Washington Breakthrough Survey
For today's post, I'm looking at Washington. Washington has been putting out weekly reports on breakthrough cases, and breaking them down by age. This allows me to find the recent rates of covid diagnosis for Washingtonians 65+ in the vaccinated and unvaccinated populations, and thereby to get a vaccine efficacy number for just this age range.
This is important because those 65+ were given access to vaccines first, and so have had more time for their antibodies to wear off. Older people also have weaker immune responses in general, and while the trials found that vaccine efficacy in older people was just as good as in younger people, it's possible that the immunity might wear off faster in older people.
Looking at the three weeks from July 4 through July 24, the rate of covid for fully vaccinated people 65+ in Washington State was 0.029%. The rate for those not fully vaccinated was 0.613%, for a vaccine efficacy of 95%. That's based on the Washington health department's numbers for breakthrough cases, and the CDC's numbers for the percentage of the population 65+ in Washington who were fully vaccinated at this time.
I was pretty surprised to see that 95%, which is exactly the efficacy number from the first 2-3 months of vaccination for those in the trial. And I need to give my usual caveats. This analysis is not a randomized controlled trial. It is possible that people who are the most worried about the virus are more likely to be vaccinated and also more likely to take other measures, and that would tend to skew vaccine efficacy upwards. Given that older people who are unvaccinated are more likely to experience a more severe case, they are probably more likely to seek testing, whereas vaccinated older people who get infected might write their symptoms off as allergies and not get tested. This would also skew efficacy upwards.
When I do the calculation for all ages, vaccine efficacy comes out at 88%. It seems really unlikely that the vaccines are working better for older people than younger people. Aside from the above caveats, I think at least part of the explanation goes back to the problem I mentioned before. Younger people are more likely to have had a prior infection and therefore at least some immunity. That partial immunity in the unvaccinated population makes vaccine efficacy appear lower. The CDC's serosurvey data has less than 5% of those over 65 in Washington state testing positive for anti-nucleocapsid antibodies that indicate prior infection. That means that this population is a better place to test vaccine efficacy than is one with a higher level of immunity from infection.
I don't think that this means that older people who are vaccinated should be complacent about their risk. This is only a rough estimate based on imperfect data, and we are still on the upswing for the spread of the delta variant in the US. Even though the vaccines apparently still reduce your risk of catching it, if the background risk is high, then the reduced risk after vaccination might still be unacceptable. Eventually enough people in the US will have gotten the delta variant that we will have herd immunity, and the risk will come back down. It will get worse before it gets better, but it will get better.