Crisis assessment, intervention, and prevention form a connected chain of practices aimed at recognizing when someone is in danger of harming themselves or others, stabilizing the immediate situation, and reducing the likelihood of future crises. The field has grown rapidly over the past two decades, with structured screening tools, brief therapeutic interventions, and community-based response models now backed by varying degrees of evidence. No single tool or program reliably predicts or prevents every crisis, but research has clarified which approaches make a measurable difference and which fall short of their reputation.
Structured Risk Assessment Tools
When someone arrives at an emergency department or calls a crisis line expressing suicidal thoughts, clinicians need a way to gauge the severity of the situation quickly. The Columbia-Suicide Severity Rating Scale, or C-SSRS, is one of the most widely adopted screening instruments for this purpose. It asks about suicidal ideation, its intensity (including how long the thoughts last), and any history of attempts. In a study of adolescent psychiatric emergency patients, the C-SSRS intensity score predicted whether a patient would return with a suicide attempt, and the duration of suicidal thoughts was a particularly strong signal among teens who already reported active ideation at their first visit.1PubMed Central. Columbia-suicide severity rating scale: predictive validity with adolescent psychiatric emergency patients
In adult emergency settings, the C-SSRS Screen Version has been tested at a larger scale. Both its ideation severity and behavior scales were associated with death by suicide within a week, a month, and a year. The optimal cutoff on the ideation severity scale was linked to roughly four times the odds of dying by suicide within one week.2Psychological Medicine. Columbia-Suicide Severity Rating Scale Screen Version: initial screening for suicide risk in a psychiatric emergency department That sounds impressive, but the picture is more complicated. A separate study of high-risk emergency department patients found that while the C-SSRS subscales did predict subsequent suicide attempts, the odds ratios were small. Among patients who had no prior attempt history, no C-SSRS subscale predicted a future attempt at all. The single strongest predictor was simply whether the person had attempted suicide before.3Suicide and Life-Threatening Behavior. C‐SSRS performance in emergency department patients at high risk for suicide
This tension runs through the entire risk-assessment literature. Structured tools are better than nothing and better than unguided clinical intuition, but they are not crystal balls. Earlier work on structured rating scales covering ten areas of suicide and violence risk found that admitted patients scored significantly differently from discharged patients on every scale, and scores predicted things like in-hospital violence and suicide precautions on the ward.4Comprehensive Psychiatry. Violence and suicide assessment scale in the psychiatric emergency room These instruments help clinicians decide who needs immediate hospitalization. What they struggle with is forecasting longer-term outcomes in the broader population, where base rates of suicide are low and individual variation is enormous.
Mental Health Triage Is Not Medical Triage
Emergency nurses who work in psychiatric settings describe a fundamentally different sorting process compared to medical triage. In physical emergencies, the threat to the patient’s life is the primary criterion: a cardiac arrest gets the highest level. In psychiatric emergencies, clinicians focus on the risk of harm to self or others, and behavioral cues like aggression can push a patient to a higher priority even when they are medically stable.5PubMed Central. Mental Health Triage from the Viewpoint of Psychiatric Emergency Department Nurses; a Qualitative Study This distinction matters because it shapes how staff are trained, how resources are allocated, and what happens in the first minutes of a crisis encounter. A person in severe psychological distress who appears physically healthy can still be the most urgent case in the department.
De-escalation and Acute Interventions
Once risk is identified, the immediate goal is to de-escalate the situation and keep the person safe. Verbal de-escalation training for psychiatric nurses has shown clinically meaningful results. One quality improvement project found that after training, the rate of patients placed in seclusion dropped from about 6% to about 4%.6PubMed. A Quality Improvement Project Using Verbal De-Escalation to Reduce Seclusion and Patient Aggression in an Inpatient Psychiatric Unit That reduction did not reach statistical significance in the study’s sample, but any drop in seclusion matters to the patients who avoid it. The broader lesson is that communication techniques can substitute for physical restraint in a meaningful number of cases.
For patients experiencing acute suicidality in the emergency department, researchers have been exploring pharmacological options alongside psychosocial ones. A scoping review found that a single dose of ketamine and a brief psychosocial approach called Crisis Response Planning both showed promise as interventions suited to the fast-paced ED environment. Three of four ketamine studies reported short-term reductions in self-reported suicidality and depression.7PubMed Central. Recent Interventions for Acute Suicidality Delivered in the Emergency Department: A Scoping Review Ketamine’s appeal is that it works within hours, unlike traditional antidepressants that take weeks. But it remains an emerging option, not a standard of care, and its effects on suicidal thinking specifically need longer-term study.
