University Place

Vaccine Refusal and Public Health

– [Tom Zinnen] Welcome everyone to Wednesday Nite @ the Lab. I’m Tom Zinnen. I work here at the UW-Madison Biotechnology Center. I also work for UW-Extension, Cooperative Extension. And on behalf of those folks and our other core organizers, Wisconsin Public Television, Wisconsin Alumni Association, and the UW Science Alliance. Thanks again for coming to Wednesday Nite @ the Lab. We do this every Wednesday night. 50 times a year. Tonight it’s my pleasure to introduce to you Malia Jones. She works at the Applied Population Lab which is right up the street here. The Applied Population Lab is part of both UW-Madison’s College of Agriculture and Life Sciences and UW-Extension Cooperative Extension.

Malia was born on Kealakekua, on the big island of Hawaii. On the Kona side. She grew up in Philadelphia and Malvern, Pennsylvania, which is where she want to high school. Then she went to Washington, D.C. to go to undergrad at the American University where she studied philosophy, religions and ethics. (laughter) Then she went to University of California Los Angeles to study public health where she got both a Masters of public health and a Ph.D. Then she went over to the dark side at the University of Southern California. Where she did a post-doc. And then two years ago, 2015, in July, she came here to UW-Madison and UW-Extension.

She’s going to talk to us about one of the most intriguing public policy issues over the course of my professional career in science outreach. And that is the vaccination and anti-vaccine movement. She’s here to talk to us about increasing numbers of parents are refusing to vaccinate their kids. Should public health be worried? And as we’ve been saying the last four weeks, “Yes.” We’re helping you out there.

– [Malia Jones] You’re worried. (laughter)
– [Tom Zinnen] Please join me in welcoming Malia Jones to Wednesday Nite @ the Lab. (audience clapping)

– [Malia Jones] Thank you. Thanks for coming tonight. So what do you think? Are you worried?
– [Audience] Yes.

– [Malia Jones] Anybody not worried? Nobody in this room. So I’m gonna talk as Tom said about parents who don’t want to vaccinate their kids. And who they are. A little bit about myself to get you started. What I study in general is how people sort themselves into geographic space. And so what I’m really interested in in terms of people who don’t want to vaccinate their kids is how tightly clustered are they? In school settings, and in neighborhood settings. And why do we see that kind of clustering arise? I’m gonna focus mostly on California because we have really great public data from California. Although Wisconsin is actually taking some steps to make more data public at the school and school district levels. So I will briefly mention that a little bit at the end. But these trends actually apply nationwide. California is sort of on the forefront of the vaccine refusal trend.

But we’re seeing increasing rates of vaccine refusal nationally. So you can think about these things happening pretty much everywhere. So just a little bit of background to get you started as I’m sure you all know. Starting in 1798 when the Smallpox vaccine was discovered, we have been inoculating kids against infectious diseases and at this time, there’s quite a series of infectious diseases that kids are protected from by vaccines. And just to give you some background info on how effective those vaccines are, I have a couple of charts here that show the rates of different diseases through time in the United States. And then each of these charts shows how those diseases really drop off after the introduction of the vaccine. So, Polio was extremely prevalent in the United States. Especially beginning around in the post-World War II era. It was a huge social and health issue.

How many of you remember the Polio days? Yeah, so when the inactivated Polio vaccine was introduced and then the oral Polio vaccine a couple of years later. Those were required for school entry in many places and Polio rates just dropped, completely bottomed out and now we have no cases of Polio in the United States. We’re working on worldwide eradication of Polio at this time. Similar for Pertussis. In the mid-1940s, a little bit earlier, the Pertussis vaccine was licensed and introduced and Pertussis rates have declined steadily. You can see a little hint of what I’m going to be talking about later here. Pertussis, or Whooping Cough, as it’s colloquially called has actually been increasing again in the most recent years. And then, Measles is another major infectious disease that used to cause hundreds of thousands of cases each year in the United States. And when the series of Measles vaccines was introduced, the rate of Measles infections dropped off precipitously.

And this is just sort of a summary that compares 1900 to 2000 levels for a variety of different infectious diseases. And you can see Measles up there in 1900 caused about 500,000 cases of childhood Measles in the U.S. And today, it’s in a bad year, it’s really just about 500 or so. The levels are so low in fact you can barely even see those 2000 bars. Smallpox having been, and Polio also having been completely eradicated. The vaccines are really effective, but how do we make sure everybody gets them? One of the primary mechanisms for ensuring good population coverage of the vaccine is through school entry mandates. So, in the United States, all 50 states have some kind of requirement that children be vaccinated in order to attend school. The first of those mandates was introduced a really long time ago in Massachusetts, 1855 passed the first school entry vaccine mandate. And today, as I said, all 50 states, DC and Puerto Rico have mandates.

