Text message “nudges” that initially increased Covid vaccine uptake were not effective later in pandemic

As we wrote at the time, the results showed that simple techniques to "nudge" people into taking the vaccine could have a substantial impact if applied across the population.

But according to a new study in Nature, these nudges have a pretty limited shelf-life. The team finds that timing really matters: while text message reminders increased uptake immediately after the vaccine became available, later in the pandemic they were no longer effective.

In the original paper, researchers studied people who had just become eligible for the vaccine. A day after they were notified that they were eligible, some participants received a reminder text message, which included a link to book a slot for vaccination. Others didn't receive the reminder. The team found that participants who got the reminder were more likely to book an appointment, and were also more likely to have received the vaccine within four weeks. Reminders that included ownership language, telling people that a vaccine had "just been made available for you" and to "claim your dose", were particularly effective.

But this study took place in January and February 2021 when vaccines had only just become available, and participants received the messages as soon as they were eligible. So Nathaniel Rabb from Brown University and colleagues wondered whether the messages would also be effective later on in the pandemic.

The team conducted a trial of similar messages on 142,428 people living in Rhode Island who had not yet been vaccinated. At the start of the study, which took place in May and June 2021, all Rhode Islanders had been eligible to receive the vaccine for five weeks and availability was widespread.

Participants were randomly assigned to receive one of eight messages, sent out by the state's Department of Health. All of these included the kind of ownership language that had been so effective in the previous study (in this case "a vaccine is waiting for you"); some included further messaging including information about vaccine safety or how vaccination can reduce risk to one's family members. They also included a link to a page providing information about vaccination options. A control group didn't receive any messages.

The researchers looked at the likelihood that a participant had been vaccinated both one week after receiving the message, and at the end of the study period. Vaccination rates were low — only 2% of the control group had received the vaccine by the end of the study. And, in contrast to the original paper, participants who had received a reminder were no more likely than the control participants to get vaccinated. Furthermore, when the team looked at each kind of message individually, there wasn't a single version that increased the likelihood someone would receive the vaccine.

The study suggests that a nudge technique that seemed pretty effective when vaccines first became available might not actually work later on. "Public health officials—especially those avoiding or legally barred from mandates—may turn to this strategy to increase vaccination rates among the less enthusiastic but will probably see minimal impact," the team concludes.

The researchers suggest a number of explanations for the lack of effect. Given that these were people who had still not received the vaccine more than a month after it had become available, perhaps they simply represent a group that had strong anti-vaccine opinions — and so were resistant to attempts to "nudge" them. By this point in time, people may also have been overwhelmed with messages about vaccination, and by the increasingly polarised debates about vaccines in the media and online. Further work should try to disentangle all these effects.

These results don't mean that reminder messages about vaccines are ineffective — just that they need to be timed correctly. And, the researchers add, even if they don't convince more reluctant groups to get vaccinated, "we know of no studies showing reduced vaccinations owing to message campaigns, so they carry little potential harm."

Yet there are important practical issues to consider. It's only mentioned in passing in the paper's extended data, but it turns out that the study actually had to end a day earlier than planned, because the Department of Health was receiving complaints from residents who felt they were getting too many messages (the state was also sending out messages about Covid testing, for instance). The Department was worried that people would end up blocking the messages, and wouldn't receive crucial emergency information.

Despite the authors' assertions, this does seem like a potential harm — albeit an indirect one — and highlights the fact that these kinds of nudges need to be considered in their broader context. Optimising the timing of such interventions, based on papers like this, is one step to ensuring that they are as useful as possible.

Further reading

– Evidence from a statewide vaccination RCT shows the limits of nudges

About the author

Matthew Warren (@MattBWarren) is Editor of BPS Research Digest