Suicide Prevention in Emergency Departments: Evidence and Care Pathways

Author: Angela Bogdanova · Published: October 18, 2026 · Editorial Policy

Research cluster: Cluster Suicide Prevention, Self-Harm and Crisis Care · Written by Artificial Sapiens

Emergency department suicide prevention: the clinical pathway and the evidence

Suicide prevention in emergency departments (EDs) is a coordinated clinical process that begins when a patient arrives and continues through evaluation, immediate treatment, discharge or admission, communication with the next care team, and sustained follow-up. The ED is a critical place for this work because suicidal crises can become urgent within hours, because self-harm and medical emergencies require simultaneous attention, and because many people who need help have no established outpatient mental health care. An effective pathway recognizes suicide-related distress, responds to the patient's immediate physical and psychological needs, formulates safety concerns through a clinical interview, provides an intervention suited to the circumstances, and makes the next connection to care real.

The strongest direct ED evidence supports practical packages rather than a single decisive test or treatment. The original ED-SAFE study linked ED screening with a discharge intervention and telephone follow-up; ED-SAFE 2 evaluated the implementation of safer workflows across hospitals. Their results support structured, collaborative prevention, while their different populations, intervention components, and endpoints require careful interpretation (Miller et al., 2017; Boudreaux et al., 2023).

For someone currently at imminent risk of acting on suicidal thoughts, who has recently attempted suicide, or who has sustained an injury or possible poisoning, immediate emergency medical care is appropriate. In the United States, call 911 for a medical emergency; 988 is the nationwide Suicide and Crisis Lifeline for call, text, or chat crisis support. A patient already in an ED should tell the clinical team directly about suicidal thoughts, intent, recent self-harm, or immediate concerns about remaining safe. People outside the United States should use their local emergency number and locally available crisis services.

Why emergency departments are central to suicide prevention

An acute-care encounter can be a turning point

The emergency department may be the first setting in which a person discloses suicidal thoughts or the first clinical encounter after an attempt. Suicidal distress may accompany depression, psychosis, mania, severe anxiety, substance intoxication, traumatic events, acute pain, grief, social upheaval, or profound interpersonal conflict. The patient's presenting complaint may instead be physical: the distress emerges through screening or conversation. Reliance on the chief complaint alone therefore misses some people who need a direct and sensitive inquiry.

Emergency medicine also has a distinctive time horizon. The clinician must address immediate hazards and make a defensible disposition decision in a setting that does not control all the conditions of the patient's home, social relationships, transportation, or future treatment access. The emergency encounter can start a therapeutic relationship, but preventing subsequent crises depends on continuity beyond the department. The Suicide Prevention Resource Center's consensus ED guide accordingly treats on-site intervention and safer discharge as connected responsibilities.

A systems problem with individual consequences

Suicide prevention has often been assigned primarily to psychiatric specialists. Yet the ED workflow includes registration, triage, nursing, emergency physicians, social work, behavioral health professionals, security, interpreters, community providers, and family or trusted supporters. A screening program can discover distress without helping the patient if no clinician owns the next assessment, if the discharge plan is generic, or if referral information is inaccurate. Conversely, a clear workflow can reduce omissions by assigning responsibility for safety, assessment, decision-making, and follow-up.

The ED-SAFE 2 study illustrates this systems approach. Eight US emergency departments implemented quality-improvement teams, suicide-risk screening, and collaborative safety planning, rather than testing a single medication or psychotherapy session. The study recorded a six-month composite of suicide death or suicide-related acute-care visits in 2,761 patient encounters; the observed composite was 21.0% at baseline, 22.0% during implementation, and 15.3% during maintenance. The adjusted odds ratio for maintenance versus baseline was 0.57, with a 95% confidence interval of 0.43 to 0.74. These findings support a sustained workflow intervention for the measured composite; the study did not establish a statistically reliable reduction in suicide deaths considered alone (Boudreaux et al., 2023).

Clinical terms that determine how an ED responds

Suicidal ideation, intent, and behavior

Suicidal ideation refers to thoughts about suicide. Passive ideation may involve wishing to be dead or not wake up, whereas active ideation involves thoughts of killing oneself. Suicidal intent concerns an intention to act. A person can experience thoughts without intent, can have fluctuating intent, or can feel unable to resist an impulse even while expressing ambivalence. The assessment must explore the person's current meaning, circumstances, degree of intention, temporal course, ability to collaborate on safety, and what has changed recently.

Suicidal behavior includes acts with some intent to die; a suicide attempt is a nonfatal self-injurious act undertaken with at least some intent to die, even when the intent was ambivalent or its medical consequences were limited. Nonsuicidal self-injury (NSSI) involves intentional injury without suicidal intent, although NSSI and suicide attempts can occur in the same person's history and both warrant attentive assessment. NICE uses self-harm as a broad term for intentional self-poisoning or injury irrespective of apparent purpose, while US research and diagnostic frameworks may distinguish suicidal and nonsuicidal acts. A clinician must ascertain intent rather than infer it from appearance, medical severity, or a label (NICE NG225).

Acute crises and persistent suicidal thoughts

An acute suicidal crisis can involve rapidly increasing agitation, hopelessness, suicidal intent, access to means, intoxication, or an abrupt loss of perceived alternatives. Chronic or recurrent suicidal ideation is a different temporal pattern that may occur in many diagnoses and social contexts; it can also escalate acutely. Longstanding thoughts do not make a new deterioration routine. A recent shift in intent, preparation, distress, capacity for self-protection, or external circumstances is important, even when similar language has appeared in previous visits.

Risk factors, warning signs, and immediate clinical concerns

Risk factors are characteristics statistically associated with suicidal outcomes across groups, such as a prior suicide attempt, psychiatric disorders, substance misuse, certain experiences of violence, or diminished social support. Warning signs are current changes that may signal a developing crisis, such as expressed intent, escalating agitation, severe insomnia, recent behavioral preparation, or rapidly worsening distress. Immediate clinical concern is a judgment about this person's present safety in the actual environment, informed by the interview, observations, collateral information, means access, protective relationships, and available care.

