Parents’ vaccination comprehension and decisions

Summary

We report on 30 in-depth mental models interviews with parents discussing vaccination for their children, both in general terms and in response to communications drawn from sources supporting and opposing vaccines. We found that even parents favourable to vaccination can be confused by the ongoing debate, leading them to question their choices. Many parents lack basic knowledge of how vaccines work, and do not find the standard information provided to them to be particularly helpful in explaining it. Those with the greatest need to know about vaccination seem most vulnerable to confusing information. Opportunities for education may be missed if paediatricians do not appreciate parents’ specific information needs.

Introduction

Vaccines have substantially reduced, and in some cases essentially eliminated, the risk of previously common childhood diseases [1]. Current vaccination rates in the United States are high enough to achieve herd immunity for many diseases, protecting even those children who are not vaccinated [2]. However, some parents’ decision not to vaccinate their children threatens to undermine this success. Recent drops in measles-mumps-rubella (MMR) vaccination rates in the UK and Japan [3] have led to a resurgence of preventable diseases [4], [5], [6], [7]. This pattern recalls previous periods of vaccine mistrust [8], leading to reduced vaccination rates [9] and disease outbreaks [10], [11].

Some failures to vaccinate are oversights. Parents often think that their children have more current vaccinations than medical records indicate [12]. Other failures reflect socioeconomic barriers, associated with low income, multiple children, and health care divided across providers [13], [14].

In other cases, though, not vaccinating reflects conscious decisions [13], [15]. Some parents object on religious or personal grounds [16]. In rare instances, children are exempted due to medical contraindications [17]. The trade-off between long-term benefits and short-term costs, such as crying or pain, is acknowledged even by parents who have their child vaccinated [18]. Two beliefs that predict decisions not to vaccinate are that vaccines (a) are unsafe when administered [13] and (b) can have long-term negative health effects [18]. These concerns also predicted vaccination decisions in a prospective study [19].

Vaccination is also less likely among parents who see fewer benefits. Some parents believe that they will be able to protect their child from disease without immunisation [20]. Others see little disease threat, ironically, because of vaccines’ success in reducing disease incidence [21], a trend seen in previous periods with anti-vaccination movements [22]. Some parents say explicitly that the shot is riskier than the disease [23]. Others seem motivated by omission bias, whereby they would feel worse if their child were injured from their action (getting the vaccine) than from their inaction (not getting it) [20], [24], [25].

Public concerns about vaccines are not new. Early vaccines, such as smallpox in England [26] and polio in the United States [27], faced strong opposition. As a result of these protests, parents in Britain won the right to decline vaccination [28]. Activists lobbied for acellular pertussis vaccine [29], which is now the standard, even though there was never strong evidence that the whole-cell version was risky [28]. Despite continued improvements in vaccine administration safety [30], the global perception that vaccines are risky is growing [31].

Some of those fears seem to be grounded in evidence that is incompletely understood, such as early, small sample studies suggesting that the MMR vaccine can cause autism [32], [33], [34]. Many subsequent studies, using more rigorous methods, failed to substantiate this claim [35], [36], [37], [38], [39], [40], including a large epidemiological study with the statistical power to reveal even a small MMR-autism correlation [41]. Research into possible links between neurological conditions and thimerosal, a preservative used in some vaccines, has not shown consistent, replicable relationships [42]. Nonetheless, negative reports persist.

One form of negative publicity concerns adverse events among individual children who have received vaccines [43]. A second form is investigative journalism, like a widely read essay printed in Rolling Stone magazine and Salon.com criticizing thimerosal, citing environmental contaminants and questionable government motives [44]. Such accusations can spread widely on the Internet [45], potentially resonating with parents holding anti-vaccination beliefs [22]. Alternative medicine practitioners, such as those using holistic and homeopathic approaches, often recommend against vaccination [46], [47]. The impact of homeopaths’ recommendations may be amplified by their willingness to give immunisation advice by e-mail more often than physicians do [48]. School officials who advocate alternative medicine or hold common misconceptions about vaccination are also more likely to advise parents to opt out or to grant exemptions [49]. Parents who favour alternative medicine are less likely to vaccinate their children [15].

Parents’ trust in healthcare providers makes them potentially valuable educators [50], with opportunities that arise frequently, as when they encounter parents raising safety concerns [13], [51]. However, physicians may miss these opportunities due to fear of litigation [45] or preference for referring sceptical patients elsewhere [51]. And some patients are frustrated that their practitioners seem to lack the time to talk [52].

Although there has been considerable research into factors that predict parents’ perceptions of risks and benefits, relatively little attention has been paid to parents’ underlying conceptual model of vaccination and its role in their interpretation of vaccine-related communications. Flawed mental models may increase susceptibility to anti-vaccination campaigns, which often rely on flawed logic [53]. Without knowing parents’ mental models, it is difficult to develop effective communications, for delivery by providers or other sources. Under the time and conversational constraints of modern medical practice, doctors may not understand parents’ concerns about vaccination. The research reported here presents a more comprehensive picture of these concerns, based on in-depth interviews with American parents who have faced the prospect of vaccinating their children. We examine their general understanding of vaccines and response to new information.

We adopted a mental models approach to identifying parents’ patterns of beliefs regarding vaccinations. This approach [54] builds on a legacy of related approaches in cognitive psychology, tailored to the demands of understanding different phenomena [55], [56], [57], [58]. All mental models approaches begin with a formal analysis of the domain, followed by open-ended interviews structured around that analysis. Subsequent studies may include structured surveys suited to administration to large samples for estimating the population prevalence of beliefs, as well as the creation of interventions that improve lay understanding and decision making [59], [60]. The research reported here takes the first two steps, characterizing the phenomena that parents must understand (vaccine risks and benefits) and then describing parents’ mental models of these processes, as revealed in open-ended interviews.

We have applied this strategy to topics as diverse as domestic radon [61], adolescent sexual behaviour [59], mammography [62], climate change [54], electromagnetic fields [63], carbon dioxide sequestration [64], paint stripper [65], nuclear energy sources in space [66], breast implants [67], and cancer [68].

In the first step, we use a standard approach to characterizing complex, open-ended systems, creating an integrated assessment of decision-relevant knowledge. It uses the formalism of influence diagrams [69], [70], in which nodes represent variables and links represent relationships. When an arrow connects two nodes, knowing the value of the variable at its tail helps to predict the value of the variable at its head. Like other models, influence diagrams are incrementally refined. At all stages, the aspiration is a computable model, in the sense of working toward a sufficiently precise representation that would allow outcomes to be predicted if data requirements were met [71]. Whether actual computation is pursued depends on the application. Here, an intermediate level of formalism seemed best suited to ensuring that the full set of potentially relevant issues was represented, without unduly favouring those issues that are more readily quantified.

The second step uses the integrated assessment to characterize the target audience's decision-relevant beliefs. Open-ended interviews are used so that people can address any issue on their minds, in their natural mode of expression. The interviews begin very generally, and then ask if the interviewee has anything to say about each topic in the model. Such interviews lay the foundation for the content and wording of structured surveys and interventions.