Predictors of psychological improvement on non-professional suicide message boards: content analysis.

2016 
Background Suicide message boards have been at the core of debates about negative influences of the Internet on suicidality. Nothing is currently known about communication styles that may help users to psychologically improve in these settings. Method In all, 1182 archival threads with 20 499 individual postings from seven non-professional suicide message boards supporting an ‘against-suicide’, ‘neutral’ or ‘pro-suicide’ attitude were randomly selected and subject to content analysis. Initial needs of primary posters (i.e. individual who open a thread), their psychological improvement by the end of the thread, their responses received and indicators of suicidality were coded. Differences between ‘pro-suicide’, ‘neutral’ and ‘against suicide’ boards, and correlations between primary posters and respondents in terms of suicidality were assessed. Logistic regression was used to test associations with psychological improvement. Results ‘Pro-suicide’ boards ( n = 4) differed from ‘neutral’ ( n = 1) and ‘against-suicide’ ( n = 2) boards in terms of communicated contents. Indicators of suicidality correlated moderately to strongly between primary posters and respondents on ‘pro-suicide’ message boards, but less on other boards. Several communicative strategies were associated with psychological improvement in primary posters, including the provision of constructive advice [adjusted odds ratio (aOR) 4.10, 95% confidence interval (CI) 2.40–7.03], active listening (aOR 1.60, 95% CI 1.12–2.27), sympathy towards the poster (aOR 2.22, 95% CI 1.68–2.95) and provision of alternatives to suicide (aOR 2.30, 95% CI 1.67–3.18). Conclusions Respondents resemble primary posters with regard to suicidality in ‘pro-suicide’ boards, which may hinder psychological improvement. Still, opportunities to intervene in these settings using simple communication techniques exist and need to be taken and evaluated.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    37
    References
    19
    Citations
    NaN
    KQI
    []