Are you puzzled about how long you need to keep your food packages undisturbed to inactivate the virus that causes COVID-19? Then this tweet thread is for you. (1/34)
First let me start by suggesting that there's no evidence of transmission of COVID-19 from food or food packaging. There's also no reason to disinfect food packaging after arriving home from the grocery store. (2/34)
But let's say there is some theoretical risk there that we are trying to assess or manage. What would we need to know? How would we assess then manage that risk? (3/34)
We should start with the idea that virus inactivation is dependent on temperature. Inactivation happens faster at higher temperatures and slower at lower temperatures. (4/34)
I took data from a couple of different papers to try to figure out the effect of temperature on the rate of reduction of the virus. There's a paper on the original SARS virus: pubmed.ncbi.nlm.nih.gov/15350737/ (5/34)
And then there's this pre-print on the new SARS-CoV-2: medrxiv.org/content/10.110…. And the data do lineup pretty well. (6/34)
From the data in these two papers I was able to construct a table which shows the relationship between temperature and rate of reduction. It will be a side tweet off of this tweet in the thread. (7/34)
There's also some other data that I've gotten a lot of attention. These come from the NEJM nejm.org/doi/full/10.10… those authors report a worst case half-life of 6.8 hours on plastic. Let's round up to seven hours for simplicity. (8/34)
I should also note that the NEJM paper does not mention experiment temperature, so let's assume room temperature (~72 F). These data don't line up exactly with the earlier data, but they're in the right ballpark. (9/34)
What does it mean to have a "half life"? It means that the viable virus concentration drops by 50% every seven hours. Food microbiologists reading this may see that it's possible to convert "half-life" to "log reduction rate", that we commonly use. (10/34)
In this case a half-life of seven hours corresponds to roughly a one log reduction per day (actually about 23.26 hours, but you get the idea). That's actually slightly faster then the rate I calculated in the table above. (11/34)
Next, let me suggest that you stop thinking about the amount of "time" needed to make something safe. It's not about *time* it's about rate. What do I mean by that? (12/34)
"Inactivation time" is a function of the starting concentration, the rate of reduction, and the growth sensitivity of your method (i.e. detection limit), so any mention of "time" presupposes those three values. (13/34)
What would be a logical starting concentration? We know based on research with and without face masks: nature.com/articles/s4159… that 30 minutes of breathing without a mask gives about 10,000 virus particles from some individuals. (14/34)
Those virus particles spread throughout a store are mostly going to end up on the floor, but even if they end up on food they will gradually inactivate overtime and as you'll see below transfer rate to hands would be low. (15/34)
From the same paper swab samples of nose and throat (unknown volume) give ~1M virus particles, which might be a worst case sneeze or cough. But what does any of this mean in terms of risk? (16/34)
It means based on the NEJM letter that your chance of finding an infectious virus particle someone has sneezed, coughed or breathed on food drops by 50% every seven hours, or by one order of magnitude (1 logarithm) every day. (17/34)
But how do we know if we are going to get sick? What's the risk of illness from one or more virus particles? (18/34)
To answer this question we need a dose response model. We don't have a dose response model SARS-CoV-2, but we do have one that's been estimated for the original SARS. onlinelibrary.wiley.com/doi/full/10.11… (19/34)
Based on that model if you ingest 10,000 virus particles the odds of getting sick are 100%. If you ingest 284 particles your odds are about 50-50, and if you ingest a single viral particle your risk is just a fraction of 1% (0.24% to be exact). (20/34)
But what happens when the predicted number of virus particles on a surface is less than one? Does it mean there's no virus particles? No. (21/34)
If we predict that the number of virus particles is 0.1, another way of thinking about that is if we tested 10 surfaces, we would expect to find a single virus particle on one of them. (22/34)
I think it's pretty unlikely that those hypothetical virus particles sitting there on packages of food are going to suddenly jump off and aerosolize, but they could get on your fingers. (23/34)
We don't have data on cross-contamination for the virus that causes COVID-19, but based on our research on cross contamination with other organisms, I'd estimate about 1% of the virus particles (plus or minus) will transfer to the hand. (24/34)
Of course at this point it would be a really good time to wash your hands or use hand sanitizer. Probably not a good idea to stick your finger in your nose. (25/34)
So where does this leave us? If you think about the math, what it means is that nothing is ever completely safe. Sorry if that freaks you out a little bit, but that's the nature of risk and probabilities. (26/34)
So what to do? Well, start by not leaving your groceries on your porch... if it's cold outside, it won't help. I'm still not planning on disinfecting my groceries either, because the risk is very low. (27/34)
The real risk is other people coughing/sneezing/breathing one *you* in the grocery store. Now that CDC is recommending face covers (i.e. masks) cdc.gov/coronavirus/20… that risk should be less too. (28/34)
I will keep washing my hands and using hand sanitizer, especially when returning from the store or any other trip. (29/34)
If it makes you feel better to sanitize your groceries (or for that matter wash your produce in soap), go for it, but I don't want to know. (30/34)
Thanks as always for reading. If you have questions or comments, I'll do my best to read and respond. (31/34)
Trolls will be blocked. People that just want to argue will eventually be muted. (32/34)
If you think I should've made a video instead, sorry. You're free to make your own video and use this as your script. (33/34)
Stay home if you can, wash your hands and use hand sanitizer, and take care of those who need it most. (34/34)
Here’s that table: 
