key: cord-0970646-o0k7eugb
authors: Chevallier, C.; Hacquin, A.-S.; Mercier, H.
title: COVID-19 vaccine hesitancy: Shortening the last mile
date: 2021-02-09
journal: Trends Cogn Sci
DOI: 10.1016/j.tics.2021.02.002
sha: 36d7b4b7c082025a4dd250e1b5d23ec1e562477e
doc_id: 970646
cord_uid: o0k7eugb

We offer three recommendations to increase COVID-19 vaccination rates. First, use communication campaigns leveraging evidence-based levers and argumentation tools with experts. Second, use behavioral insights to make vaccination more accessible. Third, help early adopters communicate about their decision to be vaccinated to accelerate the emergence of pro-vaccination norms.

However, proving that a COVID-19 vaccine is safe and effective can only be a turning-point if uptake is high enough to stop the circulation of the virus. According to models with a vaccine efficacy of 80%, the percentage of the population that needs to be vaccinated to reach herd immunity ranges from 75 to 90% (depending on factors such as the basic reproduction number R0, vaccine-induced immunity duration and whether vaccines prevent transmission) [1] . Current vaccination intentions are far from allowing this goal to be met. A significant fraction of the global population reports they are unwilling to get vaccinated, and while vaccine hesitancy fluctuates, it remains dangerously high in many countries i . As a result, and in spite of recent improvements in some countries, rates of vaccination intention remain below the percentage required for herd immunity in most countries ii .

This raises the serious possibility that a significant chunk of the effort directed towards COVID-19 vaccine development might end up being wasted in the last mile. Rolling out efficient public health policies requires attention to first and last mile issues. The first mile is about identifying the best solution, the "what": What is the best prophylactic or therapeutic solution? What is the most efficient vaccine? The last mile is about developing the best strategy to ensure the solution reaches its target, the "how": How do we convince people to accept the treatment? How do we make treatments accessible? If the last mile is neglected, then the efforts spent to identify medical or technological solutions risk being wasted.

To anticipate and solve last mile issues, it is vital to identify barriers to vaccination uptake and to come up with a strategy that maximizes vaccine uptake. In every country for which there is data, people provide similar reasons to refuse potential COVID-19 vaccination (for example, in Canada iii , or in France [2] ; see Table 1 ), and similar to the reasons reported for refusing seasonal flu vaccination [3] ; they also map the WHO-SAGE "Three C model" emphasizing that complacency, lack in confidence, and convenience issues impede vaccination. Moreover, in some countries at least these reasons appear to vary little across socio-demographic groups [2] . Such commonalities across countries, sociodemographic groups, and infectious diseases license the development of a common strategy that capitalizes on previous research to increase vaccine uptake. We propose three main intervention routes, which should receive urgent research and governmental attention.

A mass media campaign is warranted by the facts that no group is spared by vaccine hesitancy, and that the same reasons for hesitancy are provided by all sociodemographic groups. Mass media campaigns can be effective, but messages must be tested before being broadcasted. Our first recommendation is therefore to identify the most effective messages using solid methods (See Box 1).

Previous research has identified several promising levers to reduce vaccine hesitancy. These include the motivation to be altruistic [4] , emphasis on the dangers of the disease [5] , and the anticipated regret not to have gotten vaccinated [6] .

Although mass media campaign have the benefit of reaching broad swathes of the population, their effects remain modest [7] . Discussion with a trusted source can be vastly more effective, in particular when health professionals are involved. One study found that human papillomavirus vaccine uptake went from 2% to 38% among boys whose parents had received a recommendation by a medical provider [8] . In another, an expert taking part in a Q&A reduced the percentage of audience members saying they would not get vaccinated against H1N1 from 75% to 55% [9] . It is thus important to design interventions at all scales and, when large effects are required, to make the best of the power of discussion to change minds. For example, in a recent study, a chatbot designed to address arguments against the COVID-19 vaccine had proved successful in shifting vaccination attitudes and behavioral intentions [10] .

A communication campaign is necessary but not sufficient given the existence of important intention-toaction gaps in this domain. To close these gaps, vaccination should be free and easy to access. Lifting barriers that are known to make vaccination inconvenient should increase uptake among those who are already in favor of getting a COVID-19 vaccine, and among those whose intentions will have been changed in response to the communication campaign. Practically, the following steps could be taken, in combination.

First, to minimize the hassle factor, the vaccine should be offered free of charge, with no prescription requirement. Once a vaccine is available in large quantities, wide accessibility in familiar places should be offered, for example by rerouting COVID-19 testing centers into vaccination centers.

Second, doctors should be encouraged to stock the vaccine so that people who visit their GP for other health reasons can get vaccinated without having actively seek the vaccine. This would also allow provider recommendations to be more potent. In addition, studies have found that people pre-scheduled for a flu vaccine appointment were more likely to get vaccinated, even though they could cancel their pre-set appointment if they did not want it [11] . Onsite vaccination in the workplace has also been identified as a key lever. This underlines the power of defaults and simplification in helping people close the intentionto-action gap.

Finally, a large body of work demonstrates that reminders and prompts sent in a timely manner increase uptake [12] . This can take various forms depending on the target population, e.g. text messages to remind parents that it is time for their child to receive a vaccine, emails or postcards to remind adults of the yearly influenza vaccination campaign, or even phone calls from health-care providers. In the case of the COVID-19 vaccine, these methods should be leveraged to target priority populations, with particular attention to people from lower socio-economic backgrounds who may otherwise be less informed.

hospital settings have found that having vaccinated health-workers wear a badge reading "I am vaccinated against influenza to protect you." increased vaccination rates [13] . A similar result was found in a Dutch hospital with an additional down-the-road effect on patient pneumonia and influenza morbidity [14] . These examples demonstrate that simply making positive decisions visible can be a powerful lever of change.

