Article Text

Download PDFPDF
24 An analysis of sentiment and publicity of functional neurological disorder and non-epileptic attack disorder on twitter
  1. Felix May,
  2. Rohan Kandasamy
  1. Brain Injury Rehabilitation Unit, Frenchay, Bristol


Objectives Twitter may provide a platform for clinicians and allied health professionals to publicise Functional Neurological Disorder (FND) and Non-Epileptic Attack Disorder (NEAD), and also provides a platform for patients and their communities to discuss the disorders. The prevalence and sentiment of discussions of these disorders have not been reported before now. We wrote a program to collect and analyse ‘Tweets’ about the subjects in their sentiment, connectivity and content.

Methods Preliminary searches and graph analyses identified the most relevant search terms. Tweets were collected automatically, along with available metadata. Sentiment analysis was performed using natural language processing with valence aware dictionary analysis, allowing automatic interpretation of text including idioms and ‘emojis’.

Results 13347 tweets were collected, with tweets not in English having been excluded from the analysis. The analysis showed a majority positive sentiment in the tweets. The most negative discourse was related to search terms: ‘Medically Unexplained Symptoms’ and ‘Psychosomatic’. Engagement with charities and tweets aiming to promote awareness of the disorders in question were common. Most frequent links to posts about FND were synonyms for the disorder, along with NEAD and charities and awareness movements. For NEAD, the most common links made were with FND, awareness campaigns, synonyms for NEAD, and Chronic Fatigue Syndrome.

Conclusions FND and NEAD have active communities on Twitter. These include both health professionals, patients and lay advocates. The overall sentiment is positive (p<0.05), but with some negativity from sceptical patients and some who are disappointed with their care, and with more negativity associated with certain search terms. (For example, more negative sentiment in tweets about ‘Medically Unexplained Symptoms’ compared to ones about ‘Functional Neurological Disorder’, p<0.0005). Public discourse analysis on websites such as Twitter may prove fruitful for monitoring patient understanding, trends in patient acceptance of diagnosis and factors contributing to these.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.