Thread by @MackayIM

Just a reminder that we have two-fifths of FA real data on how any single person or group gets infected by SARS-CoV-2.
We're not there to observe the process. We mostly can't exclude surface or droplet or aerosol. We can apply what we expect to have happened.
But remember...

...we do that using our current settings with all their glorious biases. Some of the knowledge that came before has evidence that's a degree removed - we know SARS-CoV-2 survives in aerosols & can infect primates; but we haven't infected a human that way.

We know droplets are emitted when we cough, yell & sing & not when we breathe or talk normally. But aerosols are always emitted so how do you know droplets were in play & not time (dose) of aerosol exposure if you're proposing close-up infection was only by droplets? You can't.

We know virus survives on nice clean surfaces (sometimes in the dark & in ideal conditions & big doses) but we don't know how it lasts on surfaces in the real dusty, greasy, sunny, windy, oft-contacted world.

We also know that virus-contaminated fingers can spread virus that can be recovered from light switches and other surfaces & fomites thanks to work even from decades ago. But we can't point with absolute assurance to any person having been infected that way by SARS-CoV-2.

We know SARS-CoV-2 infects in a chunky way - not every person infects 2or 3 others, some infect none, some infect many. This "overdispersion" can't always be explained by proximity or shared surfaces, but aerosols fit. Sometimes it isn't that way & multiple routes are in play.

So embrace the uncertainty here folks!
And while you see scientists picking faults in what looks like any scientific paper from anywhere, make sure you ask them to tell you where the better data are as well. Or, if there are no better data, then that's also important to know