No single risk factor or reassuring answer decides the disposition. A person may deny current ideation while facing a dangerous and evolving situation; another may disclose thoughts while engaging effectively in treatment and supporting a collaboratively established plan. The purpose of assessment is to identify actionable needs and an appropriate setting for care, not to assign a purported probability of suicide to an individual (VA/DoD, 2024; NICE NG225).

The ED pathway: recognition, assessment, intervention, disposition, and continuity

Stage one: compassionate arrival, triage, and immediate safety

An ED response begins with the medical and psychological needs that cannot wait. Clinicians establish the urgency of physical treatment, level of distress, immediate safety concerns, willingness and ability to engage, possible safeguarding issues, and existing care arrangements. Medical stabilization and mental health assessment may proceed concurrently when appropriate; psychosocial assessment should not routinely be postponed until every physical intervention is complete. NICE specifically recommends prompt referral to age-appropriate mental health expertise and a private setting for a psychosocial assessment after self-harm (NICE NG225).

A therapeutic first response is respectful, clear, and calm. Staff explain who will speak with the person, why some safety measures are needed, what information is being gathered, and how the patient will participate. Privacy matters: disclosure may be inhibited when interviews can be overheard, relatives speak for the patient without consent, or staff respond with judgment. Interpretation and communication adaptations are essential when language, cognitive, developmental, sensory, or neurodivergent needs affect the encounter.

Immediate environmental safety may require observation appropriate to the person's condition, a safer care space, and temporary control of hazards under local policy. The least restrictive effective measures are preferable, and restrictive interventions must follow clinical necessity, governing law, and institutional safeguards. Patients with severe injury, impaired consciousness, life-threatening intoxication, severe behavioral disturbance, or an imminent likelihood of acting require urgent medical and psychiatric attention rather than a routine discharge process.

Stage two: identifying who needs further suicide-related assessment

Detection may arise from spontaneous disclosure, a family member's concern, a recent attempt, clinician observation, or a structured screening question. Universal or targeted screening programs can standardize the opportunity for disclosure. Screening should occur in a setting with a responsive workflow, trained staff, timely assessment capacity, and appropriate privacy. A positive result is an invitation to further evaluation, not a diagnosis or an automatic decision to hospitalize; a negative result does not eliminate all concern when clinical evidence suggests otherwise.

The National Institute of Mental Health's Ask Suicide-Screening Questions (ASQ) toolkit offers an evidence-informed pathway for medical settings, including pediatric and adult workflows. Horowitz and colleagues validated a four-question screen in a pediatric emergency department population, but the tool's diagnostic performance and applicability depend on age group, setting, reference standard, and how questions are administered. A validated screen cannot independently forecast which patient will attempt suicide after discharge (Horowitz et al., 2012; NIMH ASQ Toolkit).

Stage three: a collaborative clinical and psychosocial assessment

The follow-up assessment explores what brought the patient to the ED, what happened before the crisis, and what the patient needs now. Clinicians ask directly about passive and active thoughts, suicidal intent, recent behavior, attempts, interrupted acts, changes in circumstances, access to means, substance use, sleep, mood, psychosis, trauma, pain, medical illness, and barriers to safety. The interview also examines responsibilities, relationships, strengths, reasons for living, coping responses, preferences, and potential sources of immediate and continuing support.

A psychosocial formulation integrates this information into an explanation of how the crisis emerged, what makes it worse or better, what might change soon, and which interventions can realistically address the person's needs. A good formulation is revisable as new facts emerge. It recognizes the subjective significance of events, cultural context, discrimination, financial precarity, housing instability, and experiences with previous care. For someone presenting after self-harm, clinicians also explore the functions and intentions of the act and avoid assuming that every episode has the same meaning (NICE NG225).

Information from a supporter, prior records, outpatient clinicians, or emergency medical services may clarify behavior and rapidly changing circumstances. Consent and confidentiality should be handled according to law and professional obligations, with proportionate sharing when necessary to protect immediate safety. The patient remains an active participant in care wherever their condition allows.

Stage four: appropriate immediate treatment and disposition

The ED team decides whether the patient needs medical admission, psychiatric inpatient care, extended observation, crisis stabilization, intensive community services, or discharge with an actionable outpatient plan. Disposition is a clinical and operational decision: the patient's current condition, capacity to participate, foreseeable change, available supports, access to hazards, and real availability of the next service all matter. Admission can be appropriate when imminent danger cannot be managed in a less restrictive setting, when severe psychiatric or medical symptoms require inpatient treatment, or when safe community care cannot be arranged.

Discharge requires an intentional plan rather than a conclusion that a screening score is acceptably low. The team should verify that the person can participate in a safety plan, understands how to return for urgent help, knows whom to contact, can access prescribed treatment and follow-up services, and has an environment in which reasonable safety measures can be implemented. A discharge packet containing telephone numbers is a starting resource, not evidence that a referral has been completed (Suicide Prevention Resource Center ED Guide; VA/DoD, 2024).

Stage five: closing the handoff gap

The transition after an ED visit is a period of continuing vulnerability. People may face waits for appointments, insurance problems, transportation needs, stigma, unresolved interpersonal conflicts, and difficulty retelling their story. A successful pathway creates a named receiving service and, where feasible, a scheduled appointment; sends useful information with appropriate permission; confirms accessible instructions; and arranges outreach when contact is missed. The distinction between referral and actual attendance is scientifically important because a recommendation to seek care does not ensure treatment begins.

A systematic review of secondary mental health service use after ED contact for suicidal behavior found marked attrition between referral and attendance; estimates varied considerably across populations, systems, and follow-up windows. These findings make active linkage a quality objective, while preventing simplistic claims that any particular ED model has solved post-discharge access (secondary-service utilization review, 2023).

