Correlates of COVID-19 vaccination intentions and opinions about mandates among four groups of adults in South Africa with distinct vaccine intentions: evidence from a large national survey.

INTRODUCTION: Despite a high number of recorded COVID-19 infections and deaths in South Africa, COVID-19 vaccine coverage remained low in March 2022, ten months into the national vaccine roll-out. This study provides evidence on the correlates of vaccine intentions, attitudes towards vaccination and opinions about mandates. METHODS: We used data from the second COVID-19 Vaccine Survey (CVACS), a telephone survey conducted February-March 2022 among 3,608 South African adults who self-reported not being vaccinated against COVID-19. The survey instrument was designed in consultation with government, policymakers, and civil society; and segmented the sample into four distinct groups with different vaccine intentions (synonymous with vaccine hesitancy levels). Kruskal-Wallis and Mann-Whitney tests were used to examine the sociodemographic characteristics, attitudes and behaviours associated with the different vaccination intentions groups. Thematic coding of responses to open-ended questions elicited insights on reasons for not being vaccinated and attitudes towards mandates. RESULTS: Intentions to get vaccinated were greater among individuals with lower socio-economic status (Mann-Whitney Z = -11.3, p < 0.001); those believing the vaccine protects against death (Kruskal-Wallis Χ2 = 494, p < 0.001); and those who perceived themselves at risk of COVID-19-related illness (Χ2 = 126, p < 0.01). Vaccine intentions were lower among individuals who believed that the vaccine causes death (Χ2 = 163, p < 0.001); believed that the vaccine is unsafe for the babies of pregnant/breastfeeding mothers, or the chronically ill (Χ2 = 123, p < 0.01); those not trusting government health information about COVID-19 and the COVID-19 vaccine (Kendall's τ = -0.41, p < 0.01); and those in opposition to mandates (τ = 0.35, p < 0.001). Only 25% supported mandates, despite 48% thinking mandates would work well, with 54% citing individual rights as their main reason for mandate opposition. CONCLUSION: The profile of individuals not vaccinated against COVID-19 as of March 2022 varied markedly by self-reported vaccination intentions, underscoring the importance of tailored demand-creation efforts. This paper highlights several factors which differ significantly across these groups. These findings could inform the design of future vaccination campaigns, potentially increasing their likelihood of success. This is an important policy objective given widespread vaccine hesitancy, and further work is required on this topic. Mandates remain an option to increase coverage but need to be carefully considered given extensive opposition.