Real-world uptake of COVID-19 vaccination among individuals expressing vaccine hesitancy: a registry-linkage study

ABSTRACT

BackgroundUptake of COVID-19 vaccination remains suboptimal in the United States and other settings. Though early reports indicated that a strong majority of people were interested in receiving the COVID-19 vaccine, the association between vaccine intention and uptake is not yet fully understood. MethodsDuring 24 February-5 December 2021, we enrolled California residents receiving molecular tests for SARS-CoV-2 infection who had not yet received any COVID-19 vaccine doses. Unvaccinated participants provided information on their intentions to receive COVID-19 vaccination in a telephone-administered survey. We matched study participants with a state-wide immunization registry and fit a Cox proportional hazards model comparing time to vaccination among those unvaccinated at study enrollment by vaccination intention (willing, unsure, or unwilling). FindingsAmong 864 participants who were unvaccinated at the time of interview, 272 (31%) had documentation of receipt of COVID-19 vaccination later; including 194/423 (45.9%) who had initially reported being willing to receive vaccination, 41/185 (22.2%) who reported being unsure about vaccination, and 37/278 (13.3%) who reported unwillingness to receive vaccination. Adjusted hazard ratios (aHRs) for registry-confirmed COVID-19 vaccination were 0.49 (95% confidence interval 0.32-0.76) and 0.21 (0.12-0.36) for participants expressing uncertainty and unwillingness to receive vaccination, respectively, as compared with participants who reported being willing to receive vaccination. Time to vaccination was shorter among participants from higher-income households (aHR 3.30 [2.02-5.39]) and who reported co-morbidities or immunocompromising conditions (aHR 1.54 [1.01-2.36]); time to vaccination was longer among participants who tested positive for SARS-CoV-2 infection (aHR 0.60 [0.43-0.84]). Sensitivity of self-reported COVID-19 vaccination status was 82% (80-85%) overall, and 98% (97-99%) among those referencing vaccination records; specificity was 87% (86-89%). InterpretationParticipants stated willingness to receive COVID-19 vaccination was an imperfect predictor of real-world vaccine receipt. Improving messaging about the importance of COVID-19 vaccination, regardless of previous SARS-CoV-2 infection status, may improve vaccine uptake among populations who express hesitancy to initiate vaccination. RESEARCH IN CONTEXTO_ST_ABSEvidence before this studyC_ST_ABSWe searched PubMed and medR{chi}iv for variations and combinations of the terms "vaccine hesitancy", "vaccine confidence", "vaccine uptake", "COVID-19", and "SARS-CoV-2" to identify original research articles published by March 8, 2022. The majority of screened articles were cross-sectional surveys conducted prior to or after implementation of COVID-19 vaccines to assess trends or predictors of participant-reported COVID-19 vaccine hesitancy. While some studies included random population-based samples, many were conducted within subgroups like health care professionals, parents of school aged children, or college students. Evidence about the association between COVID-19 vaccine intentions and subsequent vaccine uptake remains scarce. Three observational studies quantified associations between willingness to receive COVID-19 vaccination and subsequent initiation of vaccination; however, in these studies, follow-up time was limited to the period prior to widespread availability of COVID-19 vaccination or initiation of vaccine mandates in workplaces, schools, and other public places. Therefore, it was unclear whether remaining unvaccinated at follow-up in these studies was a choice or a consequence of the lack of universal access to COVID-19 vaccines. Additionally, most efforts to identify subsequent vaccine uptake relied on self-reported vaccination status, which may be subject to reporting or interviewer bias. We also searched PubMed and medR{chi}iv with variations and combinations of the terms "self-reported", "vaccination", "accuracy", and "COVID-19" and did not discover any articles validating self-reported COVID-19 vaccination status against immunization registry data; whereas, such studies were available for other vaccine-preventable pathogens including influenza, Streptococcus pneumoniae, and human papillomavirus. Added value of this studyWe linked data collected through an ongoing case-control study and a comprehensive state-wide immunization registry to evaluate the association between COVID-19 vaccination intention and subsequent uptake. We also assessed the reliability of self-reported COVID-19 vaccination status by linking participant records with a state-wide immunization registry. We are not aware of another published study assessing predictors of COVID-19 vaccine uptake spanning over 7 months of age-eligible follow-up time and adjudicating the use of self-reported COVID-19 vaccination status. We found that expressing hesitancy to receive COVID-19 vaccination was associated with lower adjusted hazards of subsequent vaccine uptake as compared with expressing willingness to receive vaccination (aHR 0.49; 95% CI 0.32-0.76), although uptake was also suboptimal among individuals who expressed willingness (45%). Participants from lower income households or who had recently tested positive for SARS-CoV-2 were slower to initiate vaccination than from higher income households or who had recently tested negative. People who were pregnant and initially deferred vaccination were faster to receive vaccination than participants who did not cite pregnancy as a reason for refusal. Upon assessing the accuracy of self-reported vaccination status, we found referencing a vaccination card or another calendar reference source improved sensitivity of self-reported vaccination status. Implications of all available evidenceWe provide an evaluation of predictors of COVID-19 vaccine uptake and assess the validity of self-reported COVID-19 vaccination status in comparison with a state-wide immunization registry. We identified that self-reported vaccination intent was a strong but imperfect predictor of subsequent vaccine initiation. However, no single reason for participants to express vaccine hesitancy predicted their likelihood of eventual vaccine receipt. As such, public health campaigns addressing multiple factors underlying vaccine hesitancy including those correcting sources of misinformation, and allaying concerns about short- or long-term side effects and vaccine safety remain important tools to improve acceptance in hesitant populations. Future studies reliant on the use of self-reported COVID-19 vaccination status should strive to utilize additional reference sources like COVID-19 vaccination cards or vaccination registries to reduce misclassification of vaccination status.