Comparing Options For Safer Events

post by jefftk (jkaufman) · 2021-11-01T16:30:04.983Z · LW · GW · 19 comments

Yesterday I wrote about how indoor contra dancing, with one set of precautions, seems to be safe enough that I would be ok resuming organizing dances. How safe it is, however, depends a lot on which precautions organizers take. I looked over a range of potential options, many of them suggested in the comments, trying to figure out which ones are worth it. I'm going to try to keep this post general, though the specific scenario I have in mind is a social dance.

Summary: requiring vaccination, requiring (and offering) surgical masks or better, using air purifiers, not serving food, and discouraging cheering are worth it; checking vaccine cards and requiring rapid tests are not.

For each option, let's start by estimating the cost in time and money, and the benefit as a percentage reduction in transmission. For the latter, all numbers are from microcovid unless otherwise specified.

• Vaccination

  • Require participants to be fully vaccinated. Relative to no vaccination, reduces transmission by 0.5x and reception by 0.17, for a total of 0.08x. Relative to no requirement, depends a lot on your community. Since 81% of adults here are full vaccinated, requiring vaccination has a marginal effect of 0.3x.

    The vaccine is free to individuals, and takes two doses at under an hour each, plus, say, a 50% chance of losing a day to side effects after the second dose. This cost is shared over many activities and events, not just a single event. Let's scale the cost down by at least 365x to adjust for that sharing, though I could see arguments for much higher scaling factors. Cost is then 2.3min, for the 19% of unvaccinated possible attendees, or ~30s per person on average.

  • Verify vaccination status, by asking to see vaccine cards, physically or via photo. Whether this is worth it is going to depend a lot on how trustworthy your community is. At very high levels of trust, you can just take people's word. At extremely low levels of trust, and for a high demand event, you start to run into problems with people making fake vaccine cards. I'm going to estimate that about 10% of unvaccinated people (2% of population) would casually lie if asked but not go to the trouble of making a fake card. Effect is then 0.8x. Time is ~30s/person in checking and 2min/person in participant hassle.

    When estimating the harms of covid, at the end of the post, I ignore harm to people who lie about their vaccination status.

• Masks

  • Require participants to wear a mask. Assuming a mix of 1/3 snug cloth (1/3x on transmission, 2/3x on reception), 1/3 surgical (1/4x on transmission, 1/2x on reception), and 1/3 kn95+ (1/6x on transmission, 1/3x on reception), the overall effect from random pairings is 1/8x (same as assuming all surgical). At this point everyone already has a mask, because so many things require it, so no marginal cost on participants. This does make them event less enjoyable, and some people will not be able to attend due to breathing difficulties. This is much harder to adjust for.

  • Require participants to wear a surgical mask or better. These cost ~$0.08 in bulk, and could easily be provided by the event. This moves the mix to 2/3 surgical and 1/3 kn95+. Relative to the previous scenario, it's 0.8x.

  • Require participants to wear kn95 or better. These cost ~$0.40 in bulk and again could be made available at the event. Relative to the previous scenario, it's 0.6x. Some participants might find these unpleasant.

• Testing

  • Require participants to take a rapid test right before coming or on entry. Microcovid doesn't have this, but this estimate gets 1/4x. $12 each, plus 10min in hassle and testing. [This looks different outside the US where tests are often cheaper and more widely available.]

  • Require participants to present a recent PCR result. With the large delay between testing and results, I think this one is nearly useless unless combined with pre-event isolation.

• Air

  • Hold the event outdoors. 1/20x generally, but not for people within 1ft. For contra in particular this comes to ~1/5x. More hassle for organizers, worse surface for dancing, more weather risk, impractical in many places for much of the year.

  • Heavily ventilate the space. 1/4x, if you can get a level of ventilation equivalent to losing a full wall. How practical that says it's going to depend a lot on the location and weather.

  • Air purifiers. Microcovid has 1/4x for HEPA filters handling 5x the room volume hourly. I'm not sure using the room volume is correct here, since all things being equal I would expect doubling the volume of the room while keeping filtration capacity and occupancy constant to be safer, but this would say it's less safe? Microcovid cites Testing mobile air purifiers in a school classroom: Reducing the airborne transmission risk for SARS-CoV-2. The room was 27x20 with a 12ft ceiling, with 29 people. This is 233 ft^3/person, and their purifier was running at 600 CFM. For extrapolating to very differently shaped rooms (much taller) with very different occupancy (fewer people per sqft) I think it's probably more accurate to figure 20 CFM/person (600 CFM / 29). Posted on the issue.

