In the first 10 months of availability, COVID-19 vaccines prevented around 235,000 deaths and 1.6 million hospitalizations in the U.S. However, by April 2023, 19% of Americans still hadn’t received their first vaccine dose and 65% hadn’t completed all recommended boosters, leading to tens of thousands of preventable deaths. With annual booster recommendations likely, increasing booster uptake is essential to prevent future hospitalizations and deaths.
In a new paper, research led by Katy Milkman of the University of Pennsylvania looks at the efficacy of COVID-19 vaccination efforts.
“For our megastudy, we teamed up with CVS and we purchased rides from Lyft to see if reducing the friction associated with getting to and from a vaccine appointment could make a difference,” says Milkman, the James G. Dinan Professor at Penn’s Wharton School and co-director of the Behavior Change for Good Initiative at Penn, where this research was conducted. “And to our surprise, there was no added benefit of the free transportation over and above sending a reminder.”
The study, published in the journal Nature, involved a massive field experiment including 3.66 million CVS pharmacy patients who had previously received their primary COVID-19 vaccination series. Participants received different types of text reminders encouraging them to get their recommended COVID-19 booster, with some also being offered free round-trip Lyft rides to their vaccination appointments.
Despite high expectations from both behavioral science experts and lay forecasters, the results showed that providing free transportation to vaccination sites did not boost vaccination rates any more than merely sending well-designed reminder messages. Milkman and her team found that their text reminders increased 30-day COVID-19 booster uptake by 21% on average and had a positive spillover effect, increasing 30-day flu vaccinations by 8%.
The researchers identified three types of reminder messages that were the most useful for boosting vaccination rates: messages encouraging recipients to make a vaccination plan that suggested a specific date, time, and location based on their last vaccination; messages highlighting high local COVID-19 transmission rates; and messages sent on behalf of the patient’s local pharmacy team.
Milkman says that, if someone’s last vaccine was on a Tuesday at 3 p.m. at a specific CVS, for instance, the next reminder would closely match that time and location under the assumption that that is probably a convenient time and place.
“Another message highlighted high infection rates in their county, and the third message was designed to make it feel like it was coming from their local pharmacy team letting them know there’s a vaccine reserved for them,” Milkman says.
The study’s data analysis showed that, while all text reminders increased vaccination rates, the messages offering free round-trip Lyft rides to and from vaccination sites did not produce any measurable benefit over the reminders alone, suggesting that the primary barrier to COVID booster vaccination is not transportation.
“This kind of science is really important because the government invested heavily in free-ride programs,” Milkman says. “It’s crucial to understand what works and what doesn’t so we can avoid wasting resources and better address the problem. Our study contributes to this by showing that reducing transportation barriers is not the key lever we need to pull right now.”
Looking ahead, Milkman and her team at the Behavior Change for Good Initiative feel that future research should explore more varied strategies for boosting vaccination rates. For instance, it could be worth investigating the value of mobile vaccination units that bring vaccines directly to communities, especially in underserved areas. Additionally, studying more personalized communication methods delivered by various trusted sources, alternative incentives, and other behavioral nudges designed to address specific barriers to vaccination could provide valuable new insights.
“There’s always more to learn about what motivates people and how we can effectively encourage health-promoting behaviors,” Milkman says.
Katy Milkman is the James G. Dinan Professor in the Department of Operations, Information, and Decisions at the Wharton School at the University of Pennsylvania and holds a secondary appointment in Penn’s Perelman School of Medicine. She co-founded and co-directs the Behavior Change for Good Initiative at Wharton and the School of Arts & Sciences at Penn.
Other authors are Sean F. Ellis, Dena M. Gromet, Youngwoo Jung, Alex S. Luscher, Madeline K. Paxson, and Ramon A. Silvera Zumaran of the Behavior Change for Good Initiative; Robert Kuan, Christophe Van Blute, Angela Duckworth, and Ron Berman of the Wharton School; Kevin G. Volpp of the Perelman School of Medicine; Maryann V. Beauvais, Johnathon K. Bellows, and Cheryl A. Marandola of CVS Health; Rayyan S. Mobarak of the University of Maryland; Neil A. Lewis Jr. of Cornell University; John A. List of the University of Chicago; and Mitesh S. Patel of Ascension Health.
The research received funding from Flu Lab, the Social Science Research Council’s Mercury Project (with funding from the Rockefeller Foundation, the Robert Wood Johnson Foundation, Craig Newmark Philanthropies, and the Alfred P. Sloan Foundation) and the Wharton Behavioral Laboratory. Support for this research was also provided in part by the AKO Foundation, J. Alexander, M. J. Leder, W. G. Lichtenstein and A. Schiffman and J. Schiffman.
