How likely are you to get COVID these days?
I confess that I don’t do this kind of calculation but I instinctively come down pretty much in the same place anyway. I’m fully vaxed and boosted, I have avoided the virus so far, would certainly get Paxlovid if I do get it and am healthy. But at my age, you can’t be too careful so I’m pretty much where Dr. Bob Wachter is in this thread. Perhaps the more nerdy among you will enjoy his more elaborate risk calculation:
Covid (@UCSF) Chronicles, Day 915
As I hoped https://mobile.twitter.com/Bob_Wachter/status/1558234980706164736 , cases in the U.S. have dropped steadily. Up until now, I’ve avoided indoor dining and worn a mask in all crowded indoor spaces. I’m now ready to eat indoors & (selectively) remove the mask. Here’s why: (1/25)As I said recently, my threshold to liberalize my behavior is <5 cases/100K/d in my region. (After accounting for home tests, 5/100K/d is really ~25/100K/d.) U.S. is now at 19, CA is at 12, & SF is at 6 (& fell 64% in past 2 wks). Find your # here:
https://www.nytimes.com/interactive/2021/us/covid-cases.html (2/25)
Asymptomatic test positivity rate @UCSFHospitals now 1.6% (⬇from 4-6% in early Aug). This means that ~1/60 people who feel OK would test pos. for Covid. We use a PCR-like test, which will stay pos. for ~14 days on average, whereas the avg. period of infectiousness is ~8d. (3/25)
So while 1/60 asymptomatic people will test positive for Covid today in SF, ~1/100 will be infectious. Let’s use that (1%) as our #.
If each person has a 1/100 chance of being infectious, in a group of 10 people there’s a 9.6% chance that at least one will be infectious. (4/25)(In a group of 150, say on a packed plane, there’s a 78% chance ≥1 is infectious – which is why I’ll continue to wear a mask on planes & in crowded indoor spaces when I don’t need to talk to anyone, probably forever. I figure, why take ANY risk in situations like these?) (5/25)
OK, so if I dine today in a restaurant in SF, there’s a 1 in 10 chance that at least one infectious person will be nearby. How risky is that?
The household attack rate for Omicron is about 40%… and that’s with prolonged indoor exposure to a family member or roommate. (6/25)How risky is a meal-length exposure to an infectious person? Dunno; so many variables – ventilation, distance, vax status. Let’s say there’s a 10% odds of infection. If so (& 1-in-10 odds of an infectious person), that’s a 1% chance of being infected during an indoor meal. (7/25)
Now’s when your own risk status & tolerance kicks in. Is 1% chance of getting infected too high? For my 86-year-old mom I think it is, since (even with her 5 vax’s–she got her bivalent last wk) she has a ~5x ⬆chance than me of dying if she gets Covid. And she’s in Boca,…(8/25)
…which is running 20 cases/100K/d, or 3.5x SF’s rate of 6 – making her chances of exposure that much higher.
(By the way, she’s mostly not listening to me since for her the costs of social isolation are too high and her friends are all eating indoors. Which I understand.)(9/25)For me, as a healthy vaxxed/boosted 64-year-old in SF, I think the risk has dropped to acceptable levels – though I’ll still favor eating outdoors when feasible (where the risk is ~nil).
How does booster status influence this? I’ve had 2 primary shots, 2 boosters, &… (10/25)… got my bivalent yesterday. I’ll start crediting my booster as further ⬇my odds of infection next week; it’ll reach peak efficacy after 2 wks. How much will it lower the odds? Studies of the mRNA vaccines in Omicron era show that its efficacy in preventing infection… (11/25)
…is ~50% – far less than the 90% we saw pre-Omicron, but still something. Problem is that effect wanes to zero after 2 months. (The efficacy in preventing SEVERE infection/death is better & lasts much longer – the main rationale for boosters.)
Will the new bivalent vax…(12/25)… prevent Covid for more months? Probably but no human data yet.
