How can we be sure that the current measles outbreak is really due to too few people getting vaccinated? And what was the immunization coverage of visitors at Disneyland during the days the virus was circulating there?
A study in JAMA today addresses both these questions with some pretty clever calculations. The second question sounds like an impossible one to answer, but, in fact, we know enough about measles and its vaccine to come up with a pretty good estimate of coverage at Disneyland and in the various communities where cases are occurring.
“The authors have done an interesting and well thought out mathematical experiment to explore what coverage would have to be in order to produce the case counts we’ve seen, given what we know about the transmission dynamics of measles,” said Jessica Atwell, a Ph.D. candidate in epidemiology and global health at Johns Hopkins Bloomberg School of Public Health. And what they find, she said, is that the communities where these cases are occurring are well under the herd immunity threshold needed to prevent the disease from traveling, somewhere between 50 percent and 86 percent instead of the minimum 95 percent typically recommended. “It highlights the importance of making sure we can address the concerns of vaccine hesitant parents and continue our efforts to keep coverage as high as possible,” she said.
Atwell has similarly researched how much low vaccination rates contribute to outbreaks of a particular disease, but she has typically done what most researchers do: compare immunization coverage rates to clusters of disease cases to see if there’s an association between them. This study came at the issue from a different angle — practically in reverse, actually — and for good reason. We don’t really have the data we need on exactly how many people are immunized against measles in individual communities.
Most data on U.S. vaccine refusal rates are based on children’s non-medical exemptions to required school immunizations. But using these as a metric is often problematic: parents might seek an exemption if a child received no vaccines at all or even if the child received all but one vaccine, and information about which vaccines the child did and didn’t receive is not always available to researchers. It’s therefore impossible to know, at least across all communities, exactly what exemption rates mean in terms of how many vaccines those children did or didn’t receive and which ones we’re talking about. Further, even the data that are available usually aren’t reported at all geographical levels. Overall national and state data on exemptions and on kindergartners’ immunization rates (by vaccine) are available, and some states require counties or schools to publicly report these data. But in many states, data at the county level isn’t readily available, much less at the city, zip code or neighborhood levels. And knowing children’s immunization status in kindergarten doesn’t tell us their vaccination status when they’re older.
So given all these limitations of the usual route, these researchers tried a different approach. They used what we know about how effective the vaccine is, how contagious measles typically is, how long it takes for one case to lead to another, and where cases occurred to estimate coverage, basically going the other direction: How many people must be skipping the MMR for us to see the number of cases we’re seeing?
“Trying to figure out the coverage level at a school is one thing,” Atwell said. “Knowing or measuring what the effective measles coverage might have been among a bunch of random people attending Disneyland over the course of a few days in December is a totally different thing, not to mention coverage among the nebulous ‘population’ of people exposed to those primary cases from Disneyland in the weeks during their infectious period.”
So lead author Maimuna Majumder and her colleagues at Boston Children’s Hospital used an equation to model the possibilities. They used 95 percent for the MMR vaccine’s effectiveness, a conservative estimate based on the protection seen with one dose. (The second dose brings the effectiveness up to 97 to 99 percent.) They used 10, 12 and 14 days as the possible timeframes that passed from the start of one infection to the start of another originating from the first. And they calculated scenarios in which one person infected three others, infected four others or infected six others. This data all comes from what’s been seen in the current outbreak. (In a community where no one is immune, one person would typically infect 11 to 18 others, but fortunately, there are no major areas with zero immunization coverage.)
They found that for one person to infect at least three others, the vaccination rates range from 75 to 86 percent. When one person is infecting four others, then vaccination rates range from 66 to 81 percent. And if one person with measles is giving it to almost six others, just 50 to 71 percent of the community is vaccinated. The bottom line? We ain’t got herd immunity, not by a long shot. And that’s why we’re seeing the cases grow so quickly.
“As the authors explain in detail, due to the extremely contagious nature of measles, the level of immunity needed in a population to stop transmission is very high, upwards of 95 percent,” Atwell said. “While overall immunization coverage at the state level is above this coverage threshold – less than three percent of California schoolchildren have personal belief exemptions – we know there are many communities in California and elsewhere in the U.S. where coverage is much lower, well below 95 percent, and it is within these pockets that disease transmission can get a foothold.”
From what the authors calculate in this study, most of the communities seeing measles transmission have vaccination coverage below 86 percent.
“This preliminary analysis indicates that substandard vaccination compliance is likely to blame for the 2015 measles outbreak,” the authors wrote. “Clearly MMR vaccination rates in many of the communities that have been affected by this outbreak fall below the necessary threshold to sustain herd immunity, thus placing the greater population at risk as well.”
The answer is not always so simple. With pertussis, for example, several other factors besides vaccines uptake contribute to increasing infections, such as a vaccine whose protection wanes after five or so years and changes with the bacteria that causes whooping cough.
“We do have data to support that clustering of unvaccinated individuals is contributing to the pertussis resurgence when herd protection is lost at the local level, but experts in the pertussis community believe that there are multiple factors at play, and underimmunization is one of several contributing factors,” said Atwell, who conducted one of those studies. “We don’t have any of these issues with measles and the measles vaccine, leaving underimmunization as the likely culprit.”
Of course, these findings are pretty much what the medical and scientific community have been saying, but it’s important to have data to support a hypothesis. Data from national immunization surveys show very high rates of immunization across the U.S. as a whole. But when you dig down into the nitty-gritty of individual communities, the picture can change.
“It has been clear for some time that looking at state-wide or even county-wide immunization coverage can be misleading,” Atwell said. “People interact with others in their communities and schools, so local level immunization coverage is where we need to focus to understand where pockets of risk may be greatest.”
Measles was eliminated from the U.S. in 2000, which meant it no longer circulated in states on its own and cases only occurred when someone arrived with an infection from abroad. But measles is thriving throughout much of the world, with approximately 400 children dying each day from measles. Every hour, 16 children die from measles somewhere in the world.
“As long as measles continues to circulate in the rest of the world, the U.S. will always be at risk of importation cases, either from Americans traveling abroad or visitors coming into to the US,” Atwell said. “It may seem like these diseases are no longer a risk, but if this outbreak teaches us anything, it’s that measles can come back in a heartbeat if we get too complacent.”