All but three of those states allow some kind of exemption from the mandated vaccines series. And there are two kinds of exemption that are common. Religious exemptions and what’s called a personal belief exemption. Or what you’ll see in a lot of my charts referred to as a PBE, okay? So the three states that are exceptional here are California, West Virginia, and Mississippi. So California is actually really interesting cause it’s the first state in a long time to pass a new stricter vaccine mandate. So last year was the first school year that no personal belief or religious belief exemptions were allowed in the state of California. If you want to be exempted from vaccines in the state of California, you need a medical exemption. Which of course you have to get from a doctor, and typically the doctor has to agree that your child is allergic to the vaccine. Or has some immuno-compromised issue that cannot get the vaccine for medical reasons.

So I’m gonna be looking closely at this policy change and how it has affected the vaccine rates over time. That personal belief exemption in California used to be, before it was eliminated, extremely easy to obtain. And in fact, in Wisconsin, it’s similarly very easy to obtain. All you have to do to get an exemption, this is the immunization record that the school keeps for each child, and a parent who doesn’t want an immunization just fills this thing out, hands it in. A parent who wants an exemption from the vaccine, before the policy change, just turned that same form over and signed the back. I want an exemption from this vaccine mandate. Handed it in and that was the end of the story. And Wisconsin it’s similar. All you do is sign a brief affidavit that says that you do not want to comply with the mandated vaccines.

In 2013, a law was passed in California that made this a little bit harder. Instead of just having to sign the affidavit, you also had to get a note from a doctor saying we have discussed the risks of not vaccinating this child. And then in 2016, the most recent school year, all of the exemptions were eliminated. So there have been two pretty rapid policy changes in a row, in California, in order to try and reduce the number of exemptions from school entry vaccines. And what are we talking about here? These are the required vaccine series. In most states it includes Measles, Mumps, Rubella, Varicella or Chicken Pox, Polio, Human Influenza Type B, Hepatitis A, Hepatitis B, Diphtheria, Tetanus, Pertussis, and then in a few states, very controversial HPV vaccine is also being required. Not in California. So, why is this level of vaccine coverage really important in order to protect the population health? The issue here is this thing called herd immunity. Which I’m sure many of you have heard of before.

But just to catch everybody up, herd immunity is the risk that each vulnerable person has of acquiring an infectious disease. It’s based on their probability of bumping into a person who has the disease. So if you’re in a very crowded environment and everyone is susceptible. And you would choose one person who has the disease, then everybody is gonna be at risk of contracting the disease, right? If you’re in the same crowded environment and most people are vaccinated, or maybe they’ve had the disease before, so they’re not susceptible to catching it, then the probability even with a lot of contact between people, the probability that that infected person is gonna bump into someone who’s susceptible, is low. And that actually prevents the disease from getting enough cases going in the population to create this sort of wave of infection, and cause an outbreak. So herd immunity doesn’t actually protect individual people directly, it protects the whole population. It’s a statistical phenomenon. And this graphic here is sort of, illustrates this point. So, I’ll just go over.

In this top case, no one is immunized. Okay, the blue people indicate a person who is susceptible. They’re not immunized, but they’re healthy. And then the red people here are people who become infected. And you can see that because of this high probability that infected person is gonna come into contact with a susceptible person, the whole population is at risk. This is sort of a mix. Some of the population are immunized. There’s still a high probability that this infected person could bump into a susceptible person. And so the whole population is still at pretty high risk.

And in this case, you have the green people who are immunized but healthy. They are not susceptible to the disease. They’re sort of providing this buffer between the susceptible people and the infectious people. So that the probability that they’ll bump into one another is low. Okay, this is a really a key concept for understanding why it matters so much when a few people don’t vaccinate their kids. There are some other factors besides just the number of people who are vaccinated in what determines how many people need to be vaccinated. And those are sort of summarized under this thing called R not. R not is just a, it’s a summary statistic that if the number of people on average, who will get the disease from each person who has it. So if I were to contract Measles, R0 (R not) not would be the average number of people who get it from me.

