Suicide Risk Assessment: What the Conversation Actually Sounds Like
Most therapists were trained on the elements of suicide risk assessment. Ideation, plan, means, intent. History of attempts. Protective factors. These domains are not wrong — they are the right things to cover. What training often leaves therapists less clear on is what the assessment actually sounds like as a conversation, how to translate the clinical picture into a risk level, and what to do with the ambivalence that characterizes most suicidal clients.
Opening the conversation
The first clinical decision in a suicide risk assessment is how to open it. Many therapists avoid asking directly because they are worried the question will plant an idea or alarmed the client. The research is consistent: asking directly about suicidal ideation does not increase risk. It often provides significant relief.
“I want to ask you directly — are you having thoughts of suicide or of hurting yourself?”
“Sometimes when people are going through what you’re describing, thoughts of not wanting to be here come up. Has anything like that happened for you?”
The second version is useful for clients who might not identify their experience as “suicidal ideation” but who have passive thoughts — wishing they were not here, wishing they would not wake up. Passive ideation is not the same as active planning, but it is clinically significant and worth assessing fully.
The domains and what they tell you
Ideation type — passive (wish to be dead) versus active (thinking about killing themselves), and whether the active ideation has specificity about method. More specific ideation is more concerning, but passive ideation is not nothing. A client who says “I wish I could just go to sleep and not wake up” deserves the same careful assessment as one with an articulated plan.
Plan — whether the client has thought about how. Plan specificity matters: vague (“I don’t know, just something”) is different from researched and specific (“I’ve looked into it, I know how”). The presence of a specific plan significantly raises the clinical response.
Means — access to the method they have identified. Means restriction — putting distance between the client and the method, whether that is firearms stored with a family member or medications dispensed in smaller quantities — is a direct safety intervention that does not require hospitalization. It is underused.
Intent — whether the client is planning to act. Most suicidal clients are ambivalent. The part that is sitting in your office telling you about the ideation is evidence of ambivalence. Working with that part, rather than the part that wants to die, is usually the therapeutic entry point.
History — prior attempts are the single strongest predictor of future risk. A recent attempt (within 90 days) substantially elevates risk. Family history of suicide by death also elevates risk.
Protective factors — reasons for living, relationships, beliefs, plans, fear of death, connection to treatment. Protective factors that are strong and genuine can support an outpatient management plan. Their absence in the context of active ideation warrants more significant intervention.
Current context — what is happening right now. Recent loss, intoxication, recent discharge from hospitalization, anniversary of a significant loss, legal or financial crisis, social isolation — these are all risk-elevating context factors.
Translating the picture into a response
Risk stratification is not a formula. It is a clinical judgment that weighs all of these domains together for this specific person in this specific moment.
Lower risk — passive ideation without plan, strong protective factors, no access to means, engaged in treatment — can generally be managed with a collaborative safety plan, increased contact frequency, and thorough documentation.
Moderate risk — active ideation with some planning, ambivalence about intent, moderate protective factors — warrants same-day supervisor consultation, a carefully built safety plan, means restriction, and assessment of whether current level of care is adequate.
High risk — active ideation with specific plan and means, low protective factors, poor engagement with treatment, history of prior attempts — requires immediate consultation and likely involves higher level of care. A client who is high risk should not leave the session without a plan that holds, or without initiating a higher level of care.
If you want a complete visual decision tree for suicide risk assessment — covering all seven domains, risk stratification criteria, safety plan format, documentation language, and the Columbia Protocol — browse the therapist decision guides library.