Passive Suicidal Ideation Is Not a Disposition

An established outpatient says, “Sometimes I wish I wouldn’t wake up, but I would never kill myself.” The note records passive suicidal ideation, denies plan, denies intent, safety plan reviewed, follow-up in several weeks.

This fictional composite is not a patient case. It is a thin disposition: a label, two negations, a form, and a calendar slot. The record still has to answer a harder question:

A denial of current plan or intent is one finding. It is not, by itself, a level-of-care decision.

Clinicians need language for what the patient is thinking and doing. CDC surveillance definitions treat suicidal ideation as thoughts of engaging in suicide-related behavior and distinguish ideation from intent, attempt, and death. They do not turn “passive” versus “active” into a disposition rule.

Clinical instruments do draw that line. The Columbia-Suicide Severity Rating Scale starts with a wish to be dead—not alive, or to fall asleep and not wake up—then asks about active thoughts of killing oneself, method, intent, and a specific plan, and separately records preparatory acts and actual, aborted, or interrupted attempts. The APA’s 2016 adult evaluation guideline asks about current ideas, plans, and intent, including active or passive thoughts of suicide or death, plus prior ideas, plans, attempts, and intentional self-injury without suicidal intent.

Those distinctions are useful. They are incomplete as a decision rule.

Intent is whether the patient expects to act and believes the act would be lethal. A plan is more than a passing method: timing, location, lethality, availability, and preparatory work. Preparatory behavior is action toward an attempt, not merely thinking about a method. Recent suicidal behavior includes attempts, aborted or interrupted attempts, and other suicidal self-directed violence. Passive ideation names thought content. It does not grade the rest of the state.

It also does not tell you who will die. Franklin and colleagues, across 365 longitudinal studies, found individual risk factors predicted suicidal thoughts and behaviors only slightly better than chance, with no improvement over fifty years. Large and colleagues found about half of later suicides among patients not labeled high-risk, and that most patients so labeled did not die by suicide over years of follow-up. Belsher and colleagues found prediction models often classify the non-event well and still have extremely low positive predictive value for suicide death. VA CPG implementation materials state the operational limit: assessments have poor predictive value but can identify contributing factors for treatment planning. The 2024 Work Group found insufficient evidence to recommend a specific tool or method for determining level of risk. Assessment is for today’s formulation and level-of-care decision. It is not a forecast.

The question is not “Does the patient have a plan?” It is “Why is this level of care appropriate for this patient today?”

SAMHSA’s SAFE-T is a clinician sequence, not a score that disposes the case: identify modifiable risk factors; identify protective factors that can be enhanced; inquire about thoughts, plans, behavior, method, and intent; determine risk and choose an intervention; document the assessment, rationale, intervention, and follow-up. The APA initial-evaluation statements cover current ideation, plan, and intent; prior attempts and self-injury; hopelessness, impulsivity, anxiety, substance use, psychosis, and stressors; and, when ideas are present, the intended course if symptoms worsen, access to methods including firearms, motivations, and reasons for living—then an estimate of risk and the factors influencing it. The 2024 VA/DoD guideline suggests assessing self-directed-violence thoughts and behaviors, psychiatric conditions and treatment, symptoms, social determinants and adverse events, lethal means, physical health, demographics, and protective factors.

Integrate what is actually in front of you: current ideation (content, frequency, intensity, duration, controllability); recent trajectory, not only today’s denial; intent and planning, including rehearsal; access to means, especially firearms; preparatory behavior; prior attempts, including aborted or interrupted attempts; recent self-harm; substance use, intoxication, and withdrawal; agitation, psychosis, command hallucinations, mania or mixed features; severe insomnia and hopelessness; acute losses; adherence and the ability to use a safety plan; follow-up reliability; and supports that are functioning now.

For Joint Commission–accredited behavioral health organizations, NPSG.15.01.01 requires an evidence-based assessment that asks about ideation, plan, intent, suicidal or self-harm behaviors, risk factors, and protective factors, plus documentation of overall risk and a mitigation plan. That is an accreditation requirement for organizations under that program—not a nationwide legal standard for every outpatient note, and not a rule that a screening score is a disposition.

A present-tense denial is one data point. Direction of travel is another.

Chronic, unchanged passive thoughts with stable supports, sleep, substances, and follow-through are not the same state as new ideation after a job loss, breakup, medical diagnosis, or medication interruption. Progression toward method, intent, or preparatory behavior is a different problem than a long-standing wish that has not moved. Denial today after yesterday’s overdose, rehearsal, or farewell behavior is a discrepancy, not reassurance. Ideation that fluctuates with intoxication, withdrawal, or impulsivity is not captured by one “denies plan/intent” line.

Write what changed, over what interval, and whether the current statement is consistent with that course. A checkbox cannot do that work.

A patient can deny a current plan and still present with a recent attempt, preparatory behavior, agitation, intoxication, psychosis, command hallucinations, mixed or manic features, marked impulsivity, inability to engage in planning, severe hopelessness, an unreliable history, a sudden loss of support, or no timely follow-up.