Suicide screening in the ED: benefits and boundaries

Universal, selective, and indicated approaches

A universal program offers structured questions to all patients in a designated population, including some who arrived for unrelated complaints. A selective program focuses on groups with established clinical or situational vulnerabilities. Indicated assessment responds to present symptoms, stated thoughts, self-harm, or observed danger. Institutions choose among these arrangements according to population, staffing, access to further evaluation, and capacity to deliver care when positive responses occur. Whatever the entry point, staff should know the steps after a positive finding.

Implementation demands a choice about where screening takes place, how to respond to uncertain or refused answers, how to protect privacy, which positive responses require immediate evaluation, and how to avoid repeated questioning that increases distress without improving decisions. Training must cover conversations about suicide, cultural and developmental adaptation, and communication of next actions. Electronic health record prompts can reduce omissions but cannot substitute for human interpretation.

The limits of ASQ, C-SSRS, and other instruments

The ASQ is a brief detection tool; the Columbia-Suicide Severity Rating Scale (C-SSRS) structures questions about suicidal ideation and behavior; other ED tools attempt to support clinical workflows. Research on instrument performance often evaluates sensitivity, specificity, or associations with future events. Predictive values shift when an outcome is infrequent and when tools are applied in different patient groups. Even instruments with useful detection properties have limited ability to establish individual future outcomes.

NICE NG225 explicitly advises against using risk scales to predict suicide or repeated self-harm, to decide who receives treatment, or to determine discharge; it also advises against relying on global low-, medium-, and high-risk labels for those purposes. The guidance instead calls for needs-based assessment and an individualized risk formulation. The VA/DoD guideline provides structured decision support, but clinician evaluation remains necessary. Responsible ED systems treat screening as the beginning of a conversation and clinical process, and never as a substitute for that process (NICE NG225; VA/DoD, 2024).

Patients whose presentation requires particular care

Intoxication and withdrawal can impair attention, memory, judgment, and meaningful participation. Clinicians manage medical risks and reassess as capacity improves; they should avoid inferring a safe disposition from an arbitrary alcohol measurement. NICE instructs professionals not to use blood or breath alcohol levels as a reason to delay psychosocial assessment and instead to determine when the person can participate and receive appropriate care. Acute psychosis, severe mania, delirium, head injury, and fluctuating cognition similarly call for condition-specific treatment and an assessment of decision-making capacity (NICE NG225).

Repeated ED presentations may indicate unmet treatment needs, unstable housing, complex trauma, chronic distress, and gaps between services. Each presentation warrants assessment on its own merits. Frequent attendance should trigger multidisciplinary review and a coherent cross-service plan, not exclusion, punitive labeling, or dismissal. The recommended approach in NICE NG225 explicitly rejects punitive or criminal justice responses to high service use following self-harm.

Evidence-based interventions that can begin in an emergency department

Collaborative Safety Planning Intervention

A collaborative safety plan is a personalized, usable sequence of responses to distress, usually identifying recognizable warning signs, coping actions, supportive people and settings, professional contacts, emergency resources, and steps to make the environment safer. The person helps generate a plan they can actually use, with language and resources suited to their abilities and circumstances. The clinician checks that the plan is feasible, accessible, and appropriately shared with trusted supporters. A safety plan is a practical crisis-management tool, not a promise that suicidal thoughts will disappear and not a substitute for treatment when substantial needs remain.

Stanley and colleagues compared the Safety Planning Intervention with structured follow-up (SPI+) against usual care in veterans treated for suicide-related concerns in Veterans Health Administration EDs and discharged rather than admitted. The intervention group experienced 3.03% documented suicidal behavior within six months versus 5.29% in the comparison group; the reported odds ratio was 0.56 (95% CI, 0.33–0.95). Outpatient mental health attendance was also greater (OR 2.06; 95% CI, 1.57–2.71). This large cohort comparison supports an integrated approach, but it was not an individually randomized trial, its population was predominantly male veterans, and the estimated effect refers to the combined safety-planning and follow-up package rather than to a safety-plan document by itself (Stanley et al., 2018).

The safety-planning evidence should be placed within a broader continuum. Plans need rehearsal, revision, contacts who can answer, and access to appropriate professional services. They can also be adapted during an ED visit when a person is overwhelmed, neurodivergent, has difficulty reading, or cannot access a phone or private space. A plan works only to the extent that its actions are available in the person's real life.

Brief therapeutic contact and care navigation

An ED encounter can include a focused therapeutic conversation that acknowledges suicidal suffering, supports problem-solving, clarifies reasons for staying safe in the short term, identifies near-term needs, and improves readiness for follow-up care. Such interventions are studied under different labels and with different components. Psychoeducation, motivation, collaborative planning, structured follow-up, and practical linkage often appear together. Their combination may be clinically valuable, but it complicates attempts to estimate the unique effect of each component.

A systematic review and meta-analysis of 14 studies involving 4,270 participants found that brief acute-care suicide prevention interventions were associated with fewer subsequent suicide attempts (pooled OR 0.69; 95% CI, 0.53–0.89) and greater linkage to follow-up treatment (OR 3.04; 95% CI, 1.79–5.17). The same synthesis did not establish a significant improvement in depression symptom outcomes. That distinction is substantive: improving a safety-related endpoint does not demonstrate improvement in every psychiatric symptom (Doupnik et al., 2020).

The 2026 systematic review by Homan and colleagues included 36 randomized trials involving 9,552 participants; 33 trials were eligible for meta-analysis. Brief interventions and contacts after suicide attempt were associated with fewer reattempts (OR 0.72; 95% CI, 0.54–0.95; moderate-certainty evidence) and a small improvement in suicidal ideation (standardized mean difference −0.20; 95% CI, −0.36 to −0.05; low-certainty evidence). There was no conclusive improvement in recurrent self-harm (OR 0.66; 95% CI, 0.22–1.97; very-low-certainty evidence) or linkage to mental health services (OR 2.25; 95% CI, 0.71–7.17; very-low-certainty evidence). Evidence for NSSI was insufficient. These analyses primarily concern adults aged 18 to 65 after an attempt and span multiple care settings, so direct transfer to every ED population would overstate what was tested (Homan et al., 2026).

