Suicidal Ideation Response Guidance

Suicide is a leading cause of death in the United States, with 45,979 deaths in 2020. The University of Massachusetts Amherst Institutional Review Board (IRB) supports research initiated by investigators aimed at better understanding the factors associated with self-harm, suicidal ideation, and suicide. At the same time, the IRB has an important role in advocating for the health and safety of research participants. IRB deliberations must carefully weigh the importance of the research in this area and our desire to avoid placing barriers to suicide research with participant welfare and the safeguarding of vulnerable populations. The IRB often reviews research protocols utilizing surveys and “questionnaires focused on mental health, psychological functioning, and mood. Some of those studies focus directly on suicidal thoughts and behaviors or include one or more individual questions about suicidality. Research protocols that involve the identification of suicidal ideation in participants raise concerns about potential risks for those participants. This guidance gives the general action plans that will be required for all studies that involve questions about suicidality, but is particularly meant to support researchers who are not experts in the specific field of depression and suicidality.   

Action Plan

The centerpiece of any study which includes prompts regarding self-harm, suicidal ideation, or suicide will be an IRB-required “Action Plan.” The IRB will require that you develop an “Action Plan” addressing several contingencies in order to provide support and resources to any participant disclosing that they may be at risk of imminent harm. “Imminent risk” must be defined by the researcher and refers to a heightened and concerning desire or intent to self-harm; as such, the definition of “imminent risk” may vary depending on the measure or question(s) administered by the researcher, but the IRB will consider daily or nearly daily thoughts of suicide or the existence of a plan to carry it out as imminent risk requiring action. If researchers are unsure how they would identify imminent risk, it is advisable to work with an expert in the field of suicide studies and suicidal intent. While the IRB is also concerned when a participant expresses any suicidal ideation, it is imperative that timely action is taken when the participant discloses imminent risk of suicide or other self-harm.

The “Action Plan” should at a minimum address the following points:

  1. Ensure that the question(s) asking about self-harm or suicidal ideation are reviewed as soon as possible but no longer than 24 hours after the participants’ response being captured.
  2. Ensure that anyone who expresses a high degree of suicidal ideation or self-harm is identified (when participants are identifiable) and provided help. If the participants are students and with the researcher, the help should be really hands on and immediate (consulting the Maroon folder, calling emergency services, walking them to the CCPH if necessary, etc.). In the case of remote methods (surveys, etc), there should be outreach to the participant and services sought for them.  The plan should also mention whether confidentiality will be broken in such instances and the consent documents should disclose under what circumstances confidentiality will be broken.
  3. If the participants are not students, they should be referred to mental health resources, and if the threat is considered imminent, confidentiality should be broken, and emergency medical services engaged. All consent documents should also make clear that in these instances confidentiality cannot be assured, and mental health or other medical services may be alerted.

General Considerations:

Including Suicide Measures:

  • If using a measure such as the Beck Depression Inventory or the Patient Health Questionnaire-9, consider whether the questions about suicidality are necessary; of course, such questions are important for research on depression and suicidality, but for studies that are not focused on these topics, the researcher may decide that either another measurement would be a more appropriate “fit” for their study, or may consider whether individual questions can potentially be omitted from standard instruments if they are not relevant to the study, with the rest of the instrument being used.  These decisions are ones for the researcher to consider. When considering altering an instrument, check the copyright status of the measure to see if it can be modified and consider whether the removal of questions would undermine the instrument’s validity as a diagnostic tool for the research at hand.

Populations Involved:

  • Consider the populations involved in the study; certain groups have higher likelihood of experiencing suicidal ideation and/or intent.  These populations may require additional protections, and/or additional training to ensure cultural sensitivity.  If working with these populations and asking personal/sensitive questions, such as questions about personal history and emotional states – even if suicide is not explicitly asked about – a clear Action Plan should be outlined in the protocol’s “Risks and Risk Management” section, and the informed consent document should make it clear that confidentiality may be violated if a risk to self or others is disclosed.
  • If children are being asked questions involving suicide, the parental consent form and child assent form should both contain all of the following: the fact that there will be questions about suicide; the Action Plan for imminent risk, if the child indicates being suicidal; and when/how parents/guardians will be informed of a finding of such imminent risk.  If working in partnership with schools, additional precautions and support services might be necessary, and any FERPA regulations that would be enacted should be considered.
  • For actively suicidal populations, several additional factors may be considered including ability to consent (this may be dictated by state law in cases where individuals are involuntarily committed to psychiatric facilities or hospitals) and concerns about access to firearms if and when imminent risk of suicide has been determined (researchers should check ahead of time for ‘red flag’ laws within the state at the time of the research study).

