We are seeing more cases of COVID-19 in people who got vaccinated. Here’s why that’s not cause for concern.
Matthew Fox
In 1970 two towns straddling the border of Texas and Arkansas saw an outbreak of measles. In one town vaccination rates were relatively high, and they saw lots of cases among the vaccinated. Not surprisingly, this led to concerns that perhaps the vaccine didn’t work, or that it didn’t work as well as expected. It turned out, after careful investigation, that the vaccine did work just fine. And in fact this phenomenon of seeing vaccinated cases in highly vaccinated populations is to be expected and is not cause for concern. So much so that the Centers for Disease Control use this example as a teaching case in their epidemiologic surveillance training programs.
The same phenomenon is happening today with COVID-19 as we continue to scale up vaccination. As of June 21st, 2021, the CDC reports 4,115 breakthrough infections in vaccinated people in the US that resulted in hospitalization or death among over 150 million people vaccinated. But it is hard to hear about a recent outbreak among vaccinated players of the New York Yankees and not think the vaccine isn’t 90% effective as we’ve been told the Pfizer and Moderna vaccines are. Examples like this make headlines and cause wide concern. Is the vaccine failing? The answer is, in most cases, no, the vaccine is doing exactly what it is supposed to and saving lots of lives. And this can be confusing and counterintuitive. Epidemiologists who study vaccinations are very aware of this phenomenon and yet, even though we expect it, it is hard to communicate the accurate and important message to people that this isn’t a sign of vaccine failure.
So what’s going on? Imagine a town where 80% of the population is fully vaccinated and let’s say that the vaccine is 90% effective, which we will define to mean that 90% of those who got the vaccine are protected against illness (in reality the definition is a bit different, but not enough that it will matter for our purposes). Since no vaccine is 100% effective, even if we vaccinated everyone, some of the vaccinated will remain vulnerable to infection.
Back to our town. If the town had 10,000 people, 8,000 would be vaccinated and 2,000 would not. Suppose everyone was exposed to the Coronavirus (SARS-CoV-2), something that doesn’t happen in reality, but let’s see what would happen. Of the 8,000 vaccinated, 7,200 would be protected against getting sick, while 800 would be vulnerable to infection. Of the 2,000 unvaccinated, let’s say all are vulnerable to infection (in reality, some would have immunity from prior infection). If everyone who was exposed and vulnerable became infected, we’d see 800 cases in the vaccinated and 2,000 in the unvaccinated. Based on those numbers, 800 vaccinated cases and 2,000 unvaccinated cases, it would seem lots of people who were vaccinated got sick. And that could lead people to worry the vaccine doesn’t work. But they would be wrong.
The problem, as with anything in epidemiology, is the denominators. When you see people who are vaccinated who are getting infected (or even hospitalized or dying) you are missing all the people who got the vaccine and did not get sick. In our case, this was 7,200 people.
Back to our town. Among the vaccinated, 800/8,000 got sick (10%) while among the unvaccinated 2,000/2,000 got sick (100%). It’s now easy to see that if you are vaccinated, you are much less likely (in fact 90% less likely) to become sick than if you are not vaccinated. But because so many more people are vaccinated than unvaccinated, if we only focus on the infected cases and not the denominators (population vaccinated compared to population unvaccinated), we can come away with the wrong impression. And this is a pretty common phenomenon for vaccines. When a new vaccine is introduced into a population, when vaccination rates become very high, we can expect as many and sometimes more cases in the vaccinated population than the unvaccinated — even when the vaccine is doing its job and protecting people at a high rate — because the denominator of number of people vaccinated is so large.
This is not to say that there could never be a vaccine failure, or that the new variants couldn’t, theoretically, reduce vaccine effectiveness in the future. And this is why the CDC continues to conduct surveillance on the population and ensure the vaccines are working. But we should expect to see more cases among vaccinated people as the vaccine rollout continues (and the denominator increases) and this should not be cause for concern.
So when you see a headline about cases of COVID-19 in the vaccinated, take a minute and remember to ask — what’s the denominator. That simple question can provide a much more informed perspective and reassurance about our progress in fighting this disease.