What Happens After the Crisis Passes
The period immediately after an emergency department discharge is one of the most dangerous windows for people who have been suicidal. Two approaches have shown the most consistent benefits during this transition: safety planning and follow-up contact.
Safety planning involves working with the patient before they leave the ED to create a written, personalized plan listing warning signs, coping strategies, people to contact, and ways to make the environment safer. A large cohort study compared safety planning plus telephone follow-up against usual care and found that patients who received the intervention were roughly half as likely to engage in suicidal behavior over the next six months. The rate dropped from about 5% in the usual-care group to about 3% in the safety planning group.8PubMed Central. Comparison of the Safety Planning Intervention With Follow-up vs Usual Care of Suicidal Patients Treated in the Emergency Department
Simple telephone follow-up on its own also has evidence behind it. In a randomized trial, patients who received a phone call one month after discharge had significantly fewer repeat suicide attempts than controls over the following six months, with the reattempt rate dropping from 22% to 12%. No deaths by suicide occurred in the contacted group. Interestingly, when the call was delayed to three months post-discharge, the benefit disappeared, suggesting that timing matters a great deal.9BMJ. Effect of telephone contact on further suicide attempts in patients discharged from an emergency department: randomised controlled study Text-messaging interventions for suicidal youth after ED discharge have also shown feasibility and acceptability, with participants reporting that the messages helped reduce suicidal thoughts and behaviors.10PubMed. A Brief Text-Messaging Intervention for Suicidal Youths After Emergency Department Discharge
Police-Based Crisis Response and Its Limits
Much of the public conversation about crisis intervention focuses on what happens when police encounter someone in a mental health emergency. Crisis Intervention Team training, known as CIT, is the dominant model. Officers receive specialized training to recognize signs of mental illness and connect people to treatment rather than arresting them. Early reports were encouraging: one study found that after CIT was implemented, more people in mental health crises were transported to treatment facilities voluntarily, and the proportion of calls involving possible mental illness went up, suggesting officers were identifying crises they previously might have overlooked.11PubMed. Crisis intervention team training for police officers responding to mental disturbance calls
The evidence for harder outcomes, though, is weaker than many advocates realize. A systematic review and meta-analysis found no significant effect of CIT on arrest rates for people with mental illness or on officer safety.12Criminal Justice Policy Review. Do Crisis Intervention Teams Reduce Arrests and Improve Officer Safety? A Systematic Review and Meta-Analysis A separate review reached a similar conclusion: CIT likely diverts people from jail to psychiatric facilities, but there is little peer-reviewed evidence that it reduces arrests, injuries, or use of force on objective measures.13PubMed. Effectiveness of Police Crisis Intervention Training Programs CIT may change the experience of the encounter and the pathway people take through the system, but the measurable downstream effects on the people in crisis remain uncertain.
Co-responder models, in which a mental health clinician rides alongside or meets officers on scene, have gained popularity as an alternative. A systematic review found that street triage might reduce the use of involuntary detention and police custody, but implementation varies widely across programs, and the overall evidence base remains thin.14PubMed Central. A systematic review of co-responder models of police mental health ‘street’ triage A randomized trial comparing police-mental health co-response to standard police response found no significant differences in emergency medical events, jail bookings, outpatient visits, or ED visits afterward.15PubMed. Police-mental health co-response versus police-as-usual response to behavioral health emergencies: a pragmatic randomized effectiveness trial Non-police models, which remove officers entirely, have shown some promising results in youth-specific programs and home-based crisis treatment, but studies tend to be low quality and hard to compare.16PubMed. Re-examining mental health crisis intervention: A rapid review comparing outcomes across police, co-responder and non-police models
None of this means these programs are worthless, but the field is in a curious position: crisis response models are being deployed and scaled based largely on face validity and practitioner enthusiasm rather than rigorous evidence of the outcomes that matter most.
Prevention Before a Crisis Happens
Moving upstream from crisis response, prevention strategies try to reduce the number of crises in the first place. Three broad categories dominate the evidence: gatekeeper training, school-based programs, and lethal means counseling.