And that depends on several things, including how many susceptible people are there in the population. It includes how much contact do they have with each other. Under a quarantine, people have reduced contact with each other. So R0 (R not) can be lowered that way. It also depends on some properties of the disease. Like its mode of transmission. How hard is it to catch? Measles is very easy to catch because it’s an airborne transmission disease. It can hang in the air being infectious for hours after an infectious person has left the room. So it’s very easy to transmit.

A disease like Ebola actually is quite difficult to contract. You need bodily fluid contact to contract Ebola from a person. So that’s one factor. How likely is an exposed person to develop the disease? Some diseases don’t always turn into a full-blown case. And then, the last thing is for how long is an infected person walking around spreading the disease? So it’s a time element to the question. In the United States, with these childhood infectious diseases that are pretty rare, the probability that kid gets sick and goes to the doctor and then the doctor just doesn’t know what they’re looking at and doesn’t diagnose them properly actually plays into this. So they could be walking around with Measles and not know it just because doctors aren’t trained in diagnosing Measles in the United States. So these are all factors that affect– that are sort of disease specific, population specific that affect how much of the population needs to be covered by a vaccine in order for it to be really effective. And produce herd immunity.

So this is a graphic produced by the Guardian. It’s a really neat interactive graphic. And I put the link up there so that you can go look at it in its interactive form. But it gives you a sense of sort of the likelihood of contact and the likelihood of spread of disease under different rates of vaccine coverage. So this top one here says you have 10% vaccination rate. Those are the blue dots. Susceptible are the yellow dots. And then the red dots are the people who have become infected. So you can see a low vaccination rate produces a really high infection rate for Measles.

Somewhere in the 90 to 99% range, you get really good herd immunity for Measles. And infected people are unlikely to give the disease to susceptible people. So because of these variety of factors, different diseases have different thresholds for herd immunity. Polio, it’s 80 to, and they’re also estimates. R0 (R not) not turns out to be quite a hard thing to estimate. Because there all these unknowns. Like we don’t know how many susceptible people that are in the population. Immunity declines with age for some diseases. And sometimes vaccines aren’t as effective as they once were.

And blah, blah, blah. So, there’s a range here. So Polio, 80 to 86%. Diphtheria’s about 85. Mumps somewhere in the 80% range. Rubella, it’s about 84%. Pertussis and Measles are the diseases that you’re hearing a lot about on the news in terms of having a lot of outbreaks. And I think the reason for that is because they have really high thresholds for herd immunity. Pertussis, it’s in the low 90s somewhere.

And Measles, this is the low estimate, is 95%. Some people say it’s more like 98 or 99% because it’s just so easy to transmit. So I think that the reason we see these outbreaks of Pertussis and Measles right now is because we’re sort of bumping up against that herd immunity threshold in some subpopulations in the United States. Where vaccine coverage is declining. And so you can see this in the figures. This graph shows that vaccine coverage and personal belief exemption rates in California Kindergartners. And I have 15, in this case I have 13 cohorts of data. So this is 13 different Kindergarten classes in the state of California. Each class is about 600,000 kids.

And you can see what happens over time here. The pink line is the exemption rate. And so, in 2001, the exemption rate, that’s this axis over here, it was about 1 1/2 percent personal belief exemption rates. And then, each year going right up this graph, and really starting to sort of take a new shape in 2007, 2008 the exemption rate starts to go way up. Until in 2013, it was over 4% personal belief exemption rates in the 2013 Kindergarten Cohort, which is high. So 4% of kids in Kindergarten had some kind of exemption on file. Didn’t have or — Any diseases or were exempted from at least– Any vaccines or were exempted from at least one vaccine. And then you can see if you were wondering well, how do exemptions match up with specific vaccines? You can see that the specific regent vaccination rates were declining at the same time. So this exemption rate is a pretty good metric of the actual vaccine coverage in the population.

And then we saw, we’ve seen a lot of recent outbreaks. Many of you have probably see these on the news. And that’s why you all raised your hands when I asked you if you’re worried. This was a really, really high profile one. The Disneyland outbreak in December 2014. An intentionally unvaccinated person likely contracted Measles in another country where it is still endemic, and then went to Disneyland. And of course, there’s a ton of contact between people at Disneyland, right? And gave Measles to, ultimately, the outbreak spread to 141 people all across the nation, in several states. I was living in California at the time. I actually had a one year old, who was due for his MMR vaccine right in the middle of this outbreak.

