April 17, 2014


In the ED, our focus is on near term rather than long term risk of suicide, so we need to look at baseline as well as acute risk factors. Baseline, or chronic, risk factors are things that can’t be changed such as age, previous attempts and prior hospitalizations.

Acute risk factors are things that destabilize the patient and may be amenable to intervention. Examples are agitation, worsening depression, and acute suicidal ideation.

Outlined below is the TRAAPPED SILO – SAFE mnemonic for Suicide Risk Stratification


Is there a specific event or circumstance that led up to this? Often there is, but sometimes the patient can’t identify anything in particular. The patient’s subjective perception of stressful life events is far more important the quantity or type of event.8,15,36

A trigger can be anything that acts as an acute stressor. For example:

  • Loss. Loss of a Job, relationship, financial support, loved one
  • Disappointment or embarrassment -Anniversary of significant interpersonal loss
  • Recent (failed) psychiatric admission or suicide attempt
  • In teenagers, recent suicide by a peer A trigger is an acute or destabilizing risk factor.

Rational thinking loss

This is one of the most important acute risk factors and where your treatment can have a major effect. Examples of this are psychosis, agitation and debilitating anxiety, all of which increase suicide risk. There is evidence that the more severe the anxiety or agitation, the higher the near term risk of suicide. Itʼs unlikely that you are going to send an agitated or acutely psychotic patient home, but even so, there can still be an increased risk of suicide during hospitalization.5 Itʼs your job to manage the acutely agitated patient, and the assessment may change when the their condition improves. Itʼs possible theyʼll go from high risk to lower risk.


Two age groups, 15 to 24 and greater than 65, should raise a red flag. Older patients are less likely to give warning that suicide is imminent and are morelikely to use violent and potentially fatal methods. The ratios of attempts to completions speak for themselves: Young 200:1, Elderly 4:1.8,15

In the 15-24 age group, suicide is the third leading cause of death behind unintentional injuries and homicide. Possible reasons for this are age related emotional lability and impulsiveness.17,23

Access to means

Are there firearms, potentially lethal medications or other easily accessible means of suicide in the household? Suicide can be completed in many ways. Even an extension cord can be used for hanging. It’s impossible to remove all possibilities from the equation, but some methods, such as self inflicted gunshot wounds, carry a much higher likelihood of death.

In the US, firearms contribute to 60% of all suicide deaths and are the most common method across all age groups and gender.6,36

Previous Attempts

When you see someone who has had 5 suicide attempts, it can be reflexive to think, “Well, obviously they donʼt mean it. Theyʼve tried 5 times and theyʼre still alive.” Having a previous suicide attempt is associated with a 38 fold increase in suicide risk.26 Thatʼs higher than the risk associated with any psychiatric disorder including depression, schizophrenia, and personality disorders.36,43

The risk is probably higher in the first year after an attempt12, and you should account for the highest potential lethality of their most serious attempt. A lethal attempt is like breaking through any barrier. The door is open and itʼs much easier to get to that place when it’s been done before.

Previous psychiatric care

Higher levels of previous treatment intensity, such as inpatient hospitalization, increase risk. The question to ask is, “Have you ever been hospitalized for mental health issues?” This can be thought of as a proxy for the presence and severity of underlying mental illness.

What about recent hospitalization? This is a classic situation where youʼd think, “You were just hospitalized for this, and now youʼre back? Surely there must be some secondary gain.” Maybe so, but recent hospitalization can be destabilizing and increases the risk for suicide. Weʼd like to think that a patient is protected from self harm by being hospitalized for management of acute risk factors, but evidence for this is lacking. So should we stop hospitalizing high risk patients?

No, but be aware that mental health hospitalization is not a panacea, and although it may help ameliorate acute issues, it can also increase risk.9,16,19,26,30,48

Excessive ETOH or drugs

Up to 50% of patients who die by suicide have alcohol in their system. Many acutely intoxicated alcoholics either don’t remember the reason for their attempt or did it on a sudden impulse. Alcohol used in conjunction with a suicide attempt is associated with repeated suicide attempts and future suicide.12,27,36

What about the patient who comes in drunk and suicidal, but once sober, denies SI? They are still at risk for suicide down the road. If your patient is intoxicated with alcohol or drugs and ultimately discharged, substance abuse counseling needs to be part of the discharge plan. It often feels like it’s just going through the motions and while that can be true, it’s not just pro forma. Sometimes chemical dependency treatment works. On more than one occasion, I’ve seen cases where a patient makes a potentially lethal attempt after an ED visit and the main criticism of the documentation was that there was no mention of a plan to address the risk factor of alcohol abuse. Would it have made a difference? Unknown. Should