Symposia
Suicide and Self-Injury
Lily A. Brown, Ph.D.
University of Pennsylvania
Philadelphia, Pennsylvania
Yiqin Zhu, M.S.
Research Assistant
University of Pennsylvania
Philadelphia, Pennsylvania
Kevin Narine, PhD
Clinical Psychology Doctoral Student
William James College
Newton, Massachusetts
Emily Ballentine, PhD
Research Assistant
University of Pennsylvania
Chalfont, Pennsylvania
Persons living with HIV (PLWH) are at significantly higher risk for suicide, but most of the research on risk factors for suicide in this population has relied on cross-sectional designs. In this pilot study, we used a combination of active ecological momentary assessment (EMA) and passive observation with a wearable device (Fitbit) to intensively assess PLWH for changes in suicide risk. We hypothesized that we would observe significant variability in suicide risk, that suicide risk would peak in the middle of the night (consistent with two large studies on timing of suicide death), and that the assessments would be perceived as feasible and acceptable to participants.
Methods: Participants (N = 10) met these inclusion criteria: HIV+, past month suicidal ideation, access to a smartphone, and not in a manic/psychotic episode. They completed a phone screen, in-person baseline evaluation, and were provided with a Fitbit device and completed EMA of suicidal ideation, negative affect, substance use, and sleep disorder symptoms for 28 days. Most participants were Black (60.0%, n = 6; 20.0%, n = 2 White, and 20.0%, n = 2 identifying as Other). The sample identified as 70.0% (n = 7) men , 20% (n = 2) women, and 10% (n = 1) as bigender. Additionally, participants identified as 50.0% (n = 5) straight, 30.0% (n = 3) gay/lesbian, 10% (n = 1) bisexual, and 10% (n = 1) as sapio/pansexual.
Results: Participants had significant variability in suicidal ideation. Average daily range in suicide urges across all participants was 6.29 (SD=9.63) units, with a maximum range of 73. Intraclass correlation for range in suicidal urges was .341. The most frequently endorsed timing of the highest suicidal urge was 3:00 AM. There was significant variability in associations between affect and suicidal urges across individuals, with some participants demonstrating negative associations and others demonstrating positive associations. In qualitative interviews that were transcribed and double-coded, participants reported that the assessments improved their insight into triggers for suicidal ideation and prompted them to take action to reduce their risk.
Discussion: EMA and wearable device assessments for active and passive assessment of suicide risk are feasible and acceptable among PLWH. Participants demonstrated wide variability in suicide urges during the study, with some participants experiencing a 73% change in suicide urges on a given day. The middle of the night was a high-risk time for suicide urges and should be a focus of suicide prevention efforts.