    For a long-term installation it's probably worth buying a purpose-built filter, but for a shorter event the box fan approach is more promising. A Physics professor at the University of Arkansas recommends four filters per box fan and measured 470 CFM through the filters. At 19 CFM/person, for a 100 person event you'd need four of these. Each MERV-13 filter (they disagree with microcovid on whether HEPA is a good tradeoff) is $11, so this would be $176 in filters, or $1.76/person. My guess is you already have the box fans? Not that clear to me how long these last, but if your event is a few hours you might be able to reuse the same filters about 10 times, so $0.18/person.

• Activity

  • Excluding food. Many events often have food as a "nice to have", but once people start socializing over and they tend to do a lot of talking close together without masks. You can think of this as the inverse of requiring masks, skilled by the portion of the event for which people will be eating food. For example, if your event would otherwise be on average surgical masks, having 15% be socializing with food increases total risk from the event by 2.2x. For a social dance, where food is not a major component of the event, I think skipping it can make a lot of sense. For other sorts of events, you might come to a different conclusion.

  • Asking people not to cheer. A typical contra dance has some cheering, in addition to applauding. Figure a quarter of people cheer, average duration is 15 seconds, and they do it 15 times in a three hour period. This is 0.5% of person-time spent cheering. If cheering is 20x risky on its own, then allowing cheering increases total risk by 1.1x (10%). This does lose a bit of the energy of the event, but clapping is still pretty good?

  Summary, relative to a baseline indoor dance event in a cold place (so no outdoors or high-ventilation options) with masks and vaccination required, counting time at $25/hr. Sorted by cost-benefit, best first:

  • Air purifiers: 1/4x, $0.18
  • Require vaccination: 1/3x, $0.18
  • Surgical mask or better: 0.8x, $0.08
  • KN95 or better: 0.5x, $0.40
  • Don't allow cheering: 0.9x, minimal cost
  • Don't allow food: 0.5x, moderate cost
  • Check vaccine cards: 0.8x, $1
  • Rapid testing: 1/4x, $16

  Which of these are worth it? Let's say were willing to spend up to $10M to avert a covid death, and figure 0.2% of unvaccinated people and 0.001% of vaccinated people who get covid die. A microcovid (1:1,000,000) then is $0.02 if you're unvaccinated and $0.001 if you're vaccinated. Long covid is also a thing, as is missing work or other events while sick, So let's say that people are willing to pay $5,000 to avoid non-death aspects of covid. This doesn't affect the unvaccinated number appreciably, but increases the vaccinated number to $0.006.

  If we modify the calculation in yesterday's post to allow people not to be vaccinated, it goes from 120 microcovids to 400. At $0.02/microcovid for unvaccinated people and $0.006 for vaccinated, that's $3.50/person, compared to $0.72/person if we require vaccination. Since we're counting vaccination at $0.18/attendee, this clears the cost-benefit bar easily.

  For the other options, relative to the scenario in yesterday's post, we get a per-person net benefit (benefit minus cost) of:

  • Air purifiers: $0.36
  • Surgical mask or better: $0.06
  • KN95 or better: 0.5x, -$0.04
  • Check vaccine cards: -$0.86
  • Rapid testing: 1/4x, -$15.46

  Air purifiers and surgical masks are worth it, everything more expensive wouldn't be.

  On the other hand, perhaps you think my estimate from last time was too low. What if it is really 20x more dangerous, at 2400 microcovids? The risk is then $14, and we have:

  • Air purifiers: $10.32
  • KN95 or better: 0.5x, $6.60
  • Surgical mask or better: $2.72
  • Check vaccine cards: $1.80
  • Rapid testing: 1/4x, -$5.50

  Requiring vaccination, setting up air purifiers, and requiring/providing decent masks, all seem pretty good. Requiring rapid tests still does not. Implementing these would give you a total effect of 0.1x, bringing it down from 2400 microcovids to 240.

Comment via: facebook

19 comments

Comments sorted by top scores.