And ironically enough, he was late getting his MMR vaccine because I was like, no way, I’m not taking him to into a doctor’s office right now. (laughing) So the Disneyland outbreak was a really high-profile case. It has been in follow up studies, it has been linked to, expressly to low vaccination rates in the population. Though this is an excerpt from a JAMA Pediatric study that shows the estimated effect of reproductive number that’s that R0 (R not) number that I was just talking about, assuming a number of different sort of population level coverage of the MMR vaccine. And what they found was that if we had had really good vaccine coverage in California during the Measles outbreak, it never would have spread to 141 people. Okay, so the fact that it spread as far as it did suggests that MMR coverage is much lower than where we would want it to be. Most recently, I’m sure you heard about earlier this year, there was a large outbreak of Measles in Minnesota. And this particular one was really interesting because it was focused in a small socially-isolated community of Somali immigrants. Who had really low vaccination rates for MMR specifically.

And the disease ultimately spread to I think it’s 71 people. And you can see in this graph right here that the immunization rate for Minnesotan born children, all Minnesotan born children, as opposed to Somali descent children, there’s a huge gap between the immunization rates for this group and for the general population. So there’s something going on in this particular community. They’re a really tightly clustered group of kids with low vaccination rates. And they had this Measles outbreak after a foreign case came in. So here’s sort of a summary. We also have had a resurgence of Pertussis in recent years. And that’s also thought to be connected to declining vaccine coverage. As Tom mentioned in my intro, I once upon a time studied ethics.

And I find this part of the issue really interesting I think for that reason that the vaccine coverage in the population is a great example of this concept called the Tragedy of the Commons. And it’s this concept that when community interest conflicts with individual interest, then we have to produce the interest of the community by putting controls on individual behavior. Most people will choose their own interests over community interest. And so we have to control somehow the behavior of individuals in order to benefit the community as a group. Herd immunity is the common good. Vaccines on an individual basis are not without risks. There are risks associated with getting vaccines. Some of them are real. And some of them are perceived risks.

And I’m going to talk a lot more about perceived risk and how that’s changed in recent years. And so what we see happening is that as disease rates decline following the advent of the vaccination era, those risks of getting the vaccine sort of grow larger in the public mind, right? And as parents weigh the risks and benefits of vaccinating their own child, the perceived risk of vaccination seems sort of outmoded, larger than it used to be. Relative to the perceived benefit to the community. So, I don’t that parents who don’t want to vaccinate their kids are crazy or uninformed or any other sort of pejorative. I think that they’re making a decision based on bad information. And the decision is that the vaccines are somehow riskier than this counterfactual scenario where they don’t vaccinate and they probably will never be exposed to disease, right? But it’s hard I think for people to sort of scale that up and think about the population level consequences of everyone behaving that way. Where you really get into this tragedy of the commons situation. So eventually what happens people make this kind of risk benefit calculation on an individual basis and then herd immunity declines because they’re deciding not to vaccinate. So why are some kids not vaccinated? Some kids are allergic to vaccines.

Particular ingredients of vaccines or the way that they’re produced. They might be immuno-compromised and so they’re not eligible to get a vaccine. There are a number of legitimate medical contraindications. Some kids aren’t vaccinated because they don’t have access to vaccines. And that can include the cost of getting vaccinated. Some people have said to me, “Well, the federal government provides free vaccines.” And that’s true. But they do that at free clinics where you have to wait forever to get them. And not everybody has time to stand around waiting to get their child vaccinated. So there’s also the opportunity cost.

Availability of vaccines has been an issue at various times through history. And then of course, availability of appointments to get your vaccines on time. Some people don’t vaccinate their kids for religious reasons. And I’ll talk a little bit more about that in a minute. And then we have this other group, personal belief refusers or noncompliers. And I think that this can be divided into two subgroups. Whom I like to call naturalists and Libertarians. And I’ll describe each of those in just a minute. So where do these sort of fears about vaccines come from? It’s a much longer history.

There were people who thought vaccines were too risky when the Smallpox vaccine came out. There have always been people who have thought that vaccines are too risky. But in 1998, a former doctor named Andrew Wakefield produced a study. It was a pretty small study. And it was published in the Lancet, so it’s quite high profile, that linked the MMR vaccine to autism. And this was sort of picked up by subgroups within the anti-vaccine movement and distributed widely as truth. And then later on, more recently, it came out that Andrew Wakefield had invented the data he used to make these claims. The study was retracted and he was stripped of his medical license. But it took about 10 years for that to happen.

