Guidance on Conducting Research Involving Mental Health Topics

In human subjects research, many studies utilize questionnaires and assessments that address mental health, behavior or quality of life. These may include questions and assessments of the subject’s intent to harm him/herself or others. As part of its responsibility to protect research subjects, COUHES evaluates the risks and benefits of asking study participants about these topics.

This guidance document is intended to help researchers determine the appropriateness of including mental health topics in their research and considerations in developing acceptable plans for mitigating any potential risk.

Identifying Risk of Harm

Intentional Identification

Studies may be designed to gauge intent to harm self or others. This may be done through assessments such as behavioral evaluations, interviews, surveys or other measures for depression and suicidality.

When using such clinical diagnostic measures, researchers should consider if the study setting and population are appropriate with regards to the safety, risk-benefit ratio, and knowledge to be gained. If these measures identify study participants as clinically at-risk, or the study targets populations at high risk for injury to self or others, the COUHES application must include a a safety plan that describes what actions will be taken to ensure the safety of at-risk participants.

Unintentional Identification

For studies that include quality-of-life assessments or questionnaires that ask about sadness, anxiousness, or stress, the identification of a participant at-risk for harm to self or others may be unintentional. This is because these assessments are not typically designed for diagnostic purposes.

COUHES evaluates if a safety plan is needed for such studies on a case-by-case basis. A plan may be required if responses reveal acute risk (e.g. imminent danger to self or severe low mood) or if the study population is at elevated risk (e.g. receiving mental health treatment for depression, mood, or anxiety disorders).

In-Person vs. Remote Identification

When conducting research related to mental health, researchers should consider the physical environment where the study procedures will take place, and how participants will be adequately protected in that environment. Safety and ethical considerations can differ depending on whether the participants complete the intervention remotely or in-person, and whether the researchers know the identity of the participant or study participation is anonymous.

Creating a Participant Safety Plan


For studies that may identify a participant is at immediate or emerging risk for harming self or others, researchers must develop a safety plan.

The key to a safety plan is an assessment of how imminent is the risk. How this assessment is made can vary depending on: whether study procedures are carried out remotely or in-person; in a clinical, laboratory, or other setting; or whether the information is collected anonymously or not tied to an individual.

The COUHES application must explain:

  1. How the risk will be assessed;
  2. When investigators will review a subject’s response to questionnaires and assessments, and the frequency at which this review will occur; and
  3. By whom the level and immediacy of risk will be assessed.

If participants’ responses will not be individually assessed, the COUHES application should explain why the investigators believe an individually identifiable assessment will not be included.

Any researchers administering the clinical measures and assessments, or reviewing a subject’s responses, should be appropriately qualified to assess the measures and assessments, and be familiar with the safety plan. In particular, individuals that are assessing participants’ risk of harm to self or others must have appropriate training in the assessment and implementation of the safety plan.


Safety plans described in a COUHES application may include the follow, as appropriate:

  1. If participants will be provided mental health or other resources, a copy of the resource referral document, and an explanation of how/when the resources will be made accessible to participants. 
  2. For assessments that determine imminent risk, procedures on transferring the participant to appropriate crisis intervention or de-escalation resources.
  3. For assessments that determine less than imminent risk, referral or intervention procedures and how this information is communicated to participants.
  4. For research that gathers anonymous information or where responses are not tied to an individual the plan should provide for a resource referral document to be given to participants that includes mental health resources, crisis intervention services, or hotline information depending on the type of risk.
  5. Qualifications of researchers and/or clinicians involved in participant interactions, assessments, and safety interventions.
  6. Confirmation that research staff directly interacting with participants will be adequately trained on the safety plan.
  7. If participants and parents/legal guardians, as applicable, will be notified of findings.
  8. If information is reportable under state or federal law, plans for notifying the relevant authorities and/or agencies under mandated reporting requirements.

Consent Process

As part of the consent process, consent forms should:

  1. Clearly explain to potential participants the sensitive nature of any interviews or questionnaires;
  2. Describe what will happen if participants acknowledge/disclose harm to self or others;
  3. Have information on the risks and benefits of participating in this type of research. For example, for research with populations at elevated risk, include that the involvement in the research does not provide participants with “protection” against future harmful behavior, and how potential risk is mitigated; and
  4. Include any limitations on data confidentiality in the Privacy and Confidentiality section. If information collected during research must be disclosed under mandated reporting requirements, this must be included. Language regarding mandated reporting may be reviewed by MIT Office of General Counsel.

Appendix: Suggested Consent Language

For research that gathers anonymous information and researchers plan only to provide resources, the consent process should not lead participants to think that the researchers will provide immediate assistance. Suggested language to add in the risk section is as follows:

There are no anticipated risks from your participation in this study. However, some people become anxious or upset when answering questions about (behaviors, well-being, mood, views). Your responses will not be individually identified, so we cannot provide you with personal feedback or intervention based on any of your answers. If you are worried about your mood, please refer to the attached resource referral information sheet.

If responses will be individually assessed and can be linked back to participants, the consent form should explain what options the participant will have if they become upset or uncomfortable during study activities. For example:

In the event that you tell the research team you are thinking about harming yourself or others, the research team will provide you resources and may ask you more questions about these thoughts. Based on your responses, the research team may provide you will additional resources or assistance to identify appropriate follow-up. This may include working with you to contact your doctor, contacting a trusted family member or therapist to discuss your thoughts, or working with you on a plan that may include getting you to a hospital for safety.

When using clinical diagnostic or symptom severity measures, participants scores above a pre-defined threshold of the measure should be reported back to the participant with an offer for referrals and/or counseling resources. The threshold for intervention needs to be defined in the COUHES application along with when and how the study findings will be shared with participants. Researchers should be prepared to offer appropriate counseling resources, assistance in making appointments, and/or offering a list of referrals. An example email message is below:

I am part of the team for a research study you recently completed. Based in your responses to some of the questions we asked, you seem to be experiencing (sadness, stress, blue moods, etc.). We provided you some information about mental health resources, but I wanted to follow-up and offer any other information you might want to get help.


To assist investigators with identifying counseling resources, COUHES provides the following:

Conducting Research on Mental Health Topics - Participant Counseling Resources [PDF]

The links below provide additional guidance for engaging participants on the topic of mental health.