What Should Researchers Do When a Participant Discloses Self-Harm or Suicide Risk?

A Participant Mentions Self-Harm or Suicide. What Should the Researcher Do?

A participant endorses suicidal thoughts on a questionnaire. During an interview, someone says they have been thinking about harming themselves. A participant who was not enrolled in a suicide study unexpectedly describes a recent suicide attempt.

Researchers cannot simply treat this as another data point. Yet neither should every endorsement automatically be interpreted as an imminent emergency.

The challenge is to respond proportionately: recognize potentially important risk, use an appropriate assessment and escalation process, protect confidentiality within its actual limits, and avoid asking untrained research staff to make clinical judgments they are not qualified to make.

Use a Predefined Safety Protocol and Match the Response to the Assessed Risk

Suicide Risk Requires a Safety Process, Not a Single Trigger Word

Asking about suicide means researchers must be prepared for an answer

Researchers sometimes include questions about suicidal ideation or self-harm because they are scientifically relevant to depression, trauma, substance use, chronic illness, adolescent health, or other topics. Other studies may identify suicidal thoughts unexpectedly.

NIMH guidance emphasizes that studies asking about suicidal ideation or intentional self-injury need procedures for responding to identified risk and providing clinical follow-up as necessary. This applies not only to suicide-focused studies but also to research with populations not initially identified as being at elevated suicide risk.

The ethical problem therefore begins during study design. If your questionnaire can produce an answer indicating possible suicide risk, you should know what happens when someone gives that answer before the first participant completes it.

A positive screening response is not the same as an imminent crisis

Suicidal thoughts exist across a wide range of severity and circumstances. A participant might report thoughts from months ago, passive wishes about death, current suicidal ideation, recent self-harm, or circumstances suggesting acute danger.

Researchers should not collapse all of these into one category. The research instrument may identify a signal requiring further assessment, but the instrument itself may not establish the participant's immediate level of risk.

This distinction helps prevent both underreaction and overreaction. Ignoring a meaningful signal can endanger a participant, while automatically treating every positive response as an emergency can produce unnecessary coercion, distress, or loss of trust.

The protocol should define who assesses risk

Not every research staff member should be expected to perform a suicide-risk assessment. NIMH advises research teams to consider what level of professional training is required for staff involved in assessing and responding to suicide risk and how those staff will be trained to follow the study's safety procedures.

A research assistant can be trained to recognize a protocol trigger and contact the designated clinician or safety professional without being asked to independently decide whether a participant is suicidal.

Researchers should have a graded response rather than one universal reaction

The exact categories depend on the protocol and appropriate clinical framework, but the basic principle is proportionality. Information indicating no current acute concern may require a different response from information indicating a potentially immediate threat to life.

Immediate safety can take priority over ordinary research procedures

When qualified assessment indicates an acute and serious risk, participant protection becomes the immediate concern. The study should already specify whom staff contact, what emergency resources are available, how remote participants are handled, and what information may need to be disclosed.

NIMH specifically recommends that studies involving suicidal or potentially suicidal participants establish procedures for suicidal crises and clinical worsening, determine appropriate safety-monitoring frequency, and define the professional training needed to respond.

This is an instance of the broader principle governing research information that may affect participant health or safety: the response should become faster as credible risk becomes more serious and time-sensitive.

Confidentiality has limits, but those limits should not be invented during the crisis

Research participants should understand before enrollment whether information indicating serious danger may trigger disclosure or intervention. This is particularly important in studies asking directly about suicide or self-harm.

In the United States, Certificates of Confidentiality provide substantial protections for identifiable sensitive research information. NIH's current policy nevertheless permits disclosures in circumstances specified by law and other statutory exceptions. Older OHRP guidance also emphasizes that investigators who intend to make voluntary disclosures involving threatened violence to self or others should describe those practices in the consent process.

The precise legal position depends on jurisdiction and research context. Researchers should therefore define confidentiality limits with their ethics committee or IRB and institutional counsel rather than assuming that either confidentiality or emergency disclosure always wins.

Do not promise secrecy when the protocol contains a safety exception

If a participant says, “Please don't tell anyone,” a researcher should not promise secrecy and then immediately break that promise. Nor should the researcher coldly recite consent language while the participant is distressed.

The appropriate response is to explain the actual limits of confidentiality accurately and compassionately, including what information may need to be shared, with whom, and why, to the extent the situation and protocol allow.

Research staff should not become improvised therapists

Researchers can listen respectfully and facilitate appropriate care without assuming a therapeutic role they are not trained or authorized to provide.

NIMH emphasizes the importance of clinical follow-up, safety planning, linkages to care, and avoiding situations in which at-risk participants leave research without needed care. For studies enrolling people with recent or active suicidal ideation or behavior, end-of-study planning should also consider participant safety and continuity of care.

Providing an appropriate referral is therefore different from attempting to conduct therapy during a research interview.

Remote and digital research needs its own safety architecture

Online surveys, video interviews, smartphone studies, social-media research, and automated risk-detection systems complicate safety response. A researcher may identify concerning information without knowing whether the participant is still online, where they are located, or whether anyone can reach them.

NIMH highlights these problems in research using emerging technologies, including questions about how automated systems identify suicide risk, what happens when technology fails, how participants are contacted, and whether appropriate interventions are actually available.

A digital study should therefore not collect real-time suicide-risk information unless the research team has deliberately considered what it can and cannot monitor and what participants are told about that capability.

Asking about suicide is not itself generally supported as harmful

Researchers and ethics reviewers sometimes worry that asking about suicide will cause suicidal thoughts or behavior. NIMH cites evidence, including a meta-analysis of studies examining assessment of suicidality, that did not find evidence of significant iatrogenic effects from asking about suicide.

That does not eliminate the need for safeguards. It means avoiding suicide questions solely because asking is assumed to create suicidality is not well supported by the evidence NIMH reviews.

Safety response and study continuation are separate decisions

Identifying suicide risk does not necessarily mean the participant must automatically be removed from the study. In some research, suicidal ideation is part of the population or outcome under investigation.

NIMH recommends procedures and oversight for deciding whether changes in clinical status affect continued research participation, with documentation sufficient for study monitors or data and safety monitoring boards to review those decisions where appropriate.

Automatically excluding everyone who reports suicidality can also undermine the validity and relevance of research intended to understand or help precisely those populations.

A Participant Endorses Suicidal Thoughts During a Non-Suicide Study

The questionnaire item identified a reason to assess. It did not itself make the clinical determination.

Common Mistakes When Suicide Risk Emerges During Research

If Your Study Can Detect Suicide Risk, It Needs a Response Pathway

Before collecting data, determine exactly what information triggers further assessment, who receives the alert, who performs the assessment, what happens at different levels of risk, how remote participants are handled, and what confidentiality limits participants need to understand.

The central safeguard is preparation. Researchers should not be constructing a suicide-response protocol while a distressed participant waits in the interview room.

Before Collecting Information About Suicide or Self-Harm

Questions About Suicide Risk During Research

A Suicide-Risk Signal Requires a Process, Not an Automatic Conclusion

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.