10.2 Suicide Risk and Safety Planning

Key Takeaways

  • Ask about suicide directly and specifically: ideation, intent, plan, access to means, prior attempts, and acute modifiers (substance use, agitation, command hallucinations) — direct questioning does not increase risk.
  • A collaborative safety plan follows the six-step Stanley-Brown model and replaces the discredited 'no-suicide contract,' which has no evidence of preventing suicide.
  • Means restriction — especially reducing access to firearms — is the single most powerful safety-plan step and is placed last in the Stanley-Brown sequence.
  • Acute risk markers (current intent, specific plan, available lethal means, recent attempt, inability to commit to safety) point toward emergency evaluation or a higher level of care.
  • Protective factors mitigate but never cancel acute risk; never let 'reasons for living' override evidence of current intent and access.
Last updated: June 2026

Direct, structured assessment

When a client signals risk — hopelessness, wanting to "disappear," feeling like a burden, a recent loss, severe agitation, intoxication, or a prior attempt — the counselor must ask about suicide directly. Decades of research confirm that asking does not plant the idea or raise risk; avoidance is the real danger. The skill is pairing attunement with structure: reflect the pain, then ask plainly, "Have you had thoughts of killing yourself?"

Structured tools organize the data. The Columbia-Suicide Severity Rating Scale (C-SSRS) screens ideation severity (from passive wish to active intent with plan) and behavior (attempts, aborted/interrupted attempts, preparatory acts). The SAFE-T approach pairs assessment of risk and protective factors with a triage and documentation step. You do not need to recite item numbers, but you must recognize the data domains these tools cover.

Suicide risk data to gather

DomainSpecific factsClinical use
IdeationFrequency, intensity, duration, controllabilityEstablishes current severity
Intent and planWish to die, plan specificity, timeframeIdentifies acuity and urgency
Means and accessFirearms, medications, other lethal means at handDrives means-restriction steps
HistoryPrior attempts, aborted attempts, preparatory acts, self-harmA prior attempt is the strongest single predictor
Acute modifiersSubstance use, agitation, insomnia, command hallucinations, recent dischargeRaise volatility and lower threshold to escalate
Protective factorsReasons for living, supports, beliefs, responsibilities, engagementMitigate but do not erase acute risk

A prior suicide attempt is the single strongest long-term predictor of death by suicide, and the highest-risk window is the days to weeks after a psychiatric discharge. Flag these facts when a case includes them.

The Stanley-Brown safety plan and means restriction

The evidence-based intervention is the Stanley-Brown Safety Planning Intervention (SPI), a written, collaborative, prioritized plan adopted by the U.S. Veterans Affairs system. It has six steps, completed with the client:

  1. Recognize warning signs — personal thoughts, images, moods, or situations that signal a crisis is building.
  2. Internal coping strategies — things the client can do alone to distract or self-soothe (walk, music, breathing) without contacting another person.
  3. Social contacts and settings that distract — people and places that take the client's mind off the crisis.
  4. People the client can ask for help — family or friends the client can tell about the crisis.
  5. Professionals and agencies — clinician numbers, the 988 Suicide & Crisis Lifeline, and the nearest emergency department.
  6. Making the environment safe (means restriction) — reduce or remove access to lethal means.

Means restriction is the most evidence-supported step because most suicidal crises are brief and ambivalent; putting time and distance between the person and a lethal method saves lives. It is placed last deliberately: once the client has alternatives in steps 1–5, they engage more willingly in restricting means. For low-lethality methods the client may secure them; for a firearm, a responsible third party (family member, friend, or law enforcement) should store it off-site or lock it.

The "no-suicide contract" (a promise not to self-harm) is not evidence-based, does not reduce suicide, and provides no legal protection. On the exam, never choose it over assessment and a real safety plan.

Matching response to acuity

  • Lower acuity (passive ideation, no plan, no access, strong supports, willing to engage): build the safety plan, restrict means, increase contact frequency, involve supports with consent, and consult.
  • Higher acuity (current intent, specific plan, available lethal means, recent attempt, intoxication, command hallucinations, or inability to commit to safety): move toward emergency evaluation or a higher level of care per policy. If the client meets criteria for danger to self and refuses, involuntary evaluation may apply.

Apply cultural humility: shame, family roles, spiritual beliefs, and help-seeking norms vary, so ask how each protective factor actually operates for this client rather than assuming. Finally, follow-up is part of the intervention — a plan with no scheduled contact, review, or documentation is incomplete.

Static vs. dynamic factors and warning-sign recognition

Strong suicide reasoning separates static risk factors (fixed history that cannot change) from dynamic factors (current, modifiable states). Static factors include a prior attempt (the strongest single predictor), family history of suicide, chronic illness, and history of trauma or abuse. Dynamic factors include current ideation intensity, hopelessness, agitation, insomnia, intoxication, recent loss, and access to means. Static factors set the baseline; dynamic factors drive the acute decision and are the levers a safety plan actually moves.

Distinguish a risk factor (raises long-term probability, e.g., male sex, prior attempt) from a warning sign (signals near-term crisis, e.g., "I can't go on," giving away possessions, sudden calm after agitation, searching for means). A sudden, unexplained calm in a previously agitated suicidal client is an ominous warning sign — it can mean the person has decided on a plan, not that the crisis has passed.

Protective factors do not cancel acute risk

Protective factorReal valueExam caution
Reasons for living / responsibility for childrenBuffers and supports planningDoes not erase current intent + means
Strong social supportEnables outpatient safetyUseless if the client will not contact supports
Religious/cultural prohibitionMay reduce riskVaries by person; ask, do not assume
Engagement in treatmentPredicts adherence to planIrrelevant if acute intent is present right now

The trap is to let a protective factor ("but she loves her kids") override evidence of current intent, a plan, and available means. Acute risk wins. When the data show acute risk, you escalate regardless of how many protective factors are listed.

Documentation specific to suicide risk

Document the risk level (e.g., low/moderate/high acute), the specific data supporting it, the safety plan created, means-restriction steps taken (including who will secure firearms), supports involved, consultation, the level of care chosen, and the follow-up timeframe. A defensible suicide note shows your reasoning, not just your conclusion — it demonstrates that the chosen action was proportionate to the assessed risk.

Test Your Knowledge

A client says, "My family would honestly be better off without me." What should the counselor do next?

A
B
C
D
Test Your Knowledge

Within the Stanley-Brown Safety Planning Intervention, why is reducing access to lethal means placed as the final (sixth) step?

A
B
C
D
Test Your Knowledge

Which combination of facts most strongly supports emergency evaluation or a higher level of care?

A
B
C
D
Test Your Knowledge

A counselor proposes that a moderately suicidal client sign a written promise that they will not hurt themselves before the next session. What is the problem with this choice?

A
B
C
D