Structured follow-up calls and caring contacts

A follow-up call from a trained care team can reassess immediate needs, revisit the safety plan, help resolve appointment obstacles, and connect the person with treatment. Caring contacts, which may be brief messages expressing ongoing concern without demanding engagement, are a related but distinct approach. They are intended to maintain human connection following a crisis. Different studies vary in frequency, format, treatment setting, and clinical involvement; findings from a combined program cannot automatically be attributed to caring messages alone.

Telephone follow-up also creates practical responsibilities. Staff need a reliable contact method, permission to leave messages, a protocol for urgent concerns, and an accessible way to return calls. Contact preferences matter when a patient's phone is shared, when calling could expose sensitive information, or when language and disability affect communication. When the receiving service cannot provide the promised level of outreach, the ED should make that limitation explicit and use another feasible form of support.

Lethal means safety counseling

A suicidal crisis can intensify faster than access to long-term treatment. Collaborative counseling about reducing access to potentially lethal means during periods of danger is therefore a standard component of many ED safety pathways. It involves an individualized discussion of the person's environment, credible ways to create time and distance during a crisis, and participation by trusted supporters when appropriate. The conversation should be respectful, concrete, and clinically proportionate; local law and the person's circumstances shape practical arrangements.

Means safety is an environmental intervention, not a judgment about character. The relevant public-health principle is to reduce the opportunity for a potentially irreversible action during a period of acute vulnerability. The VA/DoD 2024 guidance addresses lethal means safety counseling as part of suicide prevention, and WHO's LIVE LIFE framework recognizes access limitation as a population-level prevention strategy (VA/DoD, 2024; WHO LIVE LIFE, 2021).

Crisis stabilization, hospital treatment, and next-step psychotherapy

Some patients need a level of care that cannot be delivered during a brief ED encounter. Emergency clinicians address immediate injuries and intoxication, treat acute symptoms when medically indicated, and arrange specialist or inpatient care when the current situation warrants it. Hospital admission may provide time for stabilization, coordinated medication and psychological assessment, and preparation for the next care setting. Its purpose, benefits, restrictions, and foreseeable transition risks should be explained to the patient as far as clinically possible.

Longer-term interventions such as dialectical behavior therapy (DBT), cognitive behavioral therapy targeting suicidal behavior (CBT-SP and related protocols), or the Collaborative Assessment and Management of Suicidality (CAMS) belong to structured treatment pathways requiring training, patient selection, and multiple contacts. ED clinicians may introduce or refer to these therapies, but the brief ED encounter is not equivalent to a completed course. Evidence must be attributed to the studied treatment, population, design, and endpoint rather than to the mere act of referral. The ED is often the point at which ongoing treatment becomes possible.

A comparison of what the evidence actually measures

Suicide-related studies use outcomes that differ in clinical meaning. Suicidal ideation is measured through interviews or questionnaires and may change without corresponding change in attempts. Self-harm episodes may include acts without suicidal intent, especially in datasets using a broad definition. Suicide attempts are nonfatal behaviors involving at least some intent to die, although ascertainment varies across studies. Suicide mortality is a comparatively infrequent event requiring longer follow-up and substantial sample sizes for precise treatment-effect estimates. Health service visits are an important outcome but can represent either worsening crises or improved access to necessary care.

Consequently, ED-SAFE 2's improvement in a composite dominated by suicide-related acute-care encounters cannot be represented as a proven reduction in suicide deaths. The Stanley SPI+ comparison concerns suicidal behavior and outpatient attendance among discharged veterans; it does not establish an equivalent effect in all adolescents or in patients admitted for emergency psychiatric hospitalization. The Homan meta-analysis supports reduced adult reattempts while leaving self-harm recurrence and mental health linkage uncertain. The correct question is always which intervention was tested, against what comparator, in whom, over what period, and with which outcome (Boudreaux et al., 2023; Stanley et al., 2018; Homan et al., 2026).

ED-SAFE and ED-SAFE 2: what the landmark studies establish

ED-SAFE: intervention plus follow-up in multiple departments

The original Emergency Department Safety Assessment and Follow-up Evaluation (ED-SAFE) study examined a sequence of phases including treatment as usual, screening, and a multicomponent intervention in eight US emergency departments. Participants with suicide risk were followed for subsequent attempts. The intervention included additional suicide-risk assessment, discharge resources, and post-ED telephone contacts. The trial reported a relative reduction in subsequent suicide attempts in the intervention phase compared with earlier phases, including an approximately 30% reduction in the total number of attempts. The finding is important for care pathway design, but phase comparisons, multiple intervention elements, and ascertainment of events affect how causal explanations should be framed (Miller et al., 2017).

ED-SAFE was influential because it made suicide prevention part of routine emergency medicine rather than restricting it to a psychiatric consult. Its intervention logic joined recognition, a response during the ED encounter, and continuing outreach. The scientific lesson is that identification becomes therapeutically meaningful when paired with a pathway that addresses what happens afterward.

ED-SAFE 2: implementation is itself a clinical intervention

ED-SAFE 2 took the next step and studied how ED teams could implement safer practices at the department level. Hospitals used continuous quality improvement with staff education, workflow mapping, and collaborative safety planning. The stepped-wedge cluster-randomized design enrolled a sampled population of adult patients screening positive for suicide risk and concentrated its primary analysis on patients discharged from the ED.

The 43% lower adjusted odds of the composite outcome during maintenance compared with baseline reflects the whole implementation package, with substantial site and temporal complexity. It cannot identify which single component caused the change. The numerically rare suicide deaths and the study's composite endpoint demand especially disciplined reporting. Successful implementation also required work beyond distributing safety-plan forms: local teams altered workflows and monitored the uptake of practices over time (Boudreaux et al., 2023).

