Dynamic predictors of vaccination uptake over two years of the COVID-19 vaccination campaign: lessons from Hong Kong

Abstract

Promoting high vaccine uptake is a primary public health policy in mitigating the profound impact of COVID-19 pandemic. The evolution of the COVID-19 pandemic requires a dynamic evaluation of factors that are associated with COVID-19 vaccination uptake to inform a strategic plan on vaccination programmes. Using real-time population-based data that covered the two-year COVID-19 vaccine rollout in Hong Kong, this study was aimed to systematically investigate factors associated with the uptake of the primary and booster vaccine doses at different stages of the vaccination programme. We adopted a novel network approach to map the relationships of multilevel variables involving contextual factors (number of COVID-19 cases and deaths in the community, social vulnerability of the residential areas), COVID-19 risk perceptions, COVID-19 vaccine attitudes, mental and physical health variables as well as demographics with COVID-19 vaccination uptake over four key periods (P1-P4) of the COVID-19 vaccination campaign in Hong Kong. In total, 3523, 7056, 2580 and 2020 participants in P1, P2, P3 and P4 were included in our network analyses. A total of 983 participants’ verbal reasons for rejecting vaccination were also collected to supplement our understanding of factors associated with COVID-19 vaccination uptake. Our study sheds light on how to improve the effectiveness of future vaccination programmes against a pandemic. First, while people value vaccine safety more in the initial vaccination campaign, more attention is given to vaccine effectiveness at the later stage when evidence about vaccine safety has accumulated. Second, older people and those with chronic disease status are persistently the more vaccine-hesitant groups throughout vaccine rollout. This could be due to the contexts of insufficient decisional support from healthcare workers and family members for older people and their lower health literacy, a case in Hong Kong. However, more proactive and tailored strategies can successfully boost older peoples’ vaccination uptake, which is evidenced by the implementation of home-based vaccination campaign in Hong Kong when the most severe Omicron wave occurred. Third, disease risk has a weak association with vaccination uptake when vaccines become available, suggesting that people start to shift their attention to vaccine risk from disease risk after the vaccines become available particularly when the disease incidence in the community temporarily declines. Communication should focus on addressing concerns about vaccine safety and promoting awareness of disease resurgence risk. Fourth, more adoption of non-pharmaceutical preventive measures against COVID-19 is associated with lower uptake of the vaccine. Such pattern is more evident in P1, indicating that adoption of non-pharmaceutical preventive behaviours might give people the illusion that vaccination is not important and thereby demotivates their vaccination uptake. This is particularly the case in Hong Kong as the zero-COVID policy was implemented to contain disease spread before the Omicron wave occurred. Fifth, better-educated and married participants are likely to be the early innovation adopters who are more supportive of new vaccine-related policies and thereby should be one important target group for mobilizing other groups’ vaccination uptake. Furthermore, trust in government can be an important “bridging” factor to link other variables in the networks with vaccine confidence attitudes which subsequently closely correlate with vaccination uptake.