Mentor Research Institute

Training, Consumer Protection, Value-Based Contracting, Online Screening, Public Health

503 227-2027

Suicide Risk Beyond the Questionnaire

Clinical Assessment, Risk Formulation, Documentation, and Continued Care

Overview

Introduction

Suicide-risk assessment is one of the most consequential responsibilities in behavioral healthcare. Clinicians are expected to identify meaningful risk, respond proportionately, document their reasoning, and determine whether treatment should continue, intensify, or shift to a different level of care. These decisions are often made under conditions of uncertainty. Patients may deny suicidal thoughts, provide incomplete information, fluctuate rapidly, minimize their distress, or present with warning signs that are not fully captured by a standardized questionnaire.

Validated screening instruments are valuable because they provide structure, improve consistency, and help clinicians identify patients who require further evaluation. They can support inquiry into suicidal ideation, intent, planning, prior behavior, and other clinically relevant factors. They cannot, however, determine with certainty which individual will attempt or die by suicide. They also cannot substitute for clinical observation, collateral information, changes from baseline, assessment of behavior, evaluation of access to lethal means, or consideration of the patient’s current psychological and social circumstances.

The central problem is not whether standardized instruments should be used. They should be used when appropriate. The problem arises when a screening result is treated as a complete assessment, a prediction of future behavior, or a definitive determination of medical necessity. A patient may score below a designated threshold while simultaneously demonstrating severe insomnia, agitation, escalating substance use, hopelessness, social withdrawal, recent loss, preparatory behavior, or reduced access to protective support. These findings may not accurately predict an individual suicide, but they may still be clinically significant and require intervention.

This article presents suicide-risk assessment as a process of structured clinical formulation. It distinguishes screening from comprehensive assessment, prediction, and treatment planning. It also examines how face validity, content validity, concurrent validity, and predictive validity apply differently to suicide-related information. The purpose is to help clinicians recognize meaningful evidence, evaluate conflicting information, document foreseeable risk, select proportionate interventions, and support continued care when clinically important risk is not adequately represented by a questionnaire score.

Background

Historically, suicide-risk assessment has often been organized around a relatively narrow set of questions: whether the patient is thinking about suicide, whether the patient has a plan, whether the patient intends to act, and whether the patient has access to a method. These questions remain essential, but they do not encompass the full range of information relevant to clinical decision-making.

Suicide is a low-base-rate outcome that may emerge from rapidly changing combinations of vulnerability, acute distress, access to lethal means, impaired judgment, disinhibition, interpersonal disruption, and loss of protective structure. Many people who experience suicidal thoughts do not attempt suicide, while some individuals who attempt suicide deny or do not disclose suicidal ideation during prior clinical contact. Risk may change over hours or days and may be influenced by intoxication, sleep deprivation, humiliation, relationship loss, legal or financial stress, pain, medical illness, psychosis, mixed mood states, or abrupt changes in treatment and support.

For these reasons, the predictive validity of individual questions and risk classifications is limited. A finding may be clinically important without being sufficiently precise to predict who will die by suicide. Hopelessness, entrapment, preparatory behavior, farewell communication, severe agitation, escalating substance use, and sudden social withdrawal may not provide a reliable numerical prediction for an individual patient. They may nevertheless have strong face validity, contribute to the content validity of a comprehensive assessment, and demonstrate concurrent validity when they occur alongside other evidence of deterioration.

Face validity concerns whether a question or observation appears relevant to suicide risk. Content validity concerns whether the assessment adequately covers the range of domains that may be clinically important. Concurrent validity concerns whether findings are consistent with other information available at the same time. Predictive validity concerns whether a finding accurately forecasts a later outcome. These forms of validity are related but not interchangeable. An overemphasis on predictive validity can result in the mistaken conclusion that information lacking strong individual predictive power has little clinical value.

In practice, clinicians are not asked only to predict death. They are asked to identify instability, recognize foreseeable deterioration, reduce access to lethal means, treat the conditions contributing to risk, involve protective supports, monitor changes, and determine whether the current level of care remains appropriate. These responsibilities require a broader clinical framework than a simple high-risk or low-risk classification.

Health-plan utilization criteria have sometimes reinforced narrow approaches by emphasizing current suicidal ideation, stated intent, imminent danger, or questionnaire thresholds when determining whether continued or intensive care is medically necessary. Such criteria may create pressure to reduce care once a patient denies intent or falls below a designated score. This approach can overlook the fact that treatment may be medically necessary because it is preventing further deterioration, maintaining protective structure, reducing dynamic risk, and treating the condition that produced suicidal vulnerability.

