One more time: Vaccine refusal endangers everyone, not just the unvaccinated

One of the more frequent claims of antivaccine activists often comes in the form of a disingenuous question. Well, maybe it’s not entirely disingenuous, given that many antivaccinationists seem to believe premise behind it. The question usually takes a form something like, “If your child is vaccinated, why are you worried about my children? They don’t pose any danger to you.” Of course, the premise behind that question is, ironically, one that conflicts with many of the beliefs behind antivaccinationism, in particular the belief that vaccines are ineffective. Yet, the premise behind this question is that vaccines are so effective that there’s no reason for the parents of a vaccinated child to be concerned if that child comes in contact with another child with a vaccine-preventable disease. Of course, no one ever accused antivaccine activists of being consistent in their beliefs.

Of course, another claim that antivaccinationists like to make is that it isn’t the unvaccinated who are causing outbreaks, but the vaccinated. To make this argument, they like to point out that most of the infected in an outbreak are vaccinated, which is, of course, not uncommonly true. This is, of course, a profoundly mathematically ignorant line of argument because it neglects how small the number of unvaccinated children usually are relative to the vaccinated. Raw numbers mean little. What really needs to be examined is the relative risk of infection of the unvaccinated compared to the vaccinated during an outbreak, and, depending on how effective the vaccine is, that relative risk is usually rather high. For instance, for pertussis, being unvaccinated is associated with a 23-fold increased risk of infection.

The fact is, not vaccinating children endangers them, and just this week yet another study was published that finds yet the same thing again. However, given how often antivaccine loons keep repeating their misinformation that their choice doesn’t just endanger their children but everyone’s children, it’s always good to see a new review in a high impact journal like JAMA confirming just that. This time, it’s a systematic review of the evidence for measles and pertussis by Phadke et al entitled Association Between Vaccine Refusal and Vaccine-Preventable Diseases in the United States: A Review of Measles and Pertussis. The senior author was Saad B. Omer, MBBS, MPH, PhD at Emory University. I would have discussed this one yesterday had I not, as I mentioned yesterday, crashed hard on the couch the evening before.

Basically, noting that vaccine refusal has been associated with outbreaks of invasive Haemophilus influenzae type b disease (Hib), varicella, pneumococcal disease, measles, and pertussis and that over the past 20 years, rates of nonmedical exemptions have steadily increased, the authors reviewed the existing medical literature to characterize the relationship between vaccine refusal and the epidemiology of measles and pertussis, vaccine-preventable diseases with recent outbreaks in the United States. They chose their time periods thusly: since measles was declared eliminated in the United States 16 years ago and since pertussis reached its lowest point of its incidence (after 1977). Concentrating on studies that examined risk of disease in the unvaccinated and vaccinated, they also looked at vaccine delay and exemptions, including medical and non-medical (i.e., philosophic or religious) exemptions in order to determine how vaccine refusal affects risk of disease in both the unvaccinated and the vaccinated. As a result, they were able to estimate, for example, that over half of the cases in US measles outbreaks are unvaccinated, often intentionally.

In their search, the authors identified 18 published measles studies (9 annual summaries and 9 outbreak reports). These studies described 1,416 measles cases ranging in age from 2 weeks to 84 years of age, with 178 of them younger than 12 months. Of these cases, a total of 199 cases (14%) were people with a history of being vaccinated against measles, while more than half of the total measles victims 804 (nearly 57%) had no history of measles vaccination. There were 970 measles cases with detailed vaccination data, of which 574 were unvaccinated, and, of these, 405 (71%) had nonmedical exemptions, making up 42% of the total number of cases). One particularly pertinent observation is how the unvaccinated predominate among cases early in the outbreak:

The outbreaks evaluated in the cumulative epidemic curve included cases that occurred up to 5 generations of spread after the index case, with the latest related case occurring 12 weeks after identification of the index case. When viewed by week of outbreak, unvaccinated individuals constituted a larger fraction of the total measles cases per week in the earliest weeks of an outbreak (eg, earlier generations).

So basically, most of the measles cases were in the unvaccinated, and the majority of the unvaccinated were old enough to receive the vaccine and without any medical contraindication to being vaccinated. Their parents had refused the vaccine for nonmedical reasons. Yes, being antivaccine causes harm, and existing studies allowed the authors to estimate how much these vaccine refuseniks increase the risk in the whole population.

Reviewing the relative risk of measles in unvaccinated children, the authors found studies demonstrating that the unvaccinated were anywhere from 22- to 35-fold more likely to contract the measles during an outbreak. Worse, higher rates of vaccine exemption in a community were associated with greater measles incidence in that community, among both the exempt and nonexempt population. Curious, I went back to look up the article cited by Phadke et al, which used mathematical modeling to estimate that, depending on assumptions of the model about the degree of mixing between exemptors and nonexemptors, “an increase or decrease in the number of exemptors would affect the incidence of measles in nonexempt populations. If the number of exemptors doubled, the incidence of measles infection in nonexempt individuals would increase by 5.5%, 18.6%, and 30.8%, respectively, for intergroup mixing ratios of 20%, 40%, and 60%.”

So, yes, basically this tells us what we already know, namely that an increased proportion of unvaccinated children does degrade herd immunity and does increase the risk of disease in the vaccinated. Remember, no vaccine is 100% effective. The MMR is very effective against measles, over 90%, but not 100%. Yes, the vaccinated can still be infected; it’s just that they’re much less likely to be.