Databases:

  • Note that questions about suicidal ideation and intent asked on subject pool screening forms/databases such as SONA would still need to be checked within a 24 hour period; for this reason, most databases choose not to collect such data.

Research Pre-planning:

  • Although research has shown that asking about suicidality does not increase risk, studies that include questions about suicide and assessments of suicidality require additional monitoring of participants’ responses for their safety. Keep in mind that for UMass Amherst research studies, non-anonymous surveys and interviews with questions about suicide must be checked within 24 hours; this includes both weekdays and weekends. This policy is in line with other major institutions. Please take this into consideration when planning for staffing, hourly work, vacation time, etc.
  • For studies that involve technological data gathering and/or interventions around suicidality, the protocol to the IRB should include a summary of how the system has been tested prior to deployment (beta tests with software, stress tests, field tests with hardware, etc) as well as how the systems will be regularly monitored while the study runs (i.e., regular accuracy checks and a plan to identify and correct problems in a timely fashion).

NIH Considerations:

  • NIH funded studies of suicide may require additional staffing (such as medical monitors and/or Data and Safety Monitoring Board experts) and data sharing plans.

Anonymous Surveys with Questions about Suicide:

  • The informed consent document/page should make it clear that researchers will not be able to link responses to any individual, and will not be able to reach out to provide any support services if suicidal ideation is reported
  • Informed consent must let participants know that questions about suicide will be asked
  • A list of support services should be provided to every participant at the beginning of the survey (and potentially at the end of the study as well, if the researcher would like to provide them a second time)
  • If the survey is lengthy (15 minutes or longer) and asks about suicidal ideation at different points, it is advisable to provide periodic links to support services throughout the survey
  • The researcher should be prepared to offer, in the protocol, a justification for why the benefits of asking such questions in an anonymous survey outweigh the risks

Identifiable Online Surveys with Questions about Suicide:

  • Informed consent/page must let participants know that questions about suicide will be asked
  • The informed consent document must state that confidentiality may be broken if suicidal intent is disclosed, and that outreach and an assessment from the research team will follow; describe the procedures if a participant cannot be reached for assessment
  • All submissions of the survey must be reviewed for suicide-related questions within 24 hours of their submission
  • The protocol must describe the score at which a participant is considered at imminent risk of suicide, who will do this evaluation and what their training background is, how the participant will be reached, and what the Action Plan will entail
  • For non-imminent cases, a list of resources may be shared directly with the participant
  • For imminent cases, a clear Action Plan must be submitted in the protocol and outlined in the informed consent document

Non-Anonymous In-Person Surveys and Interviews

  • Informed consent/page must let participants know that questions about suicide will be asked
  • The informed consent document must state that confidentiality may be broken if suicidal intent is disclosed and what the treatment plan would entail
  • When partnering with clinical systems, the informed consent document and protocol should outline how patient information is shared and how a positive risk screen will be handled
  • Responses to any survey items regarding suicidality and depression must be reviewed before the participant physically leaves the space
  • The protocol must describe the score at which a participant is considered at imminent risk of suicide, who will do this evaluation and what their training background is, and what the Action Plan will entail
  • For non-imminent risk, a list of resources should still be shared with the participant; if working with UMass Amherst students on campus, an offer should be extended to connect them with staff at CCPH, by email/referral or by walking them there
  • For imminent risk, the researcher must remain with the participant until official support services have been received; for students on campus, this should entail walking them directly to CCPH; off campus and for non-students, this might entail emergency services.  Note that having a family member or friend come to the participant, on its own, is not an official support service.

Reportable Events

  • A disclosure of suicidal intent may or may not count as a reportable event depending on the aims of the research study; for instance, a study whose primary research questions are around suicidality and involves only populations at a known high risk of suicide (such as individuals with recent suicide attempts) would not report suicidal ideation as an event, whereas a study about family histories wherein a participant disclosed strong depressive feelings and suicidal intent would, as it is “unexpected” in that study context; this would be reported if there is a concern that the suicide attempt is linked to the subject’s involvement in the study

Depression Guidance:

  • Some studies will ask questions about depression, but not specifically suicide. For identifiable participant responses to standardized measures (for instance, the Beck Depression Inventory), consider what responses will trigger a concern about depression, who will assess this, and what their training is. For mild depression, no intervention from the researcher is necessary. For moderate to severe depression findings, a list of resources should be provided. The researcher may choose to provide resources to all participants. The researcher can compile a list that is most relevant for the study population (for instance, UMass Students may be directed to CCPH, the Women’s Center, the Stonewall Center, etc; national studies may reference the 988 hotline, and so forth). Informed consent forms for studies with questions about depression should indicate that the researchers are not clinicians and will not be following up about participants’ possible depression scores.

References: