Abstract
IMPORTANCE: Prior studies found that financial incentives have small, positive direct effects in increasing COVID-19 vaccination rates, but unmeasured social spillovers (ie, changes in outcomes among untreated individuals who are socially exposed to policy beneficiaries) may diminish the overall effect of such policies. OBJECTIVE: To assess the spillover effects of a COVID-19 vaccination financial incentive and assess whether incorporating estimates of spillover meaningfully affects broader evaluations of policy effectiveness. DESIGN, SETTING, AND PARTICIPANTS: This population-level, address-cluster randomized clinical trial was conducted in November 2021. Participants were all adult (aged ≥18 years) residents of Ravensburg, Germany, who were randomly assigned to the treatment or control group. One resident in each address cluster was randomly selected to be an address-cluster representative. Address-cluster representatives in the treatment group received the treatment letter; all other cohabitants at that same address received the control letter. All individuals in addresses randomly assigned to the control group were mailed a control letter. Intention-to-treat data analysis was conducted from January 2022 to May 2024. INTERVENTION: Control letters informed recipients about 7 upcoming free COVID-19 vaccination events. Treatment letters were identical to control letters, except they also offered €40 (US $41.46) for getting vaccinated at one of the events. MAIN OUTCOME AND MEASURE: Primary and booster COVID-19 vaccination uptake was observed and recorded on site during the public vaccination events. Primary vaccinations were defined as either the first dose of a 1-dose vaccine or the first or second dose of a 2-doses vaccine. Boosters were defined as any dose after primary vaccination. Three types of commonly used treatment effects were analyzed: direct, spillover, and overall. RESULTS: Among 41 548 Ravensburg residents (mean [SD] age, 49.96 [19.04] years; 51.3% female), 796 (1.9%) were vaccinated at 1 of the 7 public vaccination events. The direct, spillover, and overall effects of receiving a financial incentive on primary vaccinations were all nonsignificant. For booster vaccinations, the direct effect was negative but not statistically significant (-0.32 percentage points [95% CI, -0.77 to 0.14 percentage points]; P = .17), whereas the overall effect (-0.30 percentage points [95% CI, -0.51 to -0.09 percentage points]; P = .006) was significantly negative. The spillover effect was significantly negative (-0.29 percentage points [95% CI, -0.53 to -0.06 percentage points]; P = .01), but only for the first vaccination events. CONCLUSIONS AND RELEVANCE: This trial found null direct effects on COVID-19 vaccination uptake and negative effects on booster uptake among individuals who did not receive but were indirectly exposed to the financial incentives. The timing of this spillover suggests that cohabitants of financial incentive recipients postponed booster vaccination, thereby undermining the potential effectiveness of this policy. TRIAL REGISTRATION: ISRCTN identifier: ISRCTN59503725.