Note: I have joined the “virtual class” component of Dan Kahan‘s Science of Science Communication course at Yale University. As part of this I am endeavoring to write a response paper in reaction to each week’s set of readings. I will post these responses here on my blog – my response for week five is below. Week 1 is here, and week 2 is here. (I was away for weeks 3 and 4.)
I will also be participating in the discussion on Kahan’s own blog.
This week I seek to examine several myths that Kahan and Discover blogger Keith Kloor say news media have perpetuated in the wake of the Disneyland measles outbreak.
Trends in MMR immunization
The most easily disprovable myth is that measles, mumps and rubella (MMR) immunization is falling. (This myth has been perpetuated by, among others, the Los Angeles Times.) The best source for this data is the CDC’s National Immunization survey. This data shows that since 1999, the percentage of children aged 19-35 months who had received the MMR vaccination held roughly steady, at between 90 and 93 percent (see chart). The slight fluctuations in that range have shown no particular trend.
While national immunization rates are important, more localized rates are crucial because pockets of low immunity can allow outbreaks to take hold. As the CDC notes, MMR coverage was below the 90 percent threshold in 17 states in 2013, and that definitely presents a problem. But is it a growing problem? I used CDC data to put together a quick spreadsheet of five- and ten-year trends, state by state, and at a glance found many states whose rates have dropped – but not by a statistically significant margin. The confidence intervals here are pretty high. (If anyone has done a serious trend analysis, I’d love to see it.)
Are parents concerned about vaccine safety? What does that mean?
In a January 22 piece on the Washington Post’s Wonkblog, Christopher Ingraham blames “the anti-vaccine movement” for the worrying rise in measles cases, citing an AP-GfK survey that found, as Ingraham puts it, “only 53 percent of Americans were confident that vaccines are safe and effective.” For a start, that’s a pretty big misrepresentation of the survey, in which 53 percent were very or extremely confident that childhood vaccines are safe and effective. Another 30 percent were somewhat confident.
In any case, Kahan argues that the AP-GfK survey isn’t a good measure: “Indeed, no public opinion survey of the general public can give anyone useful information on vaccine risk concerns. The only valid evidence of that generally is the National Immunization Survey, which uses actual vaccine behavior to determine vaccination rates,” he told Kloor.
I think we can agree that the NIS represents the best way to measure what proportion of parents are affected by concerns or other factors strongly enough to substantially affect their children’s immunization program. After all, if the concern is strong enough to actually affect vaccination outcomes (and surely those are the concerns we’re most interested in) then we should see that in a measure of vaccination outcomes!
However, there are several important things the NIS can’t tell us. Notably, it doesn’t give insight into the reasons behind non-vaccination. We are right to ask what economic, psychological and social factors are behind parents’ failure to immunize against measles, particularly in geographic or socioeconomic pockets that fall below target immunity, because knowing the causes of missed immunizations will help us formulate the best science communication response.
One study that has engaged with this question is Dempsey et al, “Alternative Vaccination Schedule Preferences Among Parents of Young Children.” The results of this study have been misrepresented by the Advisory Board Company and subsequently by the Post’s Petula Dvorak – the latter, for example, says “one in 10 parents are avoiding or delaying vaccines in their children because of safety concerns.”
Dempsey’s findings are more complex. She found that 13 percent of parents of young children reported using an alternative vaccination schedule (that is, they reported that did not completely adhere to the CDC vaccination schedule). Of these, a strong majority of 82 percent – but by no means all – agreed that, “Delaying vaccine doses is safer for children than providing them according to the CDC-recommended vaccination schedule.” Some parents who followed an alternative vaccination schedule may have done so because of difficulty accessing medical care, or because they simply failed to immunize on time.
I would note Dempsey’s small-ish sample size of under 800 (compared to over 13,000 in the CDC’s survey), and also potential motivated reasoning on the part of her respondents. That is, a parent whose original reason for delaying vaccine was a lack of time or simply forgetfulness might convince himself retroactively that the reason was “delaying vaccine doses is safer for children.” But, even given all these caveats, I think Dempsey makes a decent case for the power of safety concerns to drive non-vaccination – so I would be interested to hear any rebuttals.
Who doesn’t vaccinate, and why?
Having said that, safety concerns are but one part of a complex set of drivers – and some of the most destructive myths around non-vaccinators are about what drives them and, even more potently, who they are.
Many commentators have described non-vaccinators as “anti-science” or lacking “trust” in science and medicine. For the most part these are labels that the non-vaccinators would not themselves recognize. I could write another post entirely on whether that matters, but let me sum up for now by saying it does matter, at the very least, because such language is polarizing and alienating. For example, an NPR caller challenged Paul Offit on what she saw as his presumption that non-vaccinators are “yahoos who just don’t look at the scientific process at all.” Did she show a less than scientific mindset when she rejected Offit’s explanations as “pap,” arguing that “a two-year-old cannot accept this kind of chemical onslaught”? Perhaps. But she doesn’t see herself as rejecting science. She sees herself as a critical thinker – indeed, as she says, “an educated adult.”
As if that weren’t polarizing enough, we now have several stereotypes emerging about who non-vaccinators are: either rich hippies, religious nuts, or conspiracy theorists. As Dvorak writes: “The fringe who didn’t believe in medicine for religious and other reasons has exploded into a 10 percent, largely yuppie epidemic.” In the same article, George Washington University public health professor Alexandra Stewart says the non-vaxxers are primarily “white, educated populations of people with computers.”
The reality is much less homogenous – and less ideological. First, on a broad scale, MMR vaccination rates are roughly equal across most racial lines (91.5 percent for whites, 90.9 percent for black non-Hispanic, 92.1 percent for Hispanics) and a little higher for Asian-Americans (96.7 percent) and American Indians (96.3 percent). Second, as Kahan found in his study of 2,316 adults, Vaccine Risk Perceptions and Ad Hoc Risk Communication: An Empirical Assessment, “There is no meaningful correlation between vaccine risk perceptions and the sorts of characteristics that usually indicate membership in one or another cultural group.” These groups measures included a sliding scale of political orientation (liberal-conservative, Democrat-Republican) as well as two latent measures of risk-perception: tendencies to perceive risk to public safety and to social deviancy. (These risk perception measures were generated from subjects’ perception of risk from a variety of other issues, including climate change, marijuana legalization and gun ownership, and correlate with the common Hierarchy-Egalitarianism and Individualism-Communitarianism worldview scales.)
I would also point to some interesting work done by Yvonne Villanueva-Russell, who in extensive interviews found a variety of motivations for those who failed to vaccinate their children to the CDC schedule. Some of these parents seemed to fit the stereotypes (“crunchy mommas,” “ideologues”) but others simply put off making a decision for too long, or have children with health issues. Her sample size of 67 parents is small, but Villanueva-Russell’s work gives us some idea of the range of motivations we should take into account when designing further research – and before we speak out of hand about “anti-vaxxers.”