BackgroundAcross various stages of the COVID-19 pandemic and related vaccine recommendations in France, we assessed the association of the 7C-psychological antecedents with vaccine uptake/intention for booster vaccination among healthcare-sector workers (HCSWs). We also assessed whether 7C-antecedent profiles changed over time. MethodologyThe Research Group for the Prevention of Occupational Infections in Healthcare Workers (GERES) conducted three repeated web-surveys which were disseminated by email chain-referral among HCSWs throughout France. The questionnaires waves took place: July-November 2021, February-March 2022 and January-March 2023 (P2, P3 and P4). We also reanalysed data from a prior similar study conducted late 2020-early 2021 (Moirangthem et al. (2022)) (P1). To evaluate the association of 7C-items with vaccine uptake-intention for future vaccination, we estimated adjusted prevalence ratios (aPR) using robust variance Poisson regression. We report the 7C-item population attributable loss in vaccine intention. ResultsThe four surveys (P1-P4) encompassed 5234, 339, 351 and 437 participants. At earlier stages of the vaccine campaign, the principal antecedents of vaccine intention were favorable perception of vaccination benefit-risk-balance (BRB) (vs. unfavorable, aPR: 2.32), reactance to employer encouragement for vaccination (motivates vs. dissuades-me, aPR:2.23), vaccine confidence (vs. not-being-confident, aPR: 1.71) and social conformism towards vaccination (favorable vs. skeptical opinion in private environment, aPR: 1.33). Under a vaccine mandate for HCSWs, only perceiving vaccination as a collective action was associated with current vaccine status (agree vs. disagree, aPR: 2.19). At later stages of the epidemic, hypothetical booster vaccine intentions were strongly associated with BRB perception (favorable vs. unfavorable, aPR: 2.07) and perceiving vaccination as a collective action (agree vs. disagree, aPR: 1.69). Fearing a severe side effect from vaccination decreased population vaccine intention by 26.2 %. ConclusionOur results suggest that both 7C-antecedents and their association with vaccine behaviour can change over time, and underscore the importance of assuring confidence in vaccine safety.