Gatekeeper training teaches non-clinicians, such as teachers, coaches, parents, and coworkers, to recognize warning signs of suicide and connect at-risk individuals with professional help. QPR (Question, Persuade, Refer) is one of the most widely used gatekeeper programs. Training has been shown to improve participants’ attitudes toward intervening, their confidence in doing so, and their stated intention to take action.17Health Education Journal. The effectiveness of QPR suicide prevention training A small study found that before training, only about 10% of participants demonstrated adequate gatekeeper skills in observed role-plays; afterward, that figure jumped to 54%.18PubMed Central. Does a brief suicide prevention gatekeeper training program enhance observed skills?
However, a systematic review of gatekeeper training that focused specifically on randomized controlled trials found that the evidence for improvements in knowledge, appraisals, and self-efficacy after training remained unclear when held to that higher standard. Supportive results came mostly from uncontrolled pre-post studies, which are more susceptible to bias.19PubMed. Gatekeeper training for suicidal behaviors: A systematic review The gap between “people feel more prepared” and “fewer people die” has not been convincingly bridged for gatekeeper training alone. It is probably one necessary piece of a larger strategy rather than a standalone solution.
School-based programs have produced some of the strongest prevention evidence for younger populations. The SOS (Signs of Suicide) program, which combines education about depression and suicide with a screening component, was associated with roughly 40% fewer self-reported suicide attempts among participating students compared to controls over a three-month period. It did not, however, significantly change suicidal ideation or help-seeking behavior.20PubMed Central. Evaluating the SOS suicide prevention program: a replication and extension The Youth Aware of Mental Health (YAM) program, tested in U.S. adolescents, improved mental health literacy and decreased stigma around mental health, though it did not change participants’ stated intent to seek help.21PubMed Central. The Youth Aware of Mental Health Intervention: Impact on Help Seeking, Mental Health Knowledge, and Stigma in U.S. Adolescents These programs work best when woven into a school’s broader culture rather than delivered as a one-off assembly.
Lethal means counseling takes a different approach entirely: instead of changing how someone thinks or feels, it focuses on putting distance between a person in crisis and the means to act on suicidal impulses. Pediatric providers, for instance, are encouraged to integrate conversations about safe storage of firearms and medications into routine encounters. The logic is straightforward, because most suicidal crises are time-limited, and if a person cannot easily access a lethal method during the acute period, they often survive.22PubMed. Preventing Suicide Through Lethal Means Restriction in Pediatric Care Tools like the “Lock to Live” decision aid have been developed to help clinicians guide firearm owners through storage options during these conversations.23Injury Prevention. ‘Lock to Live’: development of a firearm storage decision aid to enhance lethal means counselling and prevent suicide
A Cautionary Note on Debriefing
One crisis intervention practice that has become deeply embedded in organizational culture, particularly in emergency services, is Critical Incident Stress Debriefing (CISD). In CISD, a facilitator leads a structured group session shortly after a traumatic event, encouraging participants to discuss their emotional reactions. Many organizations treat it as a standard duty-of-care response. The evidence, however, does not support this confidence. One review found that while CISD appeared to help emergency services personnel (secondary trauma victims), it was ineffective for people who had directly experienced the traumatic event, like accident or burn victims.24PubMed. The effectiveness of critical incident stress debriefing with primary and secondary trauma victims
A meta-analytic review published in The Lancet was even more damning. Comparing CISD to other interventions and to no intervention at all, it found that both non-CISD interventions and simple natural recovery improved PTSD symptoms, but CISD did not. CISD also did not improve recovery from other trauma-related conditions.25The Lancet. Psychological debriefing for prevention of post-traumatic stress and other disorders This is one of the clearest cases in the crisis field where a well-intentioned practice has persisted despite evidence that it does not work as claimed, and may in some cases interfere with the normal recovery process by re-exposing people to distressing material before they have had a chance to process it naturally.
Supporting People Bereaved by Suicide
Postvention, the work done after a suicide to support those left behind, is sometimes described as prevention for the next generation of crises. People bereaved by suicide face elevated risks of complicated grief, depression, and their own suicidal thoughts. The intervention research here is thinner than in other areas, but some patterns have emerged. A systematic review of controlled studies found some evidence that interventions targeting uncomplicated grief can be effective, though evidence for complicated grief was lacking. The most promising approaches were supportive, therapeutic, and educational, involved the bereaved person’s social network, and consisted of multiple sessions led by trained facilitators rather than one-off events.26PubMed Central. Effectiveness of interventions for people bereaved through suicide: a systematic review of controlled studies of grief, psychosocial and suicide-related outcomes
Another review found that bereavement support groups tended to reduce uncomplicated grief intensity, writing interventions helped with suicide-specific aspects of grief, and cognitive-behavioral programs were helpful specifically for bereaved people with high levels of suicidal ideation.27PLOS ONE. Grief interventions for people bereaved by suicide: A systematic review The key takeaway is that different grief responses may call for different interventions, and a one-size-fits-all bereavement program is unlikely to serve everyone well.