Regulatory actions, parity enforcement, and litigation involving behavioral-health utilization management have increased scrutiny of criteria that are more restrictive than generally accepted standards of care or that limit access to treatment through narrow interpretations of medical necessity. These developments do not establish that every adverse coverage decision is improper. They do demonstrate that health-plan criteria should not automatically displace individualized clinical judgment or prevent clinicians from documenting a broader and clinically supported rationale for continued care.

A comprehensive suicide-risk formulation integrates the patient’s current state, baseline vulnerabilities, dynamic changes, foreseeable stressors, access to lethal means, prior behavior, collateral information, protective factors, treatment engagement, and response to intervention. It also addresses contradictions. A patient’s denial of suicidal intent may be credible, partially credible, uncertain, or inconsistent with other findings. The clinician’s responsibility is not to accept or reject a statement mechanically, but to evaluate it within the full clinical context.

The purpose of this article is therefore not to diminish the value of standardized instruments. It is to place them in their proper role. Questionnaires and structured tools can improve consistency, prompt inquiry, support documentation, and reduce omission. They should supplement comprehensive clinical assessment rather than replace it. Ethical and clinically defensible care requires attention to all meaningful evidence, including signs and behaviors that may not independently predict suicide but that materially affect safety, treatment needs, and the foreseeable consequences of reducing or discontinuing care.

Part I: Screening Is Not a Declaration of Safety

A suicide screen answers a limited question: Does this person require additional evaluation? It does not answer:

  • Will this person attempt suicide?

  • Is this person safe to discharge?

  • Is continued treatment medically unnecessary?

  • Can the clinician disregard contradictory behavioral evidence?

  • Is the patient’s denial reliable under current conditions?

Brief tools such as the ASQ, PHQ-9 suicide item, and abbreviated versions of the Columbia Suicide Severity Rating Scale can efficiently identify reported thoughts or behaviors. Their efficiency is clinically valuable. Their brevity is also their limitation.

A negative response may reflect actual absence of ideation, but it may also reflect shame, fear of hospitalization, concern about employment or firearm restrictions, impaired insight, dissociation, intoxication, distrust, ambivalence, cultural differences, literal interpretation of questions, or a wish to protect family members from intervention.

Accordingly:

A negative screen is evidence to consider. It is not proof of safety.

Likewise, a positive screen generally initiates further assessment. It does not by itself determine hospitalization, predict death, or establish the precise intensity and duration of care.

Part II: Predictive Validity and Clinical Importance

Predictive validity concerns whether information accurately forecasts a future outcome. Suicide prediction is difficult because suicide is a low-base-rate event, risk changes over time, and many people who demonstrate recognized risk factors will not die by suicide.

This limitation has two implications:

  1. A scale should not be presented as capable of predicting individual suicide with certainty.

  2. Poor individual predictive accuracy does not make clinically meaningful warning signs irrelevant.

A smoke detector does not predict which room will burn or how much damage will occur. It identifies conditions requiring attention. Similarly, hopelessness, entrapment, agitation, preparatory behavior, escalating substance use, and abrupt withdrawal may require intervention even when no instrument can precisely calculate the probability of death.

Clinical importance is therefore broader than predictive validity. Clinicians must determine whether available information indicates:

  • A meaningful change from baseline;

  • Reduced self-control or judgment;

  • Increasing psychological pain;

  • Diminishing reasons for living;

  • Movement toward lethal capability or opportunity;

  • Breakdown of external protective structure;

  • A need for monitoring, treatment, safety planning, or a higher level of care.

Part III: Face, Content, and Concurrent Validity

Face Validity

Face validity asks whether an item appears, on its face, to concern the construct being evaluated.

Examples include:

  • “Have you been thinking about killing yourself?”

  • “Have you been searching for a way to die?”

  • “Do you believe your family would be better off without you?”

  • “Have you started giving away possessions because you do not expect to need them?”

Face validity is useful for communication and clinical recognition. It is not, by itself, empirical proof.

Content Validity

Content validity concerns whether an assessment adequately samples the clinically relevant domains of suicide risk.

A content-valid evaluation should ordinarily extend beyond current ideation to consider:

  • Suicidal thoughts, wishes, imagery, intent, planning, and rehearsal;

  • Prior attempts, interrupted attempts, aborted attempts, and preparatory acts;

  • Hopelessness, entrapment, unbearable psychological pain, and perceived burdensomeness;

  • Agitation, rage, panic, insomnia, impulsivity, and impaired cognitive control;

  • Substance intoxication, withdrawal, disinhibition, and rapidly changing mood states;

  • Psychosis, command hallucinations, severe depression, mixed affective states, and dissociation;

  • Access to lethal means and increased familiarity with a method;

  • Relationship loss, humiliation, legal jeopardy, financial collapse, medical deterioration, or occupational crisis;

  • Social withdrawal and loss of treatment, family, housing, employment, or spiritual supports;

  • Reasons for living, willingness to accept help, treatment engagement, and realistic safety supports.