And I think by that time a lot of the damage had been done. There were some other really high profile celebrities and doctors who also took up this cause that vaccines are too risky and specifically may be linked to autism. And really spread the word that this link exists. Which to be clear with everyone in this room, there is zero evidence that there is a link between MMR and autism. So one of these folks is Jenny McCarthy. She’s been a very outspoken advocate of vaccine choice, or refusing vaccines or eliminating them. Dr. Sears is also a proponent of not entirely missing vaccines, but doing them on a delayed schedule. So the children are, get less vaccines at one time or are exposed to less disease agents at one time during the vaccine process. And so there’s been a lot of high profile, public attention.

That I think really heightened parental concerns around vaccines and helped to produce this sort of inflated sense of risk that parents are using when they evaluate whether or not it’s safer to vaccinate their kids or not vaccinate their kids. And so, the science shows that there are actually a number of different parental concerns that are raised. 68% of parents have no concerns. So, about two in three parents, no concerns about vaccines. But at least quite a few parents who have some concerns, right? 7% of parents are worried about autism with respect to vaccinating their kids. But an even bigger proportion are worried about this concept of over toxicity or immune system overload. So somehow exposure to these vaccines is just too much for their kids’ immune systems. And so they’re worried about the safety of exposing them. 6% of parents say that it’s either healthier to get the disease naturally or the vaccines aren’t as effective as getting the disease naturally.

Or there’s no point in getting the vaccine because the diseases just aren’t that dangerous. So it’s sort of a preference issue. I’d rather just let my kid get Measles than do this vaccine. And then 7%, about an equal number of parents cite concerns about the additives in vaccines including aluminum, mercury, other preservatives and so forth. 9% had two or more of those concerns. And then these parents, this one-third of parents who have any concerns are somewhere between two and 20 times higher risk of using an alternative vaccine schedule. Or foregoing vaccines altogether. So an alternative vaccine schedule, I should back up a little bit, means you don’t do it the way the CDC and the American Academy of Pediatrics recommends. You space out the vaccines.

You do them one by one. You do them over a much longer time window. So instead of over two years, you might do them over ten years. Something like that. So another thing we know about vaccine hesitancy. It is not a binary outcome. So this idea of there being anti-vaxxers. It’s a little bit of a misdirect because it’s not just an either-or situation. Lots of parents have some concerns but go ahead and vaccinate anyway.

Some parents have some concerns about one vaccine, but they’re totally fine with all the other vaccines. They might just delay the vaccines by a little bit. So there’s a range. Parents can be located on a spectrum in terms of how they feel about vaccines. Also, parents, this is a little bit obvious, but like many things, was a surprise to social scientists, parents don’t, their attitudes are not fixed. They evolve over time. So they might have been fine with vaccines for their first child, but their second child doesn’t get any. Or maybe the other way around. Parents change their minds about these things.

Also, a lot of the work has looked at the vaccination status of children when they enter school. But actually this decision making process begins much earlier in pregnancy. So when we think about what do we do about parents who don’t want to vaccinate their kids, we really need to back up and think about much earlier on in the decision-making process than the science has done so far. And then as I said, vaccination itself is not binary. Lots of kids are partially vaccinated or they’re delayed in vaccinations. There’s sort of a range of different outcomes there. And here’s just an article that sort of talks about the different degrees of vaccine hesitancy and classifies them along a spectrum. So we have people who range from I love vaccines. Vaccines are terrific.

Like me and most of the people I went to public health school with. That’s about 33% of the population. There are go along, get alongs. People who just do what the doctor says. They don’t have a strong opinion. Agnostics needs to be convinced and sort of talked through the process. That’s about a quarter of the population. Compromisers who can be talked into doing some vaccines maybe on a different schedule. That’s around 13%.

And then, this is my term, I call them true believers. The people who will not vaccinate their kids under any circumstances. And that turns out to be a really small fraction of people, 2.6%. But when it comes to a disease like Measles, that’s an important fraction, right? Cause we’re talking about that really high threshold for herd immunity. And then there are degrees of responses as I said. Some people who don’t vaccinate, they partially vaccinate. They ask questions. And there’s a fair amount of science out there right now in the medical community about how to deal with a parent who has vaccine concerns. Because this study found that about 85% of providers have a parent who refuses a vaccine each year.

