SARS-CoV-2 vaccine hesitancy, defined as a delay in acceptance or refusal of vaccine, is still a major hurdle in achieving community immunity. Identifying target subpopulations and assessing the specific contribution of psychological variables may help to tailor effective communication strategies, improving vaccination adhesion. In this study we explore socio-demographic and psychological factors that correlate and predict vaccine hesitancy applying machine learning methodologies. In a sample of 1728 individuals, recruited from 27 February 2020 to 5 January 2021 in Italy, 26% of participants would not vaccinate. Vaccination hesitancy was predicted by older age, being women, lower education, living in less populated settlements, higher fatalistic attitude, lower perceived COVID-19 severity and threat, lower perceived efficacy of containment measures, and influence of media on protective behaviors, whereas vaccine acceptance was predicted by being a student, scientific education, and having a family member affected by pre-existing severe diseases. Vaccine hesitancy was also associated with higher moral disengagement. The model reached a balance accuracy of 67% and AUC of 72%, correctly identifying 69% of the hesitant individuals. Psychological variables emerged as important determinants of vaccine adhesion, together with socio-demographic variables that may help in identifying target populations and to tailor better communication strategies.
Vai et al. (Sun,) studied this question.