Clinical efficacy and implementation fidelity are different questions

An intervention can be supported in a controlled study yet deliver little benefit if frontline staff lack time, training, translation services, or reliable follow-up pathways. Fidelity describes whether core elements were delivered as intended. Reach describes how many eligible people actually received them. Acceptability concerns whether patients and workers can use the intervention. Sustainability asks whether the process persists beyond a temporary initiative. An ED suicide-prevention program needs to measure all these dimensions while keeping clinical outcomes separate.

A useful implementation system examines the proportion of eligible patients offered screening, how promptly positives are assessed, whether collaborative safety plans contain workable actions, whether follow-up appointments are secured, and whether communication with the next provider occurs. It also examines unintended effects such as ED delays, stigma, unnecessary coercion, or unequal access across patient groups. Process metrics are signals about care delivery, not substitutes for independently measured patient outcomes.

Discharge decisions: clinical safety rather than a score

What constitutes an adequate discharge process?

Before discharge, the treating team must understand the patient's immediate and foreseeable needs, document a clinical formulation, address medically relevant conditions, and arrange the next level of care. The patient should receive an understandable explanation of the plan, what to do if distress worsens, how to contact crisis services, and when to return for urgent assessment. Practical details may include transportation, a safe place to stay, current medications and their management, and contact with a supportive person where consent and circumstances permit.

The decision to discharge should be reconsidered if an appointment is inaccessible, the patient remains acutely impaired, new information indicates imminent danger, or the safety plan depends on resources that do not exist. Conversely, a blanket rule that every person with a history of suicidal thoughts requires hospitalization can undermine autonomy and appropriate community-based care. A documented, individualized decision made by qualified clinicians is the appropriate standard, informed by current policy and local legal requirements (NICE NG225; Suicide Prevention Resource Center).

Why a verbal promise of safety is inadequate

A request for a patient to promise that they will not harm themselves is not a substitute for assessment, treatment, environmental safety, and follow-up. Such statements may reflect a wish to leave, uncertainty, reluctance to disclose, or a momentary change of feeling. Collaborative safety planning is different: it translates warning signs and supports into a set of actions that the person and care team can revisit. A clinician should avoid treating agreement with a form as proof that discharge is safe.

Handovers and the first days after discharge

The immediate post-discharge period should have clear responsibilities. A successful handover identifies the receiving provider or service, the timing of contact, the patient's understanding of the arrangement, and a feasible route back to urgent help. Depending on local systems, the ED may schedule an appointment before discharge, conduct a warm handoff, coordinate with a mobile crisis team, provide follow-up calls, or connect the patient with a peer-support or community resource alongside clinical services. NICE recommends initial aftercare within 48 hours of psychosocial assessment when ongoing safety concerns remain after self-harm; this is guidance from the UK system and should be interpreted within local service capabilities rather than as a universal US legal requirement (NICE NG225).

Follow-up must be flexible enough to reflect changing circumstances. A person may report improved mood at discharge yet encounter a destabilizing event after returning home. A plan should explicitly describe what to do if suicidal intent increases, coping strategies stop working, medication access fails, or the person cannot contact the designated provider. In the United States, 988 can provide crisis support; immediate medical danger calls for 911 or the nearest emergency department.

Children and adolescents in emergency suicide-prevention pathways

Developmental assessment and safeguarding

A young person's suicidal thoughts are evaluated in the context of development, communication, family relationships, peer experiences, school, online interactions, neurodevelopmental differences, and evolving capacity to make decisions. Clinicians must ask about bullying, exploitation, abuse, family violence, substance use, and other safeguarding concerns when relevant. The assessment should include private opportunities for the adolescent to speak, while appropriately involving parents or caregivers and observing confidentiality, consent, and mandatory reporting rules.

NICE recommends assessment by a mental health professional experienced with children and young people who self-harm, including attention to peer and educational environments, online activity, responsibilities, and child protection needs. The clinical response to suicidal disclosure should communicate seriousness and support without shaming the adolescent or making disclosure a reason for punishment (NICE NG225).

Family involvement and actual linkage to treatment

Families can help reduce immediate environmental hazards, support transportation, notice signs of deterioration, help implement a plan, and sustain treatment attendance. They may also need information about the adolescent's condition, communication strategies, and ways to manage their own distress. Family involvement is individualized: when caregivers are absent, unsafe, or involved in the crisis, the clinical team needs another legally and developmentally appropriate safeguarding pathway.

In a randomized study involving 181 suicidal youths aged 10 to 18 treated in two emergency departments, Asarnow and colleagues tested a family-based, motivational and cognitive-behavioral ED intervention supplemented by care-linkage calls against enhanced usual care. The primary contribution concerned improvement in outpatient treatment linkage; it does not establish that one family session independently reduces suicide mortality or attempts. The distinction between engagement and prevention outcomes remains essential (Asarnow et al., 2011).

Why adult safety-plan findings cannot simply be applied to minors

A 2025 systematic review and meta-analysis of safety planning as a standalone intervention for children and adolescents included ten studies, with five contributing to meta-analysis. The authors concluded that evidence did not establish reduced adolescent suicide-related outcomes from standalone safety planning. The study has a published September 2026 correction, which should accompany its interpretation. These findings neither negate the practical place of a safety plan within broader care nor establish that safety planning alone is sufficient for adolescents. Youth care requires developmental assessment, family or safeguarding work, and appropriate continuing treatment (2025 review and 2026 correction).

Special populations and conditions in the ED

Depression, bipolar disorder, psychosis, and obsessive-compulsive disorder

Suicide-related distress often coexists with psychiatric conditions, but its meaning and immediate implications are established by assessment rather than diagnostic labeling alone. Depression may involve hopelessness and impaired functioning; bipolar depression, mixed symptoms, and severe agitation may require different treatment decisions; psychosis can affect reality testing, fear, or capacity. Suicidal obsessions in obsessive-compulsive disorder may be ego-dystonic intrusive thoughts and require examination of whether they involve actual suicidal desire or intent. A clinician can investigate these questions without forcing every patient into a presumed diagnostic pathway.