So this is really common for physicians to run into in their panel of patients. There was an interesting issue that came up a couple years ago as this vaccine refusal trend started to crest, what should doctors do about a parent who doesn’t want to vaccinate their child? Should they dismiss them from their practice? Or keep them and keep working on ’em? And this actually became a medical ethics question. And there was some really interesting work on this. The outcome of which by the way, was that they should not be dismissed. They should probably keep them in their practice and continue to talk with them. In part because we found that effective communication can help sort of convince a parent that this is the right way to go. So listening to specific parental concerns. Responding with unambiguous information are all sort of key for a physician-parent relationship to promote vaccination. So who are these folks? Risk factors for being an access noncomplier, someone who can’t access vaccines include health insurance status.

Not having health insurance. Immigration status, especially being an unauthorized resident of the United States. Language barriers. And then also, there’s a tiny bit of evidence out there that people who are migrant workers are less likely to have vaccinated children. And that might be, it’s a little unclear whether that’s their kids are actually unvaccinated or if they just don’t have the paperwork because of their migration status and their residential mobility. And then people who don’t have time. As I mentioned, this is really an under studied issue, but I think it’s an important factor for access noncompliers. We also have religious noncompliers. Risk factors for being a religious noncomplier include having a fundamentalist religious belief and being socially isolated.

So there are groups of people who live in specific religious communities. For example, Amish communities don’t believe in vaccination. And so you’ll have a whole cluster of people living in a pretty small social space that have religious reasons that conflict with vaccination. Generally these folks are more rural dwelling. They are lower income. They have younger age at first birth. They’re lower education. And then we have the personal belief noncompliers. And as I said, I think this can be divided into two different groups.

This is sort of qualitative evidence that I have observed myself. But both of these groups share some characteristics. They’re generally, they have higher education. Older age at first birth. They tend to be affluent and white and nonreligious. In fact, a lot of these people would consider themselves very science minded. And that’s one of the reasons that I say I don’t think they’re uninformed. You know, this is actually a very privileged group of people who are making a decision that we just don’t expect as public health professionals. The naturalist subgroup is politically very liberal and then the Libertarian subgroup is smaller, politically very conservative, and very understudied.

In fact, this is just me talking right now. There’s no other science out there on this, okay? (laughing) So the idea, this is really a story about ideology. And the ideology that leads a person to make a decision about vaccinating their child. I call these two groups the naturalists and the Libertarians because it’s not an isolated decision when they make a decision about vaccination. That decision fits within a broader set of ideas about the way they live their lives and the way they see the world. And the naturalists in general believe that vaccines, that the safest and best way to raise a kid is the natural way. And somehow vaccines have been sort of classified as unnatural in that ideology. Libertarians, it’s a really different ideology that leads them to the same place. And that is that the government should not have this authority over me.

I’m not going to vaccinate my kids because the government told me to. All right, so it’s quite a different sort of set of ideas that leads people there. And we can maybe see this if we look at political maps. This is not the same thing, but we’re talking about ideologies. So it’s a related concept. And there is variation in California in terms of how ideologies pattern over space. So we can get some hints about maybe what’s going on with these patterns of anti-vaccination in California. So you can see in California, there are a lot of Democrats in California. They are clustered on the coasts and in the major cities.

Republicans in California tend to be in the central valley, sort of the eastern half of the state. This next set of maps is probably more informative. These are people who voted Libertarian. Not the Libertarians I’m talking about with not vaccinating their kids. You can see that they are clustered in these really rural counties on the eastern side of the state. And then the northern coast. Green Party voters were really strongly clustered on the coast. Especially the northern coast. So there are a lot of things we don’t know about vaccine hesitant parents.

We have some. There was one study that suggested telling them about the risks of contracting diseases actually makes them less likely to vaccinate their kids. Traditional public health messaging totally ignores this variation between groups. Like the difference between people who don’t want to vaccinate because of religious belief and people who don’t want to vaccinate because they don’t like the government telling them to. And we have some lessons from market segmentation. I think we really need to understand these differences in order to target public health messaging more effectively to these various groups. Especially that naturalist group. We really I think need to better understand how people come to see vaccines as inconsistent with the rest of their parenting ideology in order to get them on board with vaccination. And the approach that’s actually been used in California and elsewhere is policy.

As I said at the beginning. We’ve had this series of policy change in California to try and stem the tide of vaccine refusal. And I think that’s gonna work for some parents. But other parents are really willing to go the distance to not vaccinate, right? There’s a commitment that is not gonna be stopped by a policy change. So the problem, vaccine rates have been falling sharply. Those rates are variable in space. We see geographic clusters of people who don’t vaccinate their children and these pockets, these spatial pockets are subgroups of people who have really poor herd immunity. Deeply compromised herd immunity. In schools and in neighborhoods and in some cities.

