While researchers and pharmaceutical companies were working around the clock to develop a COVID-19 vaccine, it became clear that not everyone was eagerly awaiting such a vaccine despite the global crisis.

The American researcher Carl Bergstrom had developed plans for the United States to prepare for a pandemic scenario about twenty years ago. At the time, his primary concern was whether vaccines would reach the people who needed them most. He feared that there could be riots because everyone would want a vaccine as quickly as possible. When the COVID-19 crisis hit, the problem turned out to be quite the opposite: even with a million COVID-19 deaths in the United States, a quarter of the adult population remains unvaccinated for various reasons. Convincing people that this vaccine would save many lives and that everyone should take it proved to be much more challenging than expected.

In Belgium, there was resistance among certain segments of society even before the vaccination campaign has properly started. The reasons are complex, and this phenomenon is not unique to COVID-19. Vaccine doubt has also been growing around the measles, mumps, and rubella (MMR) and pertussis vaccines. In fact, the World Health Organization listed vaccine hesitancy as one of the top ten threats to global health back in 2019.

Who wants (or doesn't want) a vaccine?

Based on research conducted in Flanders before the COVID-19 crisis, we learned that the decision to get oneself or one's child vaccinated depends on various factors and is primarily influenced by perceptions of vaccine side effects, followed by vaccine accessibility, its effectiveness, and the potential disease burden it prevents. Additionally, trust, or the lack thereof, is a significant element. For vaccines, this involves a balance of trust in the product (the vaccine itself), trust in the healthcare system or government, and trust in the provider (such as healthcare professionals administering the vaccine). Social media has greatly influenced the dynamics of information and misinformation spread. The majority of people view healthcare professionals as the most reliable source of health advice, more so than family members, friends, religious leaders, or celebrities. Apart from building trust in the government, healthcare professionals play a crucial role in achieving a sufficiently high vaccination rate and thus in the success of the COVID-19 vaccination campaign.

Concerns regarding COVID-19 vaccines are primarily fueled by the rapid development of new vaccines, concerns about long-term side effects, and the fact that, in many cases, these vaccines are relatively new, using mRNA technology.

To better understand the current vaccination rate, we studied the reasons behind people's decisions in more detail. From the Great Corona Study, we found that a large portion of the participants (84.2%) were willing to get vaccinated: 61.2% were certain they would accept a vaccination invitation, and 23% said they would probably accept it. In the middle of the summer, at the end of July 2020, vaccine willingness peaked at 90% before slightly decreasing and then rising again. During the period when the Great Corona Study was conducted, vaccine willingness among participants never fell below 80%.

Approximately 15.9% were hesitant or resistant to vaccination: 10% might consider getting vaccinated, 3.9% probably would not, and 2.2% definitely would not. The main reasons given for not wanting to be vaccinated were concerns about the vaccine's safety (61%), worries about its rapid development (54%), and a preference to wait until more people are vaccinated (45.9%). Over a quarter of those with doubts (28.4%) stated that their final decision would depend on the type or brand of vaccine offered.

It's important to note that the participants in the Great Corona Study may not be fully representative. However, we were later able to confirm the same trends based on the actual vaccination figures. When the vaccination program started in December 2020, it initially focused on residents of nursing homes, healthcare workers, the elderly, pregnant women, and vulnerable people. Subsequently, younger age groups were given the opportunity to be vaccinated step by step. In Flanders, in November 2021, between 86.5% and 97.1% (depending on the age group) of the adult population had been vaccinated. Overall, the vaccination campaign in Flanders was a success, with a high vaccination rate across all age groups. However, there is some variation, such as in age. Children aged 12 to 17 have the lowest vaccination rate, while those aged 65 and over have the highest, which is logical given the difference in vulnerability and disease burden for COVID-19. The vaccination rate somewhat follows the order of vaccination, although the data in our study regarding differences and patterns were collected at a time when all age groups had the opportunity to be vaccinated.