Those findings increase the reasoning the disposition must show. They do not automatically mandate hospitalization. The error is the opposite shortcut: treating the denial as the conclusion and leaving the rest of the state undocumented. If the clinician is sending the patient home, the record should show why those concerns are judged manageable in this setting, with this plan, today.

“Children, job, religion” listed under protective factors, then “low risk,” is not a formulation. The APA asks about reasons for living when ideas are present. SAFE-T asks which protective factors can be enhanced. The VA/DoD list includes connectedness, problem-solving, spirituality, purpose, employment, and access to care. None treat a template list as a subtraction from risk.

A child is not protective if the patient is isolated from that child. A job is not protective if it was lost last week. A religious affiliation is not protective if the patient no longer has a community or no longer believes it forbids the act. A safety plan on file is not protective if the patient cannot use it.

Safety planning, lethal-means counseling, crisis instructions, emergency escalation, follow-up timing, and the level-of-care decision are different acts. The Stanley-Brown Safety Planning Intervention is a brief collaborative process: warning signs, internal coping, social distraction, people who can help, professional and crisis contacts, and limiting access to lethal means. Completing the form is not a disposition.

The 2024 VA/DoD guideline found insufficient evidence to recommend for or against a crisis response plan or safety-planning intervention as a method of reducing suicide attempts. The Work Group still treats safety planning as routine risk-mitigation work in VA and DoD care, not as proof that the setting is appropriate. A 2018 VHA emergency-department cohort study found that safety planning plus structured telephone follow-up was associated with fewer subsequent suicidal behaviors and more outpatient engagement than usual care. That was a combination intervention, not a randomized trial, and not a finding that a signed plan replaces a disposition.

No-suicide contracts are a different object. Reviews have found no empirical support for them as a risk-reduction method. One randomized trial in active-duty soldiers found crisis response planning outperformed a contract for safety; that does not establish a universal effect size, and it does not revive the contract as documentation of safety.

The same visit can be written two ways.

Passive SI. Denies plan. Denies intent. Safety plan reviewed. Follow up in four weeks. Low risk.

Wish not to wake up for three days after a relationship ended; no method, intent, or preparatory behavior elicited today; last attempt two years ago, overdose. Sleep four hours; increasing alcohol this week; not intoxicated now. Patient reports firearms are currently stored outside the home and are not accessible to them. One person can stay tonight; patient can name 988 and the clinic after-hours route. Written plan the patient could repeat, including means already secured. Phone contact tomorrow; in-person visit in three days. If sleep, alcohol, or ideation worsen, or a method appears, use the crisis line and emergency department; a missed contact is an escalation.

The second note is not a universal template. It shows current state, trajectory, acute concern, what is actually mitigating risk, why this setting, and what happens next.

Disposition is not only where the patient goes. It is what happens next.

An outpatient plan should show the next contact and who owns it; how soon that contact is, given the trajectory; how lethal means are being addressed; which crisis route the patient can actually use; who, if anyone, is involved with consent; and what finding will change the level of care. “Agrees to go to the ER if worse” can be one instruction. It is not a follow-up plan, a means plan, or a reason this setting is appropriate.

This is a principle, not an algorithm. The outpatient plan loses credibility when the patient cannot or will not engage in safety planning; when the history cannot be reconciled; when intoxication, withdrawal, agitation, psychosis, or mania makes today’s denial unusable; when a recent attempt or preparatory behavior is unexplained by the current statement; when supports the plan depends on are gone; or when the next contact is too far from the tempo of the risk. Hospitalization is one response. So are same-day re-evaluation, a higher-intensity outpatient setting, longer observation, collateral information, or a documented decision that this setting can still hold the risk with a tighter plan. The record should show which option was chosen and why.

These are recording failures, not a scoring system. Correcting them does not predict or prevent suicide.

A promise not to die is not a supported risk-reduction method and does not explain the disposition.

A rating without current symptoms, trajectory, and mitigators is a label.

Yesterday’s denial pasted over today’s change hides the only information that would alter the plan.

A list that does not say what is functioning now cannot support sending the patient home.

“Denies plan today” after a near-term attempt, rehearsal, or farewell act is an incomplete sentence.

Thought content is not a level of care.

Active thoughts still require a formulation.

A positive screen starts an assessment; it does not finish one. The VA/DoD guideline found insufficient evidence that screening programs, by themselves, reduce suicide or attempts.

Escalation language without a next contact, a means plan, and a reason this setting is appropriate is not a disposition.

The paragraph—or the equivalent structured fields—should answer: current suicidal state; what changed, and over what interval; the acute concern beyond the label “passive”; which mitigators are meaningful now; why this level of care is appropriate today; and the follow-up plus what would change the decision.

If those answers are not in the record, the disposition is not documented. A checklist can prompt the questions. It cannot answer them, and it cannot predict the outcome.

The purpose of suicide-risk documentation is not to prove that the patient denied a plan. It is to show why the clinician believed the chosen level of care was appropriate—and what would change that decision.

General educational information for U.S. psychiatric prescribers and outpatient clinicians. Not legal, risk-management, or clinical advice for a specific patient; does not establish a nationwide standard of care; does not claim that a checklist, score, or note format predicts or prevents suicide. Verify current authoritative sources.