Whether it does or doesn’t, next week I’ll consider my odds of getting infected to be 50% lower. If prevalence in SF stays the same, this’ll make my new odds of being infected while eating indoors 1 in 200.(13/25)What if I do get infected? While asymptomatic infections happen, I’ll assume that – at best – I’ll feel crummy for several days, need to isolate for ~1 wk, & have a chance of infecting others – all things I’d like to avoid.
But what are odds of the 2 really bad outcomes: (14/25)1) A severe acute case of acute Covid, and
2) Long Covid?
Viz #1, the current case-fatality rate in those fully boosted is about 0.1% (similar to flu) – it’s much higher in the unvaxxed and unboosted, & it’s influenced by the usual risk factors. https://www.ft.com/content/e26c93a0-90e7-4dec-a796-3e25e94bc59b (15/25)At a 0.1% infection-fatality rate, the odds of death-by-indoor-meal would be ~1 in 100,000 (1/100 risk of infection x 1/1000 risk of death). And I’d take Paxlovid if I got Covid, which should ⬇it by another ~50% (efficacy of Pax in vaxxed people https://pubmed.ncbi.nlm.nih.gov/35653428/), (16/25)
… which would ultimately place my estimated chances of dying from my indoor maskless dinner at 1-in-200,000. This is consistent with other risks we all take to do some things we enjoy. https://theconversation.com/whats-most-likely-to-kill-you-measuring-how-deadly-our-daily-activities-are-72505 (17/25)
I’m far more worried about Long Covid. Odds of getting LC from a single case is ~10-20% in unvaxxed; it drops by ~50% in vaxxed. Let’s say it’s 5% – that’d mean that my 1/100 chance of getting Covid from a single meal would translate into a 1 in 2,000 chance of getting LC.(18/25)
I’m also persuaded that my case of Covid will increase my odds over time of having a heart attack, stroke, diabetes, & cognitive decline. https://www.nature.com/articles/s41591-022-01689-3 The magnitude of that increase isn’t well established. As a back-of-the-envelope, let’s guess it’ll be of … (19/25)
…similar magnitude to the risk of symptomatic Long Covid. If so, that would make the overall odds of a negative long-term bad effect from a case of Covid in a vaxxed/boosted person 10% per case (5% risk of symptomatic LC; 5% increase in odds of long-term bad outcome). (20/25)
So, putting it all together, at current prevalence levels in San Francisco, the rough odds following a single indoor meal here (if 10 people are in aerosol range) for someone vaccinated (but not boosted or infected in the past 2-3 months) would be: (21/25)
• Being exposed to someone with Covid at dinner: ~1/10
• Getting Covid from the dinner: ~1/100
• Getting some version of Long Covid (protracted symptoms and/or ⬆risk of a bad outcome): 1/1000
• Dying from Covid obtained at dinner (if one takes Paxlovid): 1 in 200,000 (22/25)Two weeks after taking the bivalent booster (or having a new infection), I’d lower the odds of getting Covid by 50% (to 1-in-200), which should proportionally lower the risk of both death and Long Covid. This benefit will likely last ~2 months, maybe more w/ bivalent vax. (23/25)
If you’ve been in careful mode, as I’ve been, are these risks now low enough to enjoy indoor dining in SF? To me, they are.
Are they low enough to leave the mask off when entering an uncrowded indoor space or having a small group work meeting? To me, yes.
(24/25)Are they low enough to ditch the mask in a crowded & poorly ventilated indoor space or on an airplane? For me, the answer’s still no. And would they be low enough in a place w/ >10 cases/100K/d? To me, not yet.
For you? Your call. I hope this helps you think it through. (25/end)Originally tweeted by Bob Wachter (@Bob_Wachter) on September 18, 2022.
Yep. Masks for me in crowded, indoor spaces with poor ventilation and airplanes. Indoor restaurants ok, but outdoor still preferred. Your mileage may vary.