Differences in age might explain regional differences to some extent, but ethnicity and socioeconomic status also play a role. Regions with a more diverse population and lower socioeconomic status had lower vaccination rates. One possible reason is the influence of news media from one's home country or via social media. Language barriers may also play a role, even though specific attention was paid to this issue in the vaccination campaign, both online and offline. Trust and/or the reach of the campaign can be lower for this group. A caveat to these results is that they are based on official vaccination figures, and we do not have information about those who may have been vaccinated in their countries of origin.

Compared to other European countries, Flanders performed well. Some Southern European countries such as Portugal and Spain did slightly better (possibly due to the relatively greater severity of the pandemic), while the vaccination rate in Eastern Europe was lower.

Do we change our behavior when vaccinated?

Vaccinated individuals tend to resume more social contacts. This was evident in the results of the Great Corona Study and was also confirmed based on the longitudinal CoMix study using data from 16 different European countries collected between December 2020 and September 2021. The data showed that, on average, vaccinated individuals made approx. 30% more contacts than unvaccinated individuals. The difference in behavior between those who were vaccinated and those who were not held true for all 16 countries in the study and can likely be attributed to a combination of reduced risk perception after vaccination and the COVID-19 pass, which often made it more difficult for the unvaccinated to participate in public social activities.

The Great Corona Study also revealed that various aspects of our mental well-being were influenced by whether someone was vaccinated or not. Starting in October 2021, unvaccinated individuals reported experiencing less autonomy and more social exclusion.

How effective are vaccines in containing the pandemic?

Of all public health interventions, vaccination is generally one of the most efficient investments. However, their precise economic value is often still underestimated, partly because commonly used economic models are not well-suited to assess all long-term benefits. Assessing the impact on quality of life, for example, is much more challenging for diseases in children than for those in adults. Furthermore, unlike non-communicable diseases, many indirect effects can be avoided by halting infectious diseases through vaccination.

As soon as the vaccination campaign starts, the models used to monitor and estimate the situation on a daily basis need to be adjusted. Many parameters shift: how contagious someone is and the likelihood of getting sick or being hospitalized change (depending on the vaccine's effectiveness, which itself depends on the virus variant) for a certain proportion (only those who have been vaccinated) of specific groups in society (e.g., nursing home residents, healthcare workers, etc.). The model also needs to account for the decline in protection that vaccination provides after a few months.

However, the vaccination campaign appears to yield results quickly. We immediately turned our attention to the group that was hit the hardest and vaccinated first: nursing home residents and healthcare personnel.

On December 28, 2020, 96-year-old Jos Hermans from Puurs-Sint-Amands became the first Flemish person to receive a coronavirus vaccine. In our country, nursing home residents like him were given priority in the rollout of the vaccination campaign due to their increased vulnerability to COVID-19, followed closely by healthcare workers. By the end of May 2021, nearly 96% of nursing home residents and just under 91% of healthcare workers in Flemish nursing homes were fully vaccinated. Naturally, we wanted to monitor how this impacted the spread of the virus.

Using information from 842 nursing homes, we compared the number and size of COVID-19 outbreaks before and after vaccination, in comparison to virus circulation in the rest of society. Before vaccines were available, the chance of a nursing home resident testing positive for COVID-19 was ten times greater than that for the average Flemish person. These proportions reversed after the initial rollout of the vaccination campaign. There were also fewer outbreaks, and the outbreaks that did occur infected fewer residents and were more short-lived. In 15 nursing homes, there was a COVID-19 outbreak with more than five cases. In these 15 centers, vaccinated and unvaccinated residents had a similar chance of testing positive, but vaccinated individuals had a lower chance of severe symptoms. Unvaccinated healthcare workers, however, had a greater chance of testing positive.

As more and more adults were vaccinated and COVID-19 infections decreased among the general population, the number of outbreaks in nursing homes also declined.

