A brief screening tool for suicidal symptoms in adolescents and young adults in general health settings: reliability and validity data from the Australian National General Practice Youth Suicide Prevention Project

https://doi.org/10.1016/S0005-7967(01)00017-1Get rights and content

Abstract

Using data from a nationwide project on young people in Australia aimed at assessing suicidality in general health settings, we present a brief screening tool for suicidality (the depressive symptom index suicidality subscale). Two thousand eight hundred and fifty-one (15–24 year old) patients presenting to 247 Australian general practitioners between 1996 and 1998 were assessed. In addition to the suicide screen, patients completed the general health questionnaire-12 and the Center for Epidemiological Studies depression scale. Patients' chief complaints were taken from the summary sheets completed by their general practitioners. Using inter-item correlational and factor-analytic techniques, as well as a general approach to construct validity, we show that the measure has favorable reliability and validity characteristics. We also provide results on cut-points that may facilitate its use in clinical and research settings. Because the screen is brief, easy to use, reliable, and valid, we encourage its use to combat the vexing international health problem of suicide.

Section snippets

Participants

The study population consisted of 2851 consecutive patients, aged 15–24 years, who fully completed the suicide screening measure at clinics of 247 general practitioners in the Australian States of Tasmania, Victoria, and Western Australia. Gender distribution was 66% female; mean age was 19.6 years. Ninety one percent (91%) of 15–24 year old patients presenting during the period of the study agreed to participate. Most patients were born in Australia (90.1%) with 3.9% born in Europe, 2.4% in

Inter-item characteristics and internal consistency of the suicide screen

Table 1 displays the inter-item correlations of the four items from the DSI-SS suicide screen, as well as corrected item-total correlations. As can be seen there, all inter-item correlations were high, and all item-total correlations far exceeded the guidelines provided by Nunnally and Bernstein (1994) that an item be considered “weak” if it has a corrected item-total correlation <0.30. Accordingly, it is no surprise that coefficient alpha for the DSI-SS was quite acceptable (alpha=0.90). The

Discussion

We presented the suicidality subscale of the depressive symptom index (see Appendix A, as well as Metalsky & Joiner, 1997) as a brief screening tool for suicidality in young people in general health settings. Results regarding the measure, from approximately 2800 Australian adolescents and young adults, were favorable. Specifically, the scale was clearly unitary, cohesive, and internally consistent. Moreover, it evinced construct validity by relating to several demographic and clinical indices

Acknowledgments

This work was conducted under the auspices of the Psychiatry Department, Princess Margaret Hospital for Children, Perth, Australia. Funding for this project was provided by a grant from the Commonwealth Department of Health and Aged Care, Mental Health Branch, as part of the National Youth Suicide Prevention Strategy.

References (24)

  • T. Joiner et al.

    Negative attributional style for interpersonal events and the occurrence of severe interpersonal disruptions as predictors of self-reported suicidal ideation

    Suicide and Life Threatening Behavior

    (1995)
  • T. Joiner et al.

    Disentangling the interrelations between hopelessness, loneliness, and suicidal ideation

    Suicide and Life Threatening Behavior

    (1996)
  • Cited by (279)

    View all citing articles on Scopus
    View full text