Content analysis

Psychiatry on Twitter: Content analysis of the use of psychiatric terms in French

This article was originally published here

JMIR Res. 2022 Feb 14;6(2):e18539. doi: 10.2196/18539.


BACKGROUND: With the advent of digital technology and more specifically user-generated content in social media, new ways have emerged to study the possible stigmatization of people in relation to mental health. Several works have studied the discourse conveyed on psychiatric pathologies on Twitter by considering mainly tweets in English and a limited number of terms for psychiatric disorders. This article offers the first study to analyze the use of a wide range of psychiatric terms in tweets in French.

OBJECTIVE: Our objective is to study how generic psychiatric, nosographic and therapeutic terms are used on Twitter in French. More specifically, our study has 3 complementary objectives: (1) analyze the types of use of psychiatric words (medical, misuse, or irrelevant), (2) analyze the polarity conveyed in the tweets that use these terms (positive, negative , or neural), and (3) to compare the frequency of these terms with those observed in related works (mainly in English).

Methods: Our study was conducted on a corpus of tweets in French posted from January 1, 2016 to December 31, 2018 and collected using dedicated keywords. The corpus was manually annotated by clinical psychiatrists using a multi-layered annotation scheme that includes word usage type and tweet opinion orientation. A qualitative analysis was carried out to measure the reliability of the manual annotation produced, then a quantitative analysis was carried out considering mainly the frequency of terms in each layer and exploring the interactions between them.

RESULTS: One of the first results is a resource as an annotated dataset. The initial dataset is composed of 22,579 tweets in French containing at least one of the selected psychiatric terms. From this set, psychiatry experts randomly annotated 3040 tweets that corresponded to the resource resulting from our work. The second result is the analysis of the annotations showing that the terms are misused in 45.33% (1378/3040) of the tweets and that their associated polarity is negative in 86.21% (1188/1378) of the cases. Considering the 3 types of term usage, 52.14% (1585/3040) of tweets are associated with a negative polarity. Abusive terms related to psychotic disorders (721/1300, 55.46%) were more frequent than those related to depression (15/280, 5.4%).

CONCLUSIONS: Some psychiatric terms are misused in the corpora we studied, which is consistent with findings reported in related work in other languages. Thanks to the great diversity of the terms studied, this work has highlighted a disparity in the representations and uses of psychiatric terms. In addition, our study is important to help psychiatrists become aware of the use of the term in new communication media such as social networks which are widely used. This study has the huge advantage of being reproducible thanks to the framework and the guidelines that we have produced so that the study can be repeated in order to analyze the evolution of the use of the terms. While the newly created dataset is a valuable resource for further analytical studies, it could also be used to train machine learning algorithms to automatically identify stigma in social media.

PMID:35156925 | DOI: 10.2196/18539