According to figures from Sciensano, the vaccines have shown good effectiveness, although this effectiveness gradually diminishes. The vaccine's protection against intensive care unit (ICU) admissions for all adults, including the elderly, only decreases slightly, from 93% during the first 50 days after a booster to 89% after 5 to 6 months. For hospitalizations in general, this percentage varies between 86% and 91%, depending on the age category, during the first 50 days after the booster, but it drops to 60% to 78% after a few months. The effectiveness of the vaccines against symptomatic infection is considerably lower, and the protection decreases more rapidly over time.

Vaccine effectiveness should not only be viewed in terms of disease burden but also in terms of susceptibility and transmission. For transmission, it's a combined indirect effect of how well a vaccine protects against susceptibility (people who aren't infected can't transmit the virus) and disease burden (fewer people are severely or long-term ill, resulting in less virus transmission compared to a situation without vaccination). Vaccine effectiveness is a complex interplay of factors determined by the vaccinated individual (their age, underlying conditions, prior infections, etc.), the pathogen (specifically, which variants are emerging), and the vaccination itself (number of doses, time elapsed, etc.). Researchers at Sciensano calculated, based on more than 300,000 PCR tests conducted between January and June 2021 on high-risk contacts, that fully vaccinated individuals were much less likely to become infected. The protection was found to be as effective as protection from natural infection.

The vaccination campaign altered the pandemic dynamics and the role of children

Fortunately, this pandemic turned out to be less dangerous for children. When the crisis erupted, there was hardly any time to study specific contact patterns among children. Ultimately, during that phase, their role in driving the pandemic was limited because there was no immunity built up among adults at that time. However, the role of children would change in later stages of the pandemic.

Not only contact patterns but also susceptibility and infectivity can all be age-dependent. We developed a computational method to better estimate whether children are less susceptible, less infectious, or both, based on social contact data and infection rates. Our calculations revealed that children were indeed about half as susceptible to SARS-CoV2 compared to adults, and even less so for the very youngest (our estimates ranged from 20% to 50%). In other words, children where only half as likely to become infected after exposure to the virus compared to adults. Whether they, once infected, are less likelty to transmit the virus turned out to be a more nuanced story. As the adult population was vaccinated from old to young, the transmission dynamics changed significantly. Children and young people, who were largely unvaccinated, began to play an increasingly prominent role.

This once again demonstrates how averages can mask relevant aspects in society and how seemingly simple questions, such as "do children play a role in the pandemic?" do not have a straightforward "yes" or "no" answer.

Together with Pieter Libin at the Vrije Universiteit Brussel, we developed a way to simulate the effects of various vaccination strategies based on data for Belgium. We examined the first four months of the vaccination campaign, during which a limited but increasing number of new doses of two different types of vaccines, mRNA vaccines (Pfizer and Moderna) and vector vaccines (AstraZeneca), became available each week. We investigated how the prioritization of different age groups, in combination with various types of social distancing measures, would have affected both the number of infections and hospital admissions. It became clear that the impact on society would have been smaller if children and young people had been vaccinated first. This was a significant insight because, besides safety requirements (the vaccines were initially approved only for adults), there was no support for this idea then, and it still does not have widespread backing today.

Can medication help?

We now know that even perfect contact tracing and quarantine measures are sometimes insufficient to control local outbreaks because many people are contagious even before symptoms appear. Therefore, we wanted to determine whether it would be useful and feasible to use antiviral drugs that reduce the number of viral particles and, consequently, infectivity, in the fight against COVID-19. We did this through various simulations in our models, examining how the curves would change if, on one hand, people were tested based on symptoms, or on the other hand, proactive testing and, in addition to contact tracing and quarantine measures, the use of an antiviral drug were implemented.

From these simulations, it became very clear that for every infectious disease where asymptomatic and presymptomatic infections are possible, the size of the peak could be drastically reduced when antiviral drugs were part of the intervention package. When the COVID-19 crisis began, we did not immediately have highly effective drugs. Nevertheless, this is an important lesson that can help prevent many cases and, consequently, deaths while waiting for the development of new vaccines, even when new viral infections arise in the future.