Exploring Public Sentiment of Cannabinoid Use in Epilepsy: A Quantitative and Qualitative Analysis Through Social Media
Abstract number :
2.497
Submission category :
17. Public Health
Year :
2024
Submission ID :
1389
Source :
www.aesnet.org
Presentation date :
12/8/2024 12:00:00 AM
Published date :
Authors :
Presenting Author: Justine Ker, MD – University of Texas Southwestern Medical Center
Lauren Cooper, MS – University of Texas Southwestern Medical Center
Alexander Radunsky, MPH, ScD – University of Texas Southwestern Medical Center
Christoph Lehmann, MD – University of Texas Southwestern Medical Center
Marisara Dieppa, MD – University of Texas Southwestern Medical Center
Rationale: Leveraging social media data, we aim to understand public perceptions on cannabinoid use in patients with epilepsy. We selected cannabinoid use based on its relevance to clinical practice. To date, the evidence supporting the benefit of cannabis and cannabinoids in patients with epilepsy remains limited except for one FDA approved prescription, cannabidiol, for the treatment of Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex.
Methods: A subset of English-language posts containing selected epilepsy- and cannabinoid-associated key terms were collected from X (formerly Twitter) between 1 August 2023 to 12 July 2024. We processed and analyzed the posts using Python in conjunction with the natural language processing library spaCy. The M3-Inference and Ethnicolr libraries were used to infer user demographics (age, gender, ethnicity). Posts were also assigned to a United Kingdom (UK) versus non-UK cohort. The SentiStrength library was used to assign values based on sentiment (positive or negative) of the words in the post. The Text2Emotion library used natural language processing to categorize posts into one of five emotions (happy, angry, sad, surprise, fear). Topic modeling was completed using Latent Dirichlet Allocation (LDA) from the Gensim Library.
Results:
Public Health