Monitoring Online Discussions About Suicide Among Twitter Users With Schizophrenia: Exploratory Study

Yulin Hswen; John A Naslund; John S Brownstein; Jared B Hawkins

doi:10.2196/11483

Monitoring Online Discussions About Suicide Among Twitter Users With Schizophrenia: Exploratory Study

JMIR Ment Health. 2018 Dec 13;5(4):e11483. doi: 10.2196/11483.

Authors

Yulin Hswen^{1

2}, John A Naslund³, John S Brownstein^{2

4

5}, Jared B Hawkins^{2

4

5}

Affiliations

¹ Department of Social and Behavioral Sciences, Harvard TH Chan School of Public Health, Harvard University, Boston, MA, United States.
² Informatics Program, Boston Children's Hospital, Boston, MA, United States.
³ Department of Global Health and Social Medicine, Harvard Medical School, Boston, MA, United States.
⁴ Department of Pediatrics, Harvard Medical School, Boston, MA, United States.
⁵ Department of Biomedical Informatics, Harvard Medical School, Boston, MA, United States.

PMID: 30545811
PMCID: PMC6315229
DOI: 10.2196/11483

Abstract

Background: People with schizophrenia experience elevated risk of suicide. Mental health symptoms, including depression and anxiety, contribute to increased risk of suicide. Digital technology could support efforts to detect suicide risk and inform suicide prevention efforts.

Objective: This exploratory study examined the feasibility of monitoring online discussions about suicide among Twitter users who self-identify as having schizophrenia.

Methods: Posts containing the terms suicide or suicidal were collected from a sample of Twitter users who self-identify as having schizophrenia (N=203) and a random sample of control users (N=173) over a 200-day period. Frequency and timing of posts about suicide were compared between groups. The associations between posting about suicide and common mental health symptoms were examined.

Results: Twitter users who self-identify as having schizophrenia posted more tweets about suicide (mean 7.10, SD 15.98) compared to control users (mean 1.89, SD 4.79; t₃₇₄=-4.13, P<.001). Twitter users who self-identify as having schizophrenia showed greater odds of tweeting about suicide compared to control users (odds ratio 2.15, 95% CI 1.42-3.28). Among all users, tweets about suicide were associated with tweets about depression (r=0.62, P<.001) and anxiety (r=0.45, P<.001).

Conclusions: Twitter users who self-identify as having schizophrenia appear to commonly discuss suicide on social media, which is associated with greater discussion about other mental health symptoms. These findings should be interpreted cautiously, as it is not possible to determine whether online discussions about suicide correlate with suicide risk. However, these patterns of online discussion may be indicative of elevated risk of suicide observed in this patient group. There may be opportunities to leverage social media for supporting suicide prevention among individuals with schizophrenia.

Keywords: Twitter; digital technology; mental health; schizophrenia; social media; suicide.

©Yulin Hswen, John A Naslund, John S Brownstein, Jared B Hawkins. Originally published in JMIR Mental Health (http://mental.jmir.org), 13.12.2018.

Abstract

Grants and funding