Abstract
A recent childhood vaccine promotion intervention trial showed no effects on vaccination outcomes relative to usual care. The purpose of this paper was to test assumptions and theory-based relationships underlying hypothesised mechanisms for two vaccine promotion educational websites (one tailored to parental values, beliefs, and intentions; one untailored) compared with usual care. This is a secondary analysis of a three-arm randomized controlled trial. Parental vaccine values, hesitancy, attitudes, and intention to vaccinate surveys were administered at baseline (≤2 months) and at 4–6 and 10–12 months of age. Vaccination was assessed using electronic health records. Analyses included random coefficient models and risk differences with exact confidence limits. Parental vaccine values were mostly stable over time. Vaccine attitudes were generally positive, with no differences among study arms. Both tailored and untailored website arms showed similar increases in intention to vaccinate more than usual care. Positive changes in intentions were associated with lower rates of late vaccination. Although attitudes and intentions predicted vaccination behavior and the intervention increased intention to vaccinate all on time, the web-based education and values-tailored messaging approaches were not effective at increasing vaccination rates. Intentions are necessary but insufficient targets for vaccine promotion interventions.