An assessment limited to “Are you suicidal?” lacks adequate content coverage.

Concurrent Validity

Concurrent validity concerns whether a finding is meaningfully associated with other relevant information present at approximately the same time.

Examples include:

  • Denial of suicidal ideation occurring alongside internet searches for lethal methods;

  • Stated hopefulness despite giving away valued possessions and finalizing affairs;

  • A low questionnaire score accompanied by severe insomnia, agitation, intoxication, and sudden relationship loss;

  • Denial of intent contradicted by a spouse’s report that the patient has acquired a firearm and written farewell messages;

  • Reported improvement accompanied by emotional numbing, disengagement from treatment, and an abrupt calm after intense despair.

Concurrent information does not automatically prove suicidal intent. It may, however, make a simple denial less clinically persuasive and create an obligation to inquire further.

Part IV: Clinically Significant Content From Published Instruments and Checklists

The following domains repeatedly appear across published suicide instruments, structured interviews, clinical guidelines, and warning-sign checklists.

1. Desire for Death

Examples:

  • Wishing not to wake up;

  • Believing death would provide relief;

  • Feeling that life is no longer worth continuing;

  • Imagining disappearance, accidental death, or being killed without taking direct action.

A passive wish for death is not equivalent to an active plan, but it remains clinically relevant.

2. Suicidal Ideation Characteristics

Evaluate:

  • Frequency;

  • Duration;

  • Intensity;

  • Controllability;

  • Triggers;

  • Deterrents;

  • Whether thoughts are verbal, visual, intrusive, comforting, frightening, or rehearsed;

  • Whether ideation has become more specific or compelling.

“Occasional thoughts” may represent very different levels of concern depending on these characteristics.

3. Intent, Planning, and Preparation

Examples:

  • Selecting or researching a method;

  • Choosing a place or time;

  • Acquiring or positioning lethal means;

  • Rehearsing, testing, or conducting a “trial run”;

  • Writing notes, passwords, instructions, or farewells;

  • Arranging care for children, pets, property, or finances;

  • Avoiding discovery or rescue.

Preparatory behavior may be more clinically significant than a numerical symptom score.

4. Past Suicidal Behavior

Assess:

  • Actual attempts;

  • Interrupted attempts;

  • Aborted or self-interrupted attempts;

  • Preparatory acts;

  • Medical lethality;

  • Expected lethality;

  • Rescue probability;

  • What stopped the act;

  • Similarity between the prior crisis and the current situation.

A history of behavior is particularly important when current circumstances recreate the psychological conditions surrounding a previous attempt.

5. Hopelessness and Entrapment

Examples:

  • “Nothing can change.”

  • “There is no way out.”

  • “Every option will hurt someone.”

  • “This will never end.”

  • “I cannot tolerate another day like this.”

Entrapment may be external, internal, interpersonal, financial, legal, occupational, medical, or psychological.

6. Acute Arousal and Loss of Cognitive Control

Examples:

  • Severe agitation;

  • Inability to sleep;

  • Panic or frantic distress;

  • Racing or repetitive thoughts;

  • Rage;

  • Marked irritability;

  • Dissociation;

  • Inability to generate alternatives;

  • Feeling compelled to escape immediately.

These conditions can shorten the distance between thought and action.

7. Burdensomeness and Disconnection

Examples:

  • Believing others would be relieved by one’s death;

  • Feeling replaceable, unwanted, or contaminated;

  • Withdrawal from family, treatment, work, or community;

  • Refusing contact because “there is no point”;

  • Loss of meaningful roles or reciprocal relationships.

8. Substance Use and Disinhibition

Evaluate:

  • Escalation in alcohol or drug use;

  • Intoxication during prior attempts;

  • Use of substances to become capable of acting;

  • Withdrawal symptoms;

  • Combining substances with medication or weapons;

  • Reduced impulse control;

  • Blackouts or periods of unremembered behavior.

9. Access to Lethal Means

Assessment should address actual access, not merely ownership. Relevant questions concern:

  • Location;

  • Storage;

  • Ammunition;

  • Medication quantities;

  • Recent acquisition;

  • Familiarity;

  • Ability to regain access;

  • Willingness and feasibility of temporary risk-reduction measures.

10. Protective Factors

Protective factors should be assessed realistically rather than listed mechanically.

A spouse, child, job, religion, or pet is not automatically protective. Determine:

  • Whether the patient experiences the factor as meaningful;

  • Whether it remains available;

  • Whether it increases burdensomeness or shame;

  • Whether the patient can access it during a crisis;

  • Whether it has prevented action before;

  • Whether it is stable enough to support the proposed level of care.

