This cross-sectional examined the COVID-19–related sentiments on ivermectin using Twitter data. Primary data were collected from Twitter posts about ivermectin worldwide from January 15 to 22, 2022. We searched posts with the keyword “ivermectin” and retrieved raw data with various variables. Data cleaning involved lowercasing capital letters, eliminating retweet symbols, removing punctuation marks, and other preprocessing steps to ensure data accuracy [6]. People’s sentiments were determined by creating a corpus (body of text) and loading a lexicon dictionary based on the positive and negative words. The sentiment score was calculated with the Bing method using a range of –6 to +6 and considered 3 sentiment types: positive, neutral, and negative [7,8]. The Bing lexicon was chosen for its proven effectiveness and simplicity in extensive dataset analysis. Frequency analysis of single terms (unigrams), word pairs (bigrams), and bigram networks helped identify frequently mentioned terms and explore relationships between words [7,8]. Despite its simplicity, this approach is as robust as more complex methods, as it accurately categorizes sentiments and identifies patterns through unigram and bigram frequency analysis [6]. This computational analysis–focused methodology offers valuable insights into public sentiment, complementing traditional clinical evaluation without extensive statistical validation. Data were mined using Twitter’s limited application programming interface (version 2) and analyzed using RStudio (R Foundation) and relevant packages for visualization.

There was no direct connection with Twitter users. We removed all individual data, usernames, IDs, and tweets in the manuscript and supporting material. Therefore, ethical approval was not required.

Our analysis shows that tweets about repurposing ivermectin for COVID-19 treatment are predominantly positive. Terms like “medication,” “potentially,” and “treatment” frequently appeared in positive contexts, reinforcing this finding. Phrases such as “ivermectin works,” “available ivermectin,” and “ivermectin medication” were strongly associated with positive sentiment. These results align with previous findings showing ivermectin’s popularity on Twitter for COVID-19 treatment [4,5]. However, some negative sentiment was observed, particularly concerning warnings from the Food and Drug Administration and the limited clinical evidence supporting ivermectin’s efficacy in COVID-19 prevention and treatment [9].

Twitter data have proven valuable in monitoring public responses to the COVID-19 pandemic, as evidenced by a study of millions of SARS-CoV-2–related tweets [10]. This study not only identified dominant topics like new cases, death rates, and preventive measures, but also explored the geographic distribution of sentiments through tweet-embedded geolocation data (code available in Multimedia Appendix 1).

Of note, the findings are based on Twitter data, which may not represent the entire population’s sentiments. Therefore, researchers conducted cross-validation of the sources.

This sentiment analysis highlights the polarized perception of ivermectin in COVID-19–related discourse and reflects broader public health debates. Given the regulatory advisories and the limited clinical evidence, these public sentiments must be interpreted cautiously.