In order to trigger this virtuous cycle, it is important to ensure that people have an accurate representation of the actual level of support vaccination enjoys. Pluralistic ignorance, which occurs when people underestimate or overestimate the frequency of a given behavior in the population, can slow down change. Displaying true social norms or the upward dynamics of an emerging social norm can be a powerful engine to bootstrap health behavior change. For example, informing doctors who overprescribe antibiotics that they are further away from the norm than they think reduced over-prescription in a matter of weeks. Informing people about social norms around condom use, exercising, smoking cessation, or vaccination has also been found to be an effective way of changing individual behaviors. Social norms should therefore be leveraged both towards health-care providers and towards the general population.

In high income countries, permanent refusal of all vaccines is rare, typically no more than 1 or 2% of the population and studies have shown that many people who initially refuse a vaccine eventually change their mind [15] . This suggests that, for the vast majority of the population, the "enlightened preference"-that is, the preference formed under full information-is to be vaccinated. However, how quickly we get most of the population vaccinated will have a huge impact on the final death toll. The next few months are crucial to solve last mile issues, leveraging early adopters to overcome vaccine hesitancy. We owe it to the researchers who have worked tirelessly on developing a vaccine in record time to ensure that people are well-informed about vaccination, and that they can turn their vaccination intentions into action as easily as possible, so that the last mile does not end up being longer than the first. Because of the possibility of a mutation of the virus (so the vaccine would be less effective or not effective at all).

Because a vaccine developed in an emergency is potentially dangerous, side effects are unknown, and people can get sick. Because of my distrust of the pharmaceutical industry/medical profession, the laboratories are in a logic of profitability.

Because of a lack of information, divergent opinions, no scientific consensus Because there are secret relationships between the government and the pharmaceutical industry (opinions/conspiracy theories) Because of the poor management of vaccines and/or masks by the government (negative opinions/comments regarding the government and its management of the epidemic) ANTIVAX I am against vaccination (in general), I don't get vaccinated Because vaccines are (possibly) harmful, because of the impact of vaccines on health Box 1. How to test a COVID-19 vaccine messaging strategy 1. Use the literature to identify plausibly effective messages and modes of delivery (standard messaging [4] - [6] , in person discussion [8] , [9] , chatbot [10] ). 2. Identify the outcome of interest: vaccination rates, or proxies such as vaccination attitudes and intentions, or motivation to transmit pro-vaccination messages. 3. Choose a sample. If the main interest is the effect of a message compared to no persuasive message, large convenience samples (e.g., recruited through crowdsourcing) should be sufficient. Otherwise, use a representative or other relevant sample. 4. Pre-register the experiment and the analysis plan. Include analyses of sub-populations of interest, such as the most vaccine-hesitant individuals, or individuals with different levels of trust in the government to test whether any backfire effect is observed among them. This is particularly relevant if the population's positions regarding vaccination are politically polarized. iv 5. Conduct an experiment with a control group receiving a non-persuasive message (e.g. [10] ). If measuring attitudes and intentions, pre-and post-messaging measures might create task demands, but are more informative. If possible, conduct a second wave after a delay to test for the durability of any effect. After the last post-messaging measures, give the message to the control group.

Challenges in creating herd immunity to SARS-CoV-2 infection by mass vaccination

Sharp rise in vaccine hesitancy in a large and representative sample of the French population: reasons for vaccine hesitancy

Barriers of Influenza Vaccination Intention and Behavior -A Systematic Review of Influenza Vaccine Hesitancy

The influence of altruism on influenza vaccination decisions

Countering antivaccination attitudes

Anticipated regret and health behavior: A meta-analysis

Not born yesterday: The science of who we trust and what we believe

HPV vaccination among adolescent males: results from the National Immunization Survey-Teen

Impact of information on intentions to vaccinate in a potential epidemic: Swine-origin Influenza A (H1N1)

Information Delivered by a Chatbot Has a Positive Impact on COVID-19 Vaccines Attitudes and Intentions

Opting in vs opting out of influenza vaccination

Patient reminder and recall interventions to improve immunization rates

P037: Nosocomial influenza prevention using multimodal intervention strategies; 20-years of experience

Hospital-based cluster randomised controlled trial to assess effects of a multi-faceted programme on influenza vaccine coverage among hospital healthcare workers and nosocomial influenza in the Netherlands

Increasing vaccination: putting psychological science into action

We benefited from two grants from the ANR: EUR FrontCog ANR-17-EURE-0017* to ASH, CC and HM, and BEHAVIRAL ANR-20-COVI-0060 to CC and HM.

i http://www3.weforum.org/docs/WEF_survey_vaccine_confidence_2020.pdf ii https://www.ipsos.com/sites/default/files/ct/news/documents/2020-12/global-attitudes-on-a-covid-19-vaccine-december-2020-report.pdf iii https://www150.statcan.gc.ca/n1/pub/45-28-0001/2020001/article/00073-eng.htm J o u r n a l P r e -p r o o f Journal Pre-proof iv https://www.pewresearch.org/science/2020/12/03/intent-to-get-a-covid-19-vaccine-rises-to-60-asconfidence-in-research-and-development-process-increases/