Part V: Formulation Rather Than a Risk Score

A clinically useful formulation integrates five elements:

1. Current State

What is happening now?

Include ideation, intent, planning, affective state, cognition, behavior, substance use, psychosis, sleep, impulse control, and access to means.

2. Baseline Vulnerability

What enduring factors increase susceptibility?

Examples include prior attempts, recurrent mood disorder, trauma, chronic pain, substance-use disorder, impulsivity, severe personality dysfunction, psychosis, and family history.

3. Dynamic Change

What has worsened or improved?

The direction and speed of change may be more important than an isolated score.

4. Foreseeable Stressors

What is likely to occur before the next contact?

Examples include court hearings, separation, discharge, eviction, loss of medication, anniversaries, intoxication, isolation, receipt of medical results, or access to a weapon.

5. Available Protection and Intervention

What realistically reduces danger?

Include engagement, family participation, means-safety measures, crisis access, medication management, increased contact, treatment response, willingness to disclose, and ability to follow a safety plan.

Part VI: Medical Necessity and Continued Care

Continued treatment may be clinically justified when care is needed to:

  • Stabilize a deteriorating condition;

  • Monitor dynamic suicide-risk factors;

  • Maintain gains that are not yet durable;

  • Prevent recurrence during a high-risk transition;

  • Treat the disorder or crisis producing suicidal vulnerability;

  • Strengthen coping, problem-solving, sleep, sobriety, and emotional regulation;

  • Develop and test a collaborative safety plan;

  • Reduce access to lethal means;

  • Involve family or other supports;

  • Coordinate psychiatric, medical, substance-use, and social services;

  • Provide sufficiently frequent contact to detect clinically important change.

Medical necessity should not be defined solely by current suicidal intent. Treatment may be necessary precisely because it is preventing the patient from progressing to intent, preparation, or an attempt.

A utilization reviewer may disagree with the treating clinician. The clinician should nevertheless document the complete rationale rather than altering the formulation to fit a narrow checklist.

Sample Continued-Care Rationale

“The patient denies current intent and did not exceed the questionnaire’s high-risk threshold. Continued treatment remains medically necessary because the patient demonstrates a marked change from baseline, severe insomnia, increasing alcohol use, hopelessness, perceived burdensomeness, recent marital separation, withdrawal from established supports, and access to a previously contemplated method. These converging findings indicate unstable and potentially escalating risk that is not adequately represented by the screening score. Continued care is required to treat the underlying depressive condition, monitor dynamic risk, reduce access to lethal means, restore protective structure, and evaluate response before treatment frequency can be safely reduced.”

Part VII: Ethical and Documentation Principles

A defensible record should state:

  • What the patient reported;

  • What the clinician observed;

  • What collateral sources reported;

  • Where information converged or conflicted;

  • Which factors are static, dynamic, acute, and protective;

  • Why denials or positive findings were considered credible or uncertain;

  • What foreseeable developments could increase risk;

  • What interventions were offered and accepted;

  • Why the selected frequency and level of care were reasonable;

  • What follow-up, consultation, and contingency plans were established.

Avoid unsupported statements such as:

  • “Patient contracted for safety.”

  • “Low risk because the patient denies ideation.”

  • “No risk factors.”

  • “Not suicidal.”

  • “Score proves outpatient treatment is sufficient.”

More defensible language is:

“No current intent or preparatory behavior was disclosed or identified during today’s evaluation. Risk cannot be eliminated. The present outpatient plan is reasonable because…”

Conclusion

Suicide-risk assessment is not a search for a single predictive question. It is a disciplined process of identifying clinically meaningful information, evaluating discrepancies, recognizing dynamic changes, formulating foreseeable risk, and selecting proportionate interventions.

Validated instruments improve consistency and may detect information that would otherwise be missed. They should augment—not replace—clinical inquiry, observation, collateral information, judgment, documentation, and continuing reassessment.

A clinician should not disregard a credible warning sign merely because it is absent from a health plan’s questionnaire or lacks sufficient predictive power to forecast individual death. When multiple indicators have strong face and content relevance, concur with observed deterioration, and materially affect safety or treatment needs, they provide a legitimate clinical and ethical basis for intervention and continued care.



References and Public Resources

Suicide-Risk Assessment and Clinical Management

Warning Signs, Risk Factors, and Protective Factors

National Suicide-Prevention Policy and Healthcare Systems

Health-Plan Criteria, Medical Necessity, and Mental Health Parity

Regulatory and Enforcement Examples

Key words: Supervisor Education, Ethical Charting, Barriers to Oregon’s Mental Health Services, Mental Health, Psychotherapy, Counseling, Ethical and Lawful Value-Based Care,