. Much higher than most of the comparison studies. So it's not surprising that that study also had the longest time to undetectability
1) 9 days is by far the largest number in the table, and it's from a paper that found 6–9 days
2) Look at the inoculum! 10^7. Much higher than most of the comparison studies. So it's not surprising that that study also had the longest time to undetectability
Our study looked at SARS-CoV-1 as a comparison virus for SARS-CoV-2. We found a half life for SARS-CoV-1 on plastic of between 6 and 9 hours (bottom right panel in this plot), with the most probable value just shy of 8.
How long should a virus with that half life take to decay to undetectability if you start with 10^7 TCID50 of virus? Recall that a half life is the amount of time that it takes the virus to be cut in half. Our detection limit is 10^0.5. How many halvings will it take?
We do a bit of algebra:
10^7 * (1/2)^x = 10^0.5
10^6.5 = 2^x
log10(10^6.5) = log10(2^x)
6.5 = x * log10(2)
x = 6.5/log10(2) =~21.5 halvings
10^7 * (1/2)^x = 10^0.5
10^6.5 = 2^x
log10(10^6.5) = log10(2^x)
6.5 = x * log10(2)
x = 6.5/log10(2) =~21.5 halvings
Suppose we're on the upper end of our plausible half lives, at 9ish hours. 9 hrs * 21.5 = about 8 days. So the results suddenly don't sound so different!
Lesson: read the methods section, and remember that the more virus you start with, the more virus you end up with!
(end)
Lesson: read the methods section, and remember that the more virus you start with, the more virus you end up with!
(end)
Actually, while you're all here, you should watch this awesome video. @HolyNameHealth is saving lives through DIY cleverness, and other hospitals should know about this and do the same! https://twitter.com/dylanhmorris/status/1248426857344544768
