Vaccines can prevent serious illness, disability and death, but their benefits depend on people actually receiving them. When vaccination rates fall, the usual question is how to support more people to get vaccinated. Yet that question may be too simple. A reminder, a conversation, a financial incentive or a more convenient clinic can all help, but they do not necessarily help everyone in the same way. That is the central insight emerging from a major analysis led by Dr Sarah Davies and Dr Clare French, both of the University of Bristol, and colleagues. To go beyond asking whether broad categories such as “education” or “improved access” work, the researchers took apart vaccine uptake interventions to examine their individual ingredients. It is a little like studying recipes: knowing that a cake tastes good is useful, but knowing which ingredients make the difference is far more helpful if you want to create a better cake. More
The scale of the work is striking. The researchers carried out a comprehensive systematic review of randomised controlled trials conducted in high and upper-middle income countries. For the main analysis, they examined 237 studies involving more than 4.3 million participants. The interventions were aimed at people eligible for vaccination, or their caregivers.
To make sense of such varied approaches, the team developed a framework that separated what an intervention contained from how it was delivered. To do this they worked closely with 12 members of the public from diverse underserved communities. Intervention content might include extending opportunities to be vaccinated, helping people schedule appointments, covering costs, offering incentives, providing information, using decision aids, or employing motivational interviewing. Delivery could involve personal or impersonal formats, human or automated interaction, healthcare professionals, community members, and differing numbers of contacts.
The researchers then used a method called component network meta-analysis. The name sounds formidable, but the idea is intuitive. Conventional reviews can tell us whether a package of measures works. Whereas this approach tries to identify which parts of the package are contributing to that effect, while drawing together evidence from many different comparisons. This is important because real-world interventions rarely consist of one neat action. They can be combinations of messages and practical support delivered in different ways or by different people.
Across the whole dataset, several components stood out. Making vaccination opportunities more accessible, helping with appointment scheduling, offering financial incentives, covering vaccination-related costs and using motivational interviewing were associated with improved uptake. Human interaction also appeared beneficial overall, as did involving community members in delivering interventions alongside healthcare professionals.
But the most revealing findings came when the researchers stopped treating the population as one homogeneous group. Dr Davies, Dr French and their colleagues examined the evidence separately for children, adolescents and young adults, and adults. The pattern changed substantially with age. An ingredient that looked promising for one group could have little apparent benefit for another, or even point in the opposite direction.
For children, practical barriers appeared especially important. Payments designed to cover costs were beneficial, and decision aids also showed a positive effect. Information about social factors, including social responsibility and peer support was also beneficial. There was some evidence that extending opportunities for vaccination, such as making services easier to reach or attend, could help too. For busy parents and caregivers, the obstacle may not be a lack of information or confidence in vaccines at all. It may simply be the friction of fitting vaccination into everyday life.
Among adolescents and young adults, a rather different picture emerged. Personal delivery formats were beneficial, as was involving community members alongside healthcare professionals. Information about social factors also appeared particularly useful. Intriguingly, human interaction was less beneficial than non-human interaction in this age group, while decision aids showed a negative effect. The researchers caution against overinterpreting unexpected findings, especially where relatively few studies contributed evidence. Still, the results suggest that simply importing approaches designed for adults into adolescent vaccination interventions may miss what matters to younger people.
Adults showed almost the reverse pattern in some respects. Human interaction was beneficial, while non-human interaction was less helpful than no interaction. Practical measures also mattered: extended vaccination opportunities, help with appointment scheduling and payments to cover costs all improved uptake. Motivational interviewing, a conversational approach designed to explore and strengthen a person’s motivation, was beneficial too. There was also some evidence that financial incentives and information about vaccine safety and effectiveness could help.
These contrasts are perhaps the study’s most important message. There is no universal recipe for improving vaccination rates. An intervention is not simply effective or ineffective in isolation. Its success can depend on who receives it, how it is delivered and which barrier it is trying to remove. Designing interventions around the needs of a particular population may therefore be more productive than reaching for a standard solution.
Other analyses reinforced that point. Among underserved populations, extending opportunities to receive vaccines and providing financial incentives were beneficial, with some evidence supporting community involvement, providing information on social factors, and prompts or reminders. The researchers also compared trials conducted before and after the start of the COVID-19 pandemic. Some patterns changed, although fewer trials were available for the later period, making cautious interpretation essential. The pandemic may therefore have altered not just vaccination behaviour, but the effectiveness of some strategies intended to influence it.
Effectiveness, however, is only half the policy equation. Health services also need to know what interventions cost. In a companion study, the research team searched for economic evidence linked to the randomised trials in their systematic review. They found eligible economic information for only 37 studies. Just 16 provided full economic evaluations, while others offered cost analyses or more limited information. That scarcity makes it difficult for decision makers to assess whether interventions that are effective also represent a cost-effective use of limited resources.
Future vaccine uptake trials therefore need to report intervention costs much more consistently. An intervention that increases uptake may be valuable, but policymakers must be able to compare its benefits with the staff time, technology, payments and other resources required to deliver it.
Taken together, this research replaces a deceptively simple question, “How do we increase vaccine uptake?”, with a more useful one: “Which ingredients will help this particular group get vaccinated?” The answer might be easier access for a busy parent, socially meaningful communication for a teenager, or a supportive conversation and scheduling help for an adult. By identifying the ingredients that matter, rather than treating every intervention as an indivisible package, Dr Sarah Davies, Dr Clare French and their colleagues offer a practical route towards vaccine uptake interventions that are not only more effective, but more thoughtfully designed.
