We Need to Talk: A Multilevel Theory-Informed Examination of Suicide-Related Disclosure in Social Networks

Background/Purpose: Many people experience but do not disclose their suicidal thoughts, which can lead to more deaths by suicide. This suicide risk disproportionately affects certain vulnerable groups, such as people with mood disorders. Pervasive atheoretical and person-centric approaches represent serious impediments to developing programs that facilitate suicide-related disclosure. The Disclosure Decision-Making Model (DD-MM) is a health/mental health information management model that holds promise for but has yet to be applied in a suicide context. This model is comprised of three components (Information; Receiver; and Efficacy) and can be operationalized within a conceptually appropriate multi-level framework—accounting for information about the individual (i.e., Information Component) and their relationships (e.g., Receiver & Efficacy Components) to understand disclosure. The current study investigated associations between the DD-MM components and suicide-related disclosure intent.

Methods: The current study sample was comprised of 41 adults with a mood disorder receiving traditional outpatient psychiatric care services at a large community-based mental health agency that serves people with serious mental illness. Self-report questionnaires and a social network interview were used to collect information about participants (individual-level data) and their social networks (relational-level data), respectively. We assessed all DD-MM constructs—including those that constitute the Information (i.e., Stigma; Symptom Severity), Receiver (i.e., anticipated responses/outcomes, relational quality), and Efficacy Components—as well as other pertinent control variables (e.g., race-ethnicity; social support). A series of univariable and multivariable multi-level regression analyses was used to examine associations between DD-MM constructs and suicide-related disclosure intent. Multilevel modeling was used due to clustering within personal networks.

Results: Constructs of all DD-MM components were significantly associated with disclosure intent in univariable regression models. The same largely held true for the multivariable model. For the Information Component, significant associations were found between symptom severity (b=.21, p=.004) and stigma toward suicide (b=-.43, p=.043) and disclosure intent. For the Receiver Component, significant associations were observed between both positive relationship quality (b=.40 p<.001) and positive anticipated outcome (b=.15, p=.002) and disclosure intent. For the Disclosure Efficacy Component, a positive association was found between disclosure efficacy and disclosure intent (b=.64, p<.001).

Conclusions/Implications: The decision to disclose suicidal thoughts is a poorly understood process that is consequential for managing suicide risk. Our current study begins to fill a conspicuous gap in our theoretical understanding of suicide-related disclosure. Furthermore, our results provide preliminary but much-needed guidance for the development of a suicide-related disclosure intervention. There are many different ways that the identified DD-MM correlates (i.e., stigma, relational quality, anticipated outcomes and disclosure efficacy) can potentially be addressed in adapted versions of existing interventions. For example, the Brief Disclosure Intervention—a group intervention that uses a motivational interviewing approach to help participants learn about techniques for sharing, contemplate the pros and cons of sharing, and practice sharing—is part of a recent wave of disclosure-related interventions and decision-aids developed for people with different stigmatized statuses (e.g., HIV; mental illness). Additionally, a strong case can be made for engaging people in the social networks of vulnerable groups to address DD-MM components (e.g., promoting positive responses/outcomes).