And that is a serious threat to herd immunity. We’re gonna see disease outbreaks in situations like that as we have in the last few years. So here’s a map that shows the personal belief exemption rate by county in California. And you can see that pattern of the different ideologies that I was talking about there. You have some counties in California with exemption rates as high as 15%. Especially along the northern coast here and Santa Cruz county. Not only that, we have school level data on these kids. And my colleague Allison Buttonheim and I have found that there is tremendous school level clustering. So those same ideologies that lead parents to make a decision about vaccinating their kid, also lead parents to choose a school, right? They choose a neighborhood or a school based on their parenting beliefs.

And so we see these really extreme rates of vaccine exemption in certain schools and types of schools in California. And we measure these things by two measures. One measure is clustering or the probability that an unvaccinated child is gonna bump into another unvaccinated child in their own school. And the other measure is exposure. That’s the probability that a vaccinated child is gonna bump into an unvaccinated child in their school. And that’s important because vaccines don’t always work. And so even vaccinated kids are at some risk when they encounter an unvaccinated kid. So this figure just shows the degree of current personal belief exemptions is just the number of personal belief exemptions by county in California. And then this is the aggregation index.

That measure of clustering. And you can see that although these larger population counties like Los Angeles and San Diego County don’t have a ton of clustering, there is really tremendous clustering in some of those high-risk places. And this is just an excerpt from the paper where we argued that norms around vaccination and securing exemptions are likely to be shared in these communities. And these ideas defuse through parent social networks before and after the Kindergarten enrollment process. So parents are getting the idea that they’re not gonna vaccinate their kids from one another. It’s a very sticky idea. And we don’t thoroughly understand why it spreads so easily. So here are the policy changes. As I said, the first one AB2109 required a physician’s signature for a personal belief exemption.

It allowed a free religious exemption. So the religious exemption was still really easy. You just turned the form over and signed it under that law. SB277 eliminated all personal and religious belief exemptions in California. There is still a medical exemption in California. Wisconsin and many other states have actually started to respond to these same issues by proposing some policy change. Wisconsin, a colleague of mine in here at UW, who works in pediatrics infectious disease, helped draft a bill that would have eliminated the personal belief exemption here in Wisconsin as well. And the bill was never heard on the floor but it’s sort of an ongoing effort to promote policy change. So what do we think happened in terms of prevalence in clustering of exemptions after these laws passed? Well this shows the pre exemption level of clustering and rates of personal belief exemptions.

This heat map gives you a sort of more refined sense of where the real hot spots are in terms of clustering. And then we asked this question. Was recent California policy change effective? And was it effective especially at reducing these clusters? That’s what’s really important for herd immunity. So we used data, as I said there’s great public data from California. We have, it’s actually 16 cohorts of data now from 2001 to 2015. For every single Kindergartner. So it’s not a sample. It’s all of them, right? This is super confusing to reviewers of my papers. They’re constantly giving me a hard time about this.

We do have some really small population counties where one kid can make it appear that the exemption rate spiked for one year and then went back down again the next year. But that’s because the denominators are quite small. So you do see some noise in the small population counties. We do the analysis at the county and the school levels. And then the caveat here, are some kids with a personal belief exemption actually vaccinated? Yes, some of them are. Some of them are partially vaccinated. Some of them are completely vaccinated and their parents didn’t happen to have the paperwork when they came in to enroll their kid. But we do know that they are as a group, under vaccinated. So this is a high-risk group.

Even if it’s not sort of the maximum risk group. And what did we see after these two policy changes? This is the trend in personal belief exemptions over time that I showed you before. But now it’s got a couple more years of data tacked on here. After AB2109, there was a big drop in personal belief exemptions because it got a little bit harder to get those exemptions. That policy was in place for two years. And then SB277 took effect and the exemption rate for California went all the way down to a historic low level. So, yes. This policy was quite effective. What do we know about clustering though of that 0.6%? ‘Cause that could still be really important if those parents who are willing to go the distance are gonna go out and find a way to get an exemption. One of the ways that we see that they did that was that the rate of medical exemptions has been ticking along at about 0.1% for 20 years.