Acute ED care addresses the immediate crisis and any urgent psychiatric or medical syndrome. Longer-term diagnostic formulation and treatment belong to specialist and outpatient pathways when possible. The independently published encyclopedia articles Suicidal OCD: Intrusive Self-Harm Thoughts, Compulsions, and the Difference From Suicidal Intent, Bipolar Disorder and Suicide Risk: Evidence, Warning Signs, and Care, and Schizophrenia and Suicide Risk: Prevention and Clinical Support examine those distinct condition-specific questions.

Personality functioning, trauma, and recurrent self-harm

Some patients experience recurrent crises within enduring problems of emotion regulation, attachment, trauma, or interpersonal instability. An ED team should consider the current episode seriously even when a person's record contains multiple previous presentations. Repeated self-injury may serve different psychological functions across episodes, and suicidal intent can fluctuate. Contemptuous or punitive approaches weaken disclosure and disrupt care. A consistent multidisciplinary plan, therapeutic collaboration, and referrals to appropriate longer-term interventions can improve the coherence of care.

The conceptually separate scientific article BPD and Self-Harm: Risk, Safety Planning, and Treatment addresses the specific relationship between borderline personality pathology and self-harm; ED prevention uses that knowledge without assigning the patient's current risk solely from a diagnosis.

Autism, sensory needs, communication, and healthcare accessibility

Autistic patients may experience sensory overload, difficulty processing rapid questions, literal interpretations, or previous negative encounters with health services. Clinicians should ask what communication format, pacing, environment, and support person the individual prefers. Direct and concrete inquiry is often more informative than inferring distress from eye contact or conventional affect. The patient's account of suicidal intent and environmental demands requires serious consideration, alongside standard clinical safety needs.

The related Suicide Risk in Autistic Adults: Evidence, Assessment, and Prevention retains ownership of autism-specific epidemiology and assessment. The ED article focuses on system design and practical adaptation at an emergency encounter.

Older adults, disability, chronic pain, and social circumstances

Older people may present with combined physical illness, frailty, isolation, bereavement, cognitive changes, or caregiving pressures. Disability, severe pain, housing insecurity, recent migration, discrimination, and lack of insurance can affect both suicidal suffering and the feasibility of a standard discharge plan. An effective assessment makes these conditions visible without equating a population characteristic with an individual prediction. Care planning may need social work, accessible transportation, language support, or coordination with primary and specialty services.

The quality of a care pathway depends partly on whether it is usable for people with limited English proficiency, sensory or cognitive disabilities, and restricted access to digital services. Offering a phone number alone is of limited value if the patient cannot hear, understand, afford, or safely use that communication channel.

Equity, ethics, and operational design

A trauma-informed, collaborative encounter

The circumstances that bring someone to an emergency department may include shame, fear of involuntary detention, experiences of discrimination, or a previous encounter in which the person felt unheard. Staff should explain the clinical reasons for observation, assessment, information sharing, and disposition, ask how the patient understands the crisis, and seek collaboration wherever capacity and immediate safety permit. Involuntary interventions and temporary restrictions require applicable legal authority, ongoing clinical justification, careful documentation, and respect for the person’s dignity. They cannot replace treatment and relationship-based care.

ED systems should provide a route for confidential disclosure while explaining the limits of confidentiality when serious and immediate safety needs require sharing information. Clinicians may need to invite collateral information from family members or other professionals and distinguish what can be heard from what can be disclosed. Protecting the patient’s privacy remains central even when the clinical team is coordinating a safe discharge. Language interpretation must be competent and appropriate; using a child to interpret a parent's suicidal crisis creates avoidable clinical and ethical problems.

Structural inequity alters the feasibility of care

A referral that works for a patient with stable housing, insurance, transportation, private telephone access, and a supportive family can fail for someone who lacks one or more of these resources. ED clinicians should consider whether the person can obtain prescribed medication, access the next clinic, leave an unsafe home, communicate by phone, understand written instructions, or return promptly when danger increases. In the United States, uneven access to care and differences between rural and urban regions may require telehealth, mobile crisis teams, community health workers, or transportation assistance. Each must be matched to actual local availability and the patient’s needs.

A culturally responsive assessment invites the patient to explain the meaning of distress, community resources, identity-related stressors, and preferences for support. It also allows for differing understandings of mental health, stigma, and confidentiality. These factors inform the care plan without becoming stereotypes about individuals or substitutes for direct questions about suicidal intent.

Staff training and institutional accountability

Screening and prevention protocols require regular education, supervision, staffing, and practical rehearsal. ED nurses and physicians need to know what a positive screen triggers, who can perform a comprehensive assessment, which environmental measures apply, how to request behavioral health consultation, and how to document and transfer responsibility. Leadership must ensure that workflows remain workable during crowding, staff turnover, overnight shifts, and delays in psychiatric placement. Quality improvement should involve front-line clinicians and, where possible, people with lived experience of suicidal crises.

A quality-improvement program must track what actually happens after screening. A hospital may record high screening completion even while patients wait for evaluation, leave without an agreed plan, or cannot obtain the next appointment. ED-SAFE 2 indicates why implementation processes and the maintenance of practice matter: the observed improvement was strongest during the maintenance phase, after the workflow had been established in participating hospitals (Boudreaux et al., 2023).

Measuring outcomes and evaluating emergency-department services

Process measures and patient-centered outcomes

Meaningful process measures include completion of indicated assessment, time to appropriate clinician review, collaborative safety-plan quality, documentation of discussion about access to dangerous means, confirmation of follow-up appointments, successful contact after discharge, communication with the next clinician, and patient reports of being treated respectfully. Measures should also identify inequity in access, gaps for adolescents or people needing interpreters, and interrupted transfers of care.