Machine Learning and Predictive Analytics
One of the more active frontiers in crisis prevention is using electronic health records to flag patients at elevated suicide risk before they reach a crisis point. Machine learning models trained on routine clinical data from children and adolescents have produced prediction accuracies with area-under-the-curve values ranging from 0.81 to 0.86, detecting over half of suicide-positive cases while maintaining 90% specificity.28PubMed Central. Machine learning for suicide risk prediction in children and adolescents with electronic health records When these models were validated across five different health care systems, they detected an average of 38% of future suicide attempts at 90% specificity, with AUC values between 0.71 and 0.76.29JAMA Network Open. Validation of an Electronic Health Record–Based Suicide Risk Prediction Modeling Approach Across Multiple Health Care Systems
Those numbers represent a genuine advance over unaided clinical judgment, but they come with a practical problem. Because suicide attempts are rare events, the positive predictive value of these models is low, often in the single digits. That means for every patient the model correctly flags, many more are flagged who will not go on to attempt suicide. In focus groups, health care providers expressed enthusiasm about the potential to identify high-risk patients who would otherwise be missed, but they also worried about alert fatigue, liability questions, and the added burden on already strained systems.30PubMed Central. Implementing Machine Learning Models for Suicide Risk Prediction in Clinical Practice: Focus Group Study With Hospital Providers The technology is promising, but deploying it effectively will require workflows that help clinicians act on the alerts without being overwhelmed by false positives.
Cultural Adaptation and Specialized Populations
Crisis interventions developed and tested on predominantly white, Western populations do not automatically transfer to other communities. A meta-analysis of culturally adapted trauma interventions found that adapted versions were more effective at reducing PTSD symptoms than non-adapted versions among racial and ethnic minority groups, with a moderate-to-large effect size. The adapted interventions also produced large reductions in anxiety symptoms compared to non-adapted ones, and participants reported more positive attitudes toward mental health services and greater engagement with care.31Trauma, Violence, & Abuse. Cultural Adaptations to the Assessment and Treatment of Trauma Experiences Among Racial and Ethnic Minority Groups: A Mixed-Methods Systematic Review and Meta-Analysis The gains in PTSD and anxiety did not persist beyond the immediate post-intervention period, and no significant difference was found for depression. But the finding that cultural adaptation improves at least short-term outcomes and engagement suggests that crisis services operating in diverse communities should be tailoring their approaches rather than assuming a universal model will suffice.
Peer Support in High-Risk Occupations
Certain occupations, including law enforcement, firefighting, corrections, and emergency medical services, carry elevated exposure to traumatic events and correspondingly higher rates of mental health difficulties. Peer support programs, in which trained colleagues provide emotional support and help connect coworkers with professional resources, have become increasingly common in these settings. A scoping review found that public safety workers viewed peer support as helpful for normalizing their experiences, increasing hope, and decreasing the stigma that often prevents people in these professions from seeking formal mental health care. The programs also appeared to bridge a gap between the reluctance many workers feel about seeing a therapist and the genuine need for support after critical incidents.32Workplace Health & Safety. Peer Support Programs to Reduce Organizational Stress and Trauma for Public Safety Workers: A Scoping Review In occupations where admitting vulnerability can carry professional consequences, having a trusted colleague rather than an outside clinician as the first point of contact may be the difference between getting help and suffering in silence.
Crisis stabilization units offer another piece of the infrastructure puzzle. These are short-stay facilities designed as alternatives to psychiatric hospitalization for people who need more support than an ED visit but less than an inpatient admission. One economic analysis of a regional pediatric crisis stabilization unit found an average per-admission cost of roughly $3,000, with an average stay of about 60 hours.33PubMed Central. An Economic Analysis of the Cost of a Regional Crisis Stabilization Unit That is substantially cheaper than a multi-day inpatient stay, and for many patients, a brief stabilization period with safety planning and connection to outpatient care may be all that is needed to get past the acute danger window.