And last year it jumped up to half a percent. So there were some parents who went and got a medical exemption from a doctor after SB277 was passed. What happened with clustering and exposure measures? Actually, not very much. The policies did not change clustering very much at all. The most recent data year is not on there. The 2016 data, because California changed the way that they mask some of their data. And so the public data, you actually can’t calculate this measure from the public data anymore. So, I’m waiting for them to give me the unmasked data to tell you what happened with clustering and exposure in the most, after SB277. After assembly bill 2109, clustering did not change very much.

So we do see this pattern of parents who were already clustered in schools, kids who were already clustered in schools, and had low rates of vaccination were still clustered in schools. So we still have some concerns about herd immunity there. What’s sort of the upshot here? Policy did have a really substantial impact. Probably did not fix this problem of parents resisting vaccines. There’s a substantial fraction of noncompliers who took the time and effort to go get that healthcare provider signature on their exemption form before the exemptions were eliminated. And now we’re seeing there was a substantial number of parents who went and found a doctor who would agree that there’s a medical contraindication for vaccinating their child. It’s a small fraction, but as I say, with Measles and Pertussis, those small fractions can be really really important. We also saw a substantial fraction of people took that free religious exemption under the first version of the policy change. And yeah, clustering of exempted kids was not greatly impacted.

Now, when you look at the Wisconsin data and the Wisconsin proposed policy change, we do see substantial clustering in schools for Wisconsin vaccine rates. So there are reasons to think that in Wisconsin there’s also substantial clustering. There’s also pockets of places that have really compromised herd immunity. Especially on these high risk diseases. And the bill that was proposed to amend this would not have eliminated vaccine exemptions in Wisconsin. It would have like that first California bill, it would have just made ’em a little harder to get. So the policy approaches to this I think are, it’s a step in the right direction. I don’t think it’s gonna fix things for the most high risk groups in the population. So a few limitations.

We don’t know very much about the extent to which personal belief exemption kids are vaccinated. We do know that they are more likely to be under vaccinated than the general population. So there’s some sort of wiggle in what the real risk of outbreak is here when you talk about exemptions versus you look at a kid’s medical records and you know for sure whether they got a vaccine. As I said, the Kindergarten enrollment it’s not the right time for measurement or intervention. By the time a kid is entering Kindergarten, they’re five years old. They should have completed this vaccine series when they were two. So it’s just sort of not the right time to measure. And then we really need to know something about these parents and what they’re thinking when they evaluate the risks and benefits of vaccinating their kid. And we know almost nothing about that.

So that’s sort of the future direction of my work. I’m gonna try and work with some survey data and figure out if I can tease out these different ideologies and what the context of people’s decision making process really is when they decide to vaccinate or not. So here are a few conclusions. The highest risk areas still have significant clustering of unvaccinated children. And yeah, we have a lot more work to do to understand how to address these vaccine hesitant parents and what is driving the trend of increasing vaccine hesitancy in California and in Wisconsin and in the U.S. You also see these trends happening and playing out in Europe. I’m sure you’ve seen the news about recent outbreaks there. So my future directions. I want to know what’s happening in these high-risk areas.

I’m trying to access the Wisconsin immunization record data to get some personal level information about the timing of when kids are vaccinated in Wisconsin and maybe even spatial information about how clustered they are. And then I’m really interested ultimately in looking at implications of this work for HPV vaccine. Which has a sort of a whole different ball of issues wrapped up with it. But it is likely going to be the next mandatory vaccine coming along in many new states. And ultimately what I want to do is understand how to better target and communicate with high risk parents with an effective intervention. And I’m using some new methods to do this including this really cool thing called agent-based modeling. Where you set up a simulated population of people and you can give them some behavior rules and then tell them to go interact with each other. And you can see the patterns that emerge. And this will give me the ability to test my little hypothesis that says the idea of vaccinating being too risky is actually something that moves from parent to parent through a community.

And that idea itself is very sticky. And that I think would help us understand how to better intervene and talk to these parents. So should public health be worried? Yes. Public health should be worried. This is a big problem. We’re gonna see increasing numbers of these kinds of outbreaks like the one we saw earlier this year in Minnesota. Especially in these high risk communities where there are a lot of unvaccinated kids. If you have kids or grandkids who are in a school, you can actually look on the Wisconsin Department of Public Health data at the school level vaccination rate where your kids go. So if you’re curious what the risk environment is in your kid’s or your grandkid’s school, you can go look at that up on the Department of Public Health website. Okay.
(audience clapping)