This study examined public sentiment regarding COVID-19 vaccination in Brazil using Twitter data from 2020 to 2022. It employed a lexicon algorithm to analyze sentiment expressed in tweets, considering temporal trends, vaccine brand associations, and contextual factors like circulating variants. The VADER algorithm was utilized for sentiment analysis, categorizing posts as negative, neutral, or positive. Results revealed temporal fluctuations in sentiment distribution, with variations observed across different pandemic phases. While positive sentiments initially dominated, negative sentiments became more prominent during periods of increased COVID-19 cases and deaths. Specific vaccine brands evoked distinct associations and sentiments, reflecting broader societal discussions on efficacy, safety, and accessibility. Notably, Coronavac elicited more negative sentiments due to political polarization, while Janssen received more positive sentiments. These findings had implications for health policy and decision-making, emphasizing the importance of understanding public sentiment to inform resource allocation and tailored interventions in public health contexts. Integrating social media data into Health Technology Assessment frameworks could facilitate evidence-based decision-making.