As far as pertussis goes, the numbers aren’t good either for the unvaccinated. The authors identified 32 reports of nonoverlapping pertussis outbreaks covering 10,609 cases among individuals ranging in age from 10 days to 87 years. The five largest statewide pertussis outbreaks had substantial portions of vaccinated or undervaccinated. Part of the problem that complicates the pertussis picture is, of course, the problem of waning immunity, but it’s clear with pertussis as well that being unvaccinated carries with it a substantial increased risk of developing the disease:

Three studies evaluated the individual risk of pertussis associated with vaccine refusal—1 retrospective cohort study used Colorado pertussis surveillance and immunization data from 1987-1998 and determined that those with exemptions were 5.9 times more likely to acquire pertussis compared with fully vaccinated individuals. A different case-control study analyzed pertussis cases from 1996-2007 within a large managed care organization and computed a nearly 20-fold increased risk of pertussis among individuals with exemptions—11% of the pertussis cases in that cohort were attributed to vaccine refusal. Another case-control study used pooled longitudinal data (2004-2010) from 8 Vaccine Safety Datalink sites and determined that even undervaccinated individuals had an increased risk of pertussis, with the risk being proportional to the number of missed doses of DTaP.

As is the case with measles, high rates of vaccine exemption in a community or state are associated with an increased risk of pertussis in that community or state compared to communities or states without high rates of exemptions. Worse, as was the case with measles, the risk of being infected with pertussis is higher even among the appropriately vaccinated. The authors also noted that the geospatial association between clusters of vaccine refusers and pertussis cases can’t be explained by waning immunity for the simple reason that there shouldn’t be geographic heterogeneity in the duration of protection offered by vaccines against pertussis.

Overall, the authors concluded that vaccine refusal is associated with an increased risk of vaccine among both the unvaccinated and vaccinated and that, although waning immunity to pertussis is an issue in pertussis outbreaks (as I’ve discussed before), there is still a significant contribution in some populations due to vaccine refusal.

The authors observed:

This review has broad implications for vaccine practice and policy. For instance, fundamental to the strength and legitimacy of justifications to override parental decisions to refuse a vaccine for their child is a clear demonstration that the risks and harms to the child of remaining unimmunized are substantial. Similarly, central to any justification to restrict individual freedom by mandating vaccines to prevent harm to others is an understanding of the nature and magnitude of these risks and harms. However, the risks of vaccine refusal remain imperfectly defined, and the association between vaccine refusal and vaccine-preventable diseases may be both population- and disease-specific. Vaccine refusal–specific strategies to optimize vaccine uptake could include state or school-level enforcement of vaccine mandates, or increasing the difficulty with which vaccine exemptions can be obtained.

Exactly. Depending on the vaccine and disease, the most potent one-two punch argument used by antivaccine activists, one that is often heard sympathetically among people with ideologies that take a dim view of the government and government regulations, is shown to be false. I’m referring, of course, to the claim that parents’ freedom and right to raise their children as they see fit shouldn’t be limited by vaccine mandates because their unvaccinated children are harming no one. For pertussis and particularly for measles, at least, this is clearly not true.

As is often the case with major articles like this, there was an accompanying editorial, in this case by Matthew Davis at the University of Michigan. Davis first notes that, in the case of pertussis, waning immunity and vaccine refusal are different challenges, but they are related. The reason is that nonmedical exemptions for childhood vaccination decrease overall community immunity and thus increase the risk of infection for children with waning immunity or, in the case of the children of vaccine refusers, no immunity at all. Outbreaks then occur, and these outbreaks provide “examples” that antivaccine activists can point to claim that the benefits of vaccination are being oversold and therefore not important for their children.

Davis also notes:

An important priority is to ensure high reliability in US vaccination efforts. Current US vaccination efforts are not optimally effective, as measured by outbreaks of vaccine-preventable diseases and vaccination coverage rates that fail to reach target levels. Currently, no single entity is accountable for monitoring and coordinating the multiple stakeholders with interests in maximizing vaccination rates. These multiple stakeholders include parents, physician practices, private insurance, public health institutions, community pharmacies, and government agencies. Given the public health importance of effective vaccination, a more reliable system is needed.

The airline and nuclear power industries have established a culture that values consistent and standardized practices to promote highly reliable performance. In the United States, efforts to achieve complete vaccination rates in the population do not follow the standards established by these industries. By standardizing procedures and continuously evaluating the effectiveness of new initiatives to increase vaccination rates, it may be possible to reduce exemptions and waning immunity and achieve more complete vaccination of children and adults.

Exactly. Davis almost drolly notes at the end that “without a centralized infrastructure focused on the goal of maximizing community immunity, high-reliability vaccine coverage remains challenging in the United States.” That’s an understatement. The infrastructure in this country for tracking vaccination rates could use considerable improvement. It’s a patchwork of state systems, some of which do a good job, some of which do not. In some states school-level vaccination rates are reported; in others not. Unfortunately, because it is states that are responsible for setting vaccine requirements, this is not a situation likely to be improved much any time soon.

Still, the take home message of this review article needs to be repeated over and over again. Despite what antivaccine parents claim, their choice not to vaccinated does impact more than just their children and themselves. It impacts the entire community in which they live negatively, even the vaccinated.