Impact of COVID-19 on multiple sclerosis topic discussion on Twitter

Guido Giunti*, Maëlick Claes, Enrique Dorronzoro Zubiete, Octavio Rivera-Romero, Elia Gabarron

*Corresponding author for this work

Research output: Chapter in Book/Conference proceedings/Edited volumeChapterScientificpeer-review

9 Downloads (Pure)


Introduction: Multiple sclerosis (MS) is one of the world's most common neurologic disorders. Social media have been proposed as a way to maintain and even increase social interaction for people with MS. The objective of this work is to identify and compare the topics on Twitter during the first wave of COVID-19 pandemic. Methods: Data was collected using the Twitter API between 9/2/2019 and 13/5/2020. SentiStrength was used to analyze data with the day that the pandemic was declared used as a turning point. Frequency-inverse document frequency (tf-idf) was used for each unigram and calculated the gains in tf-idf value. A comparative analysis of the relevance of words and categories among the datasets was performed. Results: The original dataset contained over 610k tweets, our final dataset had 147,963 tweets. After the 10th of march some categories gained relevance in positive tweets ("Healthcare professional", "Chronic conditions", "Condition burden"), while in negative tweets "Emotional aspects" became more relevant and "COVID-19" emerged as a new topic. Conclusions: Our work provides insight on how COVID-19 has changed the online discourse of people with MS.

Original languageEnglish
Title of host publicationPublic Health and Informatics
Subtitle of host publicationProceedings of MIE 2021
EditorsJ. Mantas
PublisherIOS Press
Number of pages5
ISBN (Electronic)9781643681856
ISBN (Print)9781643681849
Publication statusPublished - 2021

Publication series

Name Studies in Health Technology and Informatics


  • COVID-19
  • Multiple Sclerosis
  • Sentiment Analysis
  • Social Media


Dive into the research topics of 'Impact of COVID-19 on multiple sclerosis topic discussion on Twitter'. Together they form a unique fingerprint.

Cite this