Promoting Annual COVID-19 Vaccine Uptake in Unvaccinated Populations Through Effective Communication

Importance: While high levels of annual vaccination remains a key strategy for preventing the spread of new COVID-19 variants, most eligible Americans chose not to receive an updated vaccine in 2023. Objective: To assess the effectiveness of two new COVID-19 vaccine uptake communication strategies on annual COVID-19 vaccine uptake among vaccine hesitant sub-populations in a high vaccination state, and determine whether or not the messages “scale up” into an effective national messaging strategy. Design: We administered four randomized controlled trials embedded in two large and representative public opinion surveys before and following the availability of updated COVID-19 vaccines. Trials 1 and 2 were conducted in June 2023, and randomly exposed a representative cross-section of adults in a highly vaccinated state (Rhode Island; Trial 1, N = 800) and a nationally representative cross-section US adults (Trial 2, N = 1,000) with messages that normalize the risks posed by endemic COVID-19 (vs. a control group). Trials 3 and 4 were conducted in October 2023, and randomly exposed analogous state (Trial 3, N = 800) and nationally representative survey respondents (Trial 4, N = 1,200) to messages that “de-normalize” constantly-mutating viral threats posed by endemic COVID-19 (vs. a control group). Setting: Survey respondents were administered the trials online. Participants: All adults aged 18 or over, residing in either the US (Trials 1, 3) or Rhode Island (specifically; Trials 2, 4), and who were members of a large, online, opt-in survey panel recruitment pool maintained by YouGov, were eligible for inclusion in this study. Interventions: Respondents were randomly assigned to read a short message promoting annual COVID-19 vaccine uptake (vs. a control group, shown an unrelated message). In Trials 1 and 2, messages either made an effort to (a) normalize COVID-19 as endemic (“normalization” [NORM]), (b) both normalize and highlight the Long COVID health risks posed by endemic COVID-19 (“normalization + risk” [NR]), or (c) a control message unrelated to vaccination. In Trials 3 and 4, respondents were randomly assigned to either read (a) the aforementioned “NORM” message, (b) a message “de-normalizing” (DENORM) COVID-19 as a constantly mutating viral risk, or (c) a study-wide control message. Outcomes & Measures: The primary outcome variable in this study measures respondents’ self-reported intentions to receive an annual COVID-19 vaccine. The primary explanatory variable in our analyses is an indicator of experimental condition assignment, stratified by whether respondents (a) received a full COVID-19 vaccine sequence & at least one booster shot, (b) received some form of COVID-19 vaccination, but no boosters, or (c) are unvaccinated against COVID-19. Results: In Trials 1 and 2, we find that unvaccinated Rhode Islanders exposed to the NR messaging, but not NORM messaging, were significantly more likely to intend to receive an annual COVID-19 vaccine (NORM: B = 0.13, CI = -0.03, 0.30, p = 0.11; NR: B = 0.22, CI = 0.00, 0.44, p = 0.05). In Trials 3 and 4, we find that exposure to DENORM messaging, but not NORM messaging, is also associated with annual COVID-19 vaccine uptake among unvaccinated Rhode Islanders (NORM: B = 0.19, CI = -0.03, 0.41, p = 0.10; DENORM: B = 0.23, CI = 0.01, 0.44, p = 0.04). We detect no significant effects in either of our national samples, which post hoc analyses suggest might result from observed and unobserved differences in unvaccinated populations across Rhode Island and national samples. Conclusions: Our work suggests that messaging emphasizing both the endemic health risks (“NR”) and challenges of combating new COVID-19 variants (“DENORM”) may be effective at increasing annual COVID-19 vaccine uptake in unvaccinated populations, and suggest opportunities for future research regarding how differences in state public health contexts might moderate the effectiveness of COVID-19 vaccine communication.