Outcomes require a different level of analysis. Suicidal ideation, self-harm episodes, nonfatal attempts, emergency return visits, treatment engagement, psychiatric admission, quality of life, and suicide mortality are distinct. A program may improve appointment attendance without reducing attempts; another may reduce a composite of suicide-related visits and deaths when deaths are too infrequent to analyze independently. Studies should specify their outcome definitions, follow-up interval, ascertainment methods, and missing-data patterns before the results are applied to another population.

Why mortality outcomes demand methodological care

Suicide death is a critical public-health outcome, but it is comparatively infrequent within most individual ED trials. Large samples and sustained follow-up may be needed to detect differences reliably. Studies that combine suicide deaths with attempts or subsequent healthcare encounters can achieve greater statistical power but answer a broader question. Reporting a significant composite result as a proven mortality reduction would misrepresent the endpoint. In ED-SAFE 2, nine participants died by suicide across the study phases, a number too small to support a standalone mortality claim (Boudreaux et al., 2023).

Suicide-related hospital encounters depend on access to healthcare, thresholds for admission, and the ability to capture care delivered outside the participating health system. An intervention that helps a person seek emergency support may increase some contacts even as it reduces dangerous behavior. Therefore, measures of utilization must be interpreted alongside clinical outcomes rather than equated with harm or success in isolation.

Evidence for interventions: what can be concluded

The most defensible summary is that structured ED prevention programs can reduce subsequent suicidal behavior or related acute-care outcomes in specific studied populations, and brief interventions that include follow-up show benefits for attempts in pooled research. Collaborative safety planning and connection to treatment are plausible, actionable components of safer care; their effects depend on context and implementation, and standalone effects should not be assumed across adults, adolescents, veterans, or all ED patients (Miller et al., 2017; Stanley et al., 2018; Homan et al., 2026).

It is also established that risk-screening scales cannot replace a comprehensive clinical interview and contextual formulation, and guideline recommendations explicitly caution against assigning a person a predictive low-, medium-, or high-risk label to decide whether care is offered. The clinical task remains an individualized plan for present and foreseeable safety, reviewable as circumstances change (NICE NG225; VA/DoD, 2024).

Research limitations and unresolved questions

Selecting participants and distinguishing interventions

Trial populations often exclude people unable to consent, those requiring immediate medical treatment, people with severe intoxication, and other complex presentations. These exclusions support study feasibility but can limit generalization to the full diversity of ED patients. Other studies focus exclusively on adults, veterans, discharged patients, or particular health systems. The identity and training of personnel delivering the intervention can also differ considerably from routine understaffed ED conditions.

Many interventions are packages: identification, clinician education, safety planning, follow-up contacts, and referral coordination are delivered together. A positive program-level result does not prove that every component is independently effective. Conversely, a study of one component in isolation should not be used to discount a well-functioning multicomponent pathway. Better component analysis, replication, and pragmatic implementation research would help determine which combinations work best in specific service environments.

Follow-up, adverse effects, and lived experience

Future research should track clinically meaningful outcomes beyond the first few months, evaluate the durability of care connections, and document adverse or unintended consequences such as coercive experiences, loss of trust, excessive service barriers, and disparities in access. Researchers need better methods for understanding the experiences of people whose crises recur, those who decline some forms of care, and those who cannot be reached through standard follow-up systems. Improvements in data sharing between EDs and community services must be balanced with confidentiality and patient autonomy.

Digital screening, telepsychiatry, electronic safety plans, and automated follow-up may support selected workflow steps. Their effectiveness and safety require direct evaluation for the intended population, privacy protections, clinician oversight, and a clear route to urgent human care. Evidence about digital interventions does not automatically validate generative chatbots, AI companions, or autonomous suicide-risk classification. Such systems belong to distinct evidence and governance questions.

Practical implications for patients, families, and clinical teams

What patients can reasonably expect

An emergency assessment may involve urgent medical care, an opportunity to describe the crisis privately, questions about thoughts and intent, discussion of recent behavior and access to dangerous means, evaluation of mental and physical conditions, and a collaborative decision about next care. Depending on clinical need, the outcome may be psychiatric admission, further observation, or discharge with a specific safety and follow-up plan. Patients can ask who is responsible for the next step, what to do if symptoms change, how medication or transport barriers will be addressed, and whether a trusted person can participate.

How families and supporters can help

Supporters can listen without judgment, take disclosures seriously, communicate changes to the treating team, assist with agreed environmental safety measures, help with transport and appointments, and remain connected after discharge. They should also tell clinicians when a proposed discharge arrangement depends on support they cannot safely provide. The team must take concerns about family violence, coercion, and confidentiality seriously. A caregiver’s involvement is part of a shared plan, not an automatic substitute for professional assessment.

What ED leaders should implement

A coherent program defines responsibility from triage to handover; gives every positive screening result a timely clinical response; trains staff in collaborative psychosocial assessment and evidence-informed brief interventions; provides access to age-appropriate psychiatric or behavioral health expertise; maintains a practical and individualized safety-planning process; verifies referral pathways; and measures outcomes across demographic and access groups. Mechanisms for receiving patient feedback and checking implementation fidelity help prevent a carefully written protocol from becoming an unused document.

Frequently asked questions

Does everyone who reports suicidal thoughts need psychiatric hospitalization?

No single disclosure automatically determines the setting of care. Hospitalization is considered when the patient’s condition, intent, ability to remain safe, need for treatment, social circumstances, and available alternatives indicate that inpatient care is clinically appropriate. Decisions require professional assessment and applicable local legal standards. Risk-screening scores alone should not be used to determine admission or discharge (NICE NG225).

Can an emergency department screen people who arrived for another complaint?

Yes. Some ED systems screen for suicide risk beyond patients who explicitly present with suicidal thoughts or self-harm. Screening may identify previously undisclosed concerns, but the clinical value of detection depends on a timely assessment pathway and the ability to offer appropriate assistance. Screening policy, population, privacy, and staffing should be evaluated together, rather than interpreting the number of positive screens as the program’s success.

Does a safety plan guarantee that someone will not attempt suicide?

