95% Effective Vaccine

December 8, 2020 7:20 AM

I'm confused abut what 95% effective for the vaccine means in a context where social distancing and other preventative measures are in place. Is the vaccine only 95% effective in this context and if the social distancing measures are removed does it become less effective? Does that matter?
posted by edbles to Health & Fitness (15 answers total)

The 95% effectiveness measure derives from the numbers of Covid infections observed among the vaccinated group vs. the control (placebo) group in the trials. If a vaccine were 100% effective, you would expect zero cases among the vaccinated, and some number among the controls. The actual observation was that the number of cases in the vaccinated group was only about 5% of the number observed in the control group, meaning that the vaccine was 95% effective in preventing the expected number of cases. Both groups were asked to behave as they normally would with regard to masking and distancing, so the study is inconclusive with regard to masking and distancing.
posted by beagle at 7:31 AM on December 8, 2020

Yes, a randomized control group means that all other preventive mechanisms should be the same for both groups. In absolute terms, however, mechanisms work together. A great vaccine isn't very effective in the context of out-of-control spread.

The vaccine as fire hose (NYT): "An analogy may be helpful here. A vaccine is like a fire hose. A vaccine that’s 95 percent effective, as Moderna’s and Pfizer’s versions appear to be, is a powerful fire hose. But the size of a fire is still a bigger determinant of how much destruction occurs. ... At the current level of infection ... a vaccine that is 95 percent effective ... more than 160,000 would die.

... [if] the vaccine was only 50 percent effective but the U.S. had reduced the infection rate to its level in early September ... the death toll ... would be .. about 60,000. "
posted by Mr.Know-it-some at 7:55 AM on December 8, 2020

Basically, regardless of what other measures are taken, we should expect to see 95% fewer infections in a vaccinated population vs. a non-vaccinated population.

So if you have a population that is great about masking and quarantining and there's very little community spread and new daily cases are like 1 in a million, when you add in the vaccine, cases would go down to 1 in 20 million. If you have a population that is a hot mess and cases are 1 in 1000, new daily cases would go down to 1 in 20,000.

Herd immunity makes a bit more complex than that (because if enough people are immune we interrupt transmission chains), but it's useful to think of all the different preventative measures we can take as being largely independent.
posted by mskyle at 7:55 AM on December 8, 2020

I thought it meant that if I come into close contact with 100 people who are spreading covid, that I would only have a 5% chance of getting it?
posted by halehale at 8:40 AM on December 8, 2020

No. On a micro level, roughly, it means that on average you are 5% (one-twentieth) *as likely* to catch COVID, or 95% less likely, from a single encounter than if you hadn’t gotten the vaccine, *all else equal*. It says nothing about base probabilities. This is an assumed number based on averages over large sample sizes.
posted by supercres at 8:55 AM on December 8, 2020

Eh, I take that back. now I’m not so sure if “single encounter” really applies. That wasn’t tested. It’s about likelihood of catching it over the course of the trial period, so there could be zero, one, or multiple exposures for trial participants. Again, averages over otherwise-similar large samples.
posted by supercres at 8:58 AM on December 8, 2020

Is the mechanism more "the vaccine works (almost) perfectly for 95% of people and the other 5% are (mostly) unprotected by it at all" or "the vaccine works better for some people than others, but on average, they are 95% less likely to get it"?
posted by jacquilynne at 8:59 AM on December 8, 2020

I thought it meant that if I come into close contact with 100 people who are spreading covid, that I would only have a 5% chance of getting it?

Definitely not the case! You already have a lower-than-100% chance of getting COVID from those interactions. Also the vaccine might not actually reduce your personal chance of infection by 95% - it could be better or worse. For instance, a vaccine might work perfectly for 95% of people and not at all for 5% of people. It would still be a 95% effective vaccine.

To answer Jacqueline's question - I don't know if we know that yet and the answer is probably somewhere in between, but the vaccines also seem like they may prevent really severe disease. I don't think there have been any reported cases of severe disease in the vaccinated group. So it might even be "provides 100% effectiveness for 95% of people and 50% effectiveness for the other 5%" (THESE NUMBERS ARE MADE UP).
posted by mskyle at 8:59 AM on December 8, 2020

Important to note too, this is a coronavirus and those are hard to immunize against. All the major vaccine candidates are effective against severe disease, not infection. (Compared to something like the measles vaccine that also prevents infection.)

So you could still catch COVID-19 while vaccinated, but you are less likely to need to be ventilated and less likely to die because you don't develop the severe form of the disease.

(Source: Dr Norman Swan on Coronacast)

Piggy back question- surely this doesn't help with the spread of disease if not everyone is immunised? Instead of noble herd immunity you are protecting yourself from severe disease, not infection full stop.
posted by freethefeet at 9:14 AM on December 8, 2020

As noted above: as yet, the trials cannot prove that transmission is affected in any way by the current top three vaccines (Pfizer, Moderna, and AstraZeneca/Oxford.) Pfizer, for example, is 95% effective at preventing severe cases, not at stopping transmission. Unless and until research shows otherwise, all vaccinated people will have to keep up with choosing well-ventilated areas if they must gather indoors, avoiding crowds, social distancing, masks, hand-washing, etc. because even the successfully vaccinated may be able to acquire and transmit the virus.
posted by maudlin at 9:59 AM on December 8, 2020

And as soon as I post, there is some preliminary evidence for stopping transmission: tweet by Matthew Herper at STAT.

Interpretation from David Anderson at Balloon Juice:

The above graph is part of the vaccine efficacy data that is being submitted to the FDA by Pfizer. The blue line is the number of cases over time for people in the vaccine arm. The red line is the number of COVID cases by time for people in the placebo/control arm.

Looking at the curves, it looks like vaccines provide little to no protection within the first ten days. The height of the blue line is about the same as the height of the red line.

By Day 14 there is starting to be a gap.

By Day 28 the gap is a chasm.

After Day 28, it seems like the vaccinated arm will get infections here and there, but no huge leaps.

This is an amazing curve.

And from an infection breaking perspective, this is probably good enough to break infection chains and crush the reproductive rate even if we made the completely unnatural and unsupported assumption that the immunity is very short term and fades after three or four months. This would require frequent re-vaccination/boosters and fire break strategies, but even in a fairly bad case scenario that is not supported by the data, this is very good news.

posted by maudlin at 10:16 AM on December 8, 2020

So it depends on the vaccine, of course - because I just ran across an article about the Oxford vaccine reducing symptoms much more than transmission.
posted by clew at 11:04 AM on December 8, 2020

As a note, AFAICT, we don't have evidence that shows the vaccine prevents 95% of infections. We instead have evidence that the vaccine prevents 95% of the worst effects of the infection. This is a pretty big difference particularly if many people can't or won't be immunized.

Preventing infection stops the pandemic. Reducing the severity of symptoms saves live, but not all lives, and could allow the pandemic to continue.
posted by jclarkin at 11:07 AM on December 8, 2020

IMO we should stop referring to "the" rather than a particular vaccine because we're in for a year of people angrily attributing everything they've ever heard about any vaccine to whichever one they're depending on or avoiding, let's not encourage that. It's going to be confusing enough anyway.
posted by clew at 12:10 PM on December 8, 2020

95% of those who would have gotten the virus otherwise didn't get it.

100% did not get a severe infection.

source: https://therapeuticseducation.org/podcast/episode-463-covid-19-vaccine-%E2%80%93-all-numbers-so-far-and-much-more
posted by mikek at 3:47 AM on December 9, 2020