Vaccine Hesitancy: Evidence from an Adverse Events Following Immunization Database, and the Role of Cognitive Biases

"Distrust in vaccine safety and AEs and the lack of effective communication about vaccine information demand reconstructing vaccine communication."

Among factors contributing to vaccine hesitancy, concerns regarding vaccine safety and adverse events (AEs) play a leading role. An emotionally compelling story about a rare AE might cause parents to perceive that rare incident as a frequent AE; in that case, it is a cognitive bias that might nudge them toward vaccine hesitancy. This study addresses concerns regarding vaccine AEs by analysing reported AEs in the United States (US)' Vaccine Adverse Events Reporting System (VAERS) from 2011 to 2018, noting that such databases can be used as a source for evidence-based communication about vaccine safety and AEs. The study also identifies the potential cognitive biases connecting vaccine hesitancy concerns to vaccine-hesitant behaviours and details the mechanism they trigger in the vaccine decision-making process. The goal is to support healthcare providers and officials in their vaccination communication efforts.

VAERS is a passive spontaneous reporting system for vaccine safety monitoring whose primary purpose is safety signal detection and hypothesis generation about AEs following immunisation (AEFIs). VAERS is open to everyone to report voluntarily any medical incident after immunisation, even if the reporter is not sure if the vaccine caused the incident. From January 1 2011 to December 31 2018, VAERS received 293,609 reports. The paper provides a summary of data gathered during this timespan, such as population-based AEFI reports, vaccine types, and serious AEs (SAEs). These data can be used as a communicable summary (of the non-severity of the AEs: 94.5% of the reports are not serious, and the average population-based serious reporting rate over the 8 years was 25.3 reports per 1 million population) to increase trust in vaccines and boost vaccine acceptance. Also offered is an interactive dashboard for healthcare professionals to use in vaccine communications with vaccine decision-makers (patients/parents).

Even if open, this communication between health practitioners and vaccine decision-makers is not free from challenges. One challenge, particularly when the information is about vaccine safety and AEs, is the role of cognitive biases. At the individual level, people decide about vaccination based on several factors, including but not limited to their prior beliefs, perceived risk, perceived ambiguity, perceived loss, the message they receive, and factors like emotion and the vividness of the message. All these factors instigate cognitive biases that might nudge people toward sub-optimal decisions. Clinicians and other healthcare professionals are also influenced by these biases, affecting their vaccine recommendation.

A review of 29 relevant articles led the researchers to a core list of 13 cognitive biases; separate research on the topic of prevalent cognitive biases on social media led to two additional cognitive biases, shared information bias and false consensus effect, that can affect and direct social media users' behaviours. (See Table 1 in the paper for a summary list of the 15 cognitive biases that affect vaccine hesitancy.) The researchers categorise the list into three categories: Group 1: Cognitive biases triggered by processing vaccine-related information; Group 2: Cognitive biases triggered in vaccination decision making; and Group 3: Cognitive biases triggered by prior beliefs regarding vaccination. These groupings can be aligned with the Precaution Adoption Process Model (PAPM), which is a stage-based theoretical model to explain how people decide to undertake a health-protective behaviour and how they transform that decision to action through 7 stages: 1. Unaware of the issue; 2. Unengaged by issue; 3. Undecided about acting; 4. Decided not to act; 5. Decided to act; 6. Acting; and 7. Maintenance. As this paper focuses on vaccine hesitancy, which results in not vaccinating (Stage 4. Decided not to act), the researchers explain the cognitive biases based on stages 1 to 4.

Here are just a few examples:

• Group 1: Cognitive biases triggered by processing vaccine-related information - For instance, availability bias is the tendency to attribute higher weight to factors that are easier to recall. Media coverage of a rare SAE report that offers a vivid and emotionally compelling anti-vaccination message, likely to be recalled during decision making, could cause people to overestimate the probability of an AEFI. The same rule applies to detailed reports in VAERS. (The media (message) has the highest effect in moving people from the PAPM's Stage 1 to Stage 2, and from Stage 2 to Stage 3.)
• Group 2: Cognitive biases triggered in vaccination decision making - During the PAPM's Stage 3, when people are in the midst of vaccination decision-making, they can fall prey to: omission bias, ambiguity aversion, loss aversion, optimism bias, present bias, and/or protected values.
• Group 3: Cognitive biases triggered by prior beliefs regarding vaccination - At Stage 4 of the PAPM, when people have a prior belief about a topic, their formed opinion (decided not to vaccinate) will affect their reaction to new information and arguments. On social media, people's behaviour affected by these biases might create polarised groups of like-minded people, i.e., echo chambers, where they are at risk of other social-level biases.

The researchers suggest that, based on the categories provided in this paper, public health officials can customise their plans, interventions, and other forms of communication to hinder the impact of identified cognitive biases and to increase vaccine trust and acceptance. Specific suggestions include:

• Both anti-vaxxers and vaccine advocates can use Group 1 cognitive biases. Public health officials should craft campaigns to decrease the effect of these cognitive biases when/as they (frequently) appear in anti-vaccine content. Such efforts could increase the availability of vaccine-advocate information and emphasise the majority of mild AEFIs. By using some of the biases like the framing effect and authority bias, public health officials could improve vaccine trust. On social media, correction of health-related misinformation by an authority such as the US Centers for Disease Control and Prevention (CDC) may be effective (an example of authority bias).
• Public health officials should consider Group 2 cognitive biases in all their plans. They should try to decrease perceived levels of uncertainty, ambiguity, and feeling of loss about the result of vaccination in vaccine decision-makers. The summarised data of the VAERS AEFI reports could be of help here to communicate evidence-based information about vaccine safety and AEs to decrease the influence of those factors.
• Alleviating the effect of Group 3 cognitive biases that vaccine-hesitant people hold requires careful attention, as direct debunking or refuting of misinformation might backfire.

In conclusion, this paper has provided a summary of VAERS that health practitioners can use to provide evidence-based vaccine information to vaccine decision-makers (patients/parents) and to mitigate concerns over vaccine safety and AEs. In addition, by paying attention to 15 potential cognitive biases that might affect the vaccination decision-making process and nudge people toward vaccine hesitancy, health officials can develop or modify their plans, interventions, and/or messages to increase vaccination uptake so as to decrease the effect of these potential cognitive biases.