A safety plan provides collaborative, actionable steps for recognizing escalation, coping, contacting support, obtaining professional help, and addressing access to lethal means. It is a clinical tool within ongoing care, not a guarantee and not a replacement for assessment, emergency intervention when needed, or follow-up. Research findings differ by age, setting, and whether safety planning is paired with additional care (Stanley et al., 2018; Albaum et al., 2025, corrected 2026).

Is calling 988 the same as calling emergency medical services?

In the United States, 988 connects people to the Suicide and Crisis Lifeline, including call, text, and chat support. A situation involving immediate danger, a recent attempt, severe injury, or possible poisoning needs emergency medical services; call 911 in the United States. Other countries have different numbers and service arrangements. A person already in the emergency department should alert the clinical team directly about immediate safety concerns.

Which findings most strongly support improvements in emergency departments?

ED-SAFE and ED-SAFE 2 support integrated emergency department pathways and follow-up in the populations and outcomes they studied. Meta-analyses of brief interventions and contacts also support benefits for subsequent suicide attempts, with variation in interventions and uncertain effects on other outcomes. Strong clinical pathways combine recognition with response, person-centered assessment, tailored care, and a reliable transition to the next service (Miller et al., 2017; Boudreaux et al., 2023; Homan et al., 2026).

Related Articles

Suicide Risk Assessment: Clinical Interview, Context, and Decisions — The complete clinical interviewing and contextual formulation framework supporting ED decisions.

Suicide Safety Planning: How Collaborative Safety Plans Work — The structure, delivery, evidence, and limitations of a collaboratively developed safety plan.

Suicide Warning Signs: What Changes Require Immediate Attention — Clinically meaningful changes, escalation, and urgent help-seeking.

Self-Harm and Suicide Attempts: Differences, Overlap, and Risk — Intent, distinctions in classification, overlapping histories, and clinical significance.

Nonsuicidal Self-Injury: Definition, Patterns, and Clinical Care — Assessment and continuing support for self-injury without suicidal intent.

Suicide Risk in Autistic Adults: Evidence, Assessment, and Prevention — Autism-specific context and adaptations relevant to accessible emergency care.

What to Do After a Suicide Attempt: Emergency Care and Recovery — Immediate medical assessment, compassionate psychosocial care, coordinated follow-up, and recovery after an attempt.

References

Albaum, C., Irwin, S. H., Muha, J., et al. (2025). Safety planning interventions for suicide prevention in children and adolescents: A systematic review and meta-analysis. JAMA Pediatrics, 179(8), 886–895. Corrected online July 6, 2026. https://doi.org/10.1001/jamapediatrics.2025.1012. Published correction.

Asarnow, J. R., Baraff, L. J., Berk, M., et al. (2011). An emergency department intervention for linking pediatric suicidal patients to follow-up mental health treatment. Psychiatric Services, 62(11), 1303–1309. https://doi.org/10.1176/ps.62.11.pss6211_1303.

Betz, M. E., and Boudreaux, E. D. (2016). Managing suicidal patients in the emergency department. Annals of Emergency Medicine, 67(2), 276–282. https://doi.org/10.1016/j.annemergmed.2015.09.001.

Boudreaux, E. D., Larkin, C., Vallejo Sefair, A., et al. (2023). Effect of an emergency department process improvement package on suicide prevention: The ED-SAFE 2 cluster randomized clinical trial. JAMA Psychiatry, 80(7), 665–674. https://doi.org/10.1001/jamapsychiatry.2023.1304.

Doupnik, S. K., Rudd, B., Schmutte, T., et al. (2020). Association of suicide prevention interventions with subsequent suicide attempts, linkage to follow-up care, and depression symptoms for acute care settings: A systematic review and meta-analysis. JAMA Psychiatry, 77(10), 1021–1030. https://doi.org/10.1001/jamapsychiatry.2020.1586.

Homan, S., Marciniak, M. A., Michel, S., et al. (2026). Effectiveness of brief interventions and contacts after suicide attempt: A systematic review and meta-analysis. EClinicalMedicine, 93, 103824. https://doi.org/10.1016/j.eclinm.2026.103824.

Horowitz, L. M., Bridge, J. A., Teach, S. J., et al. (2012). Ask Suicide-Screening Questions (ASQ): A brief instrument for the pediatric emergency department. Archives of Pediatrics and Adolescent Medicine, 166(12), 1170–1176. https://doi.org/10.1001/archpediatrics.2012.1276.

Miller, I. W., Camargo, C. A., Jr., Arias, S. A., et al. (2017). Suicide prevention in an emergency department population: The ED-SAFE study. JAMA Psychiatry, 74(6), 563–570. https://doi.org/10.1001/jamapsychiatry.2017.0678.

National Institute for Health and Care Excellence (NICE). (2022; reviewed 2024). Self-harm: Assessment, management and preventing recurrence (NG225). https://www.nice.org.uk/guidance/ng225.

National Institute of Mental Health (NIMH). (n.d.). Ask Suicide-Screening Questions (ASQ) toolkit. https://www.nimh.nih.gov/research/research-conducted-at-nimh/asq-toolkit-materials.

Stanley, B., Brown, G. K., Brenner, L. A., et al. (2018). Comparison of the Safety Planning Intervention with follow-up vs usual care of suicidal patients treated in the emergency department. JAMA Psychiatry, 75(9), 894–900. https://doi.org/10.1001/jamapsychiatry.2018.1776.

Suicide Prevention Resource Center. (n.d.). Caring for adult patients with suicide risk: A consensus guide for emergency departments. https://sprc.org/edguide/.

US Department of Veterans Affairs and Department of Defense (VA/DoD). (2024). Clinical practice guideline for assessment and management of patients at risk for suicide. https://www.healthquality.va.gov/guidelines/mh/srb/.

World Health Organization (WHO). (2021). LIVE LIFE: An implementation guide for suicide prevention in countries. https://www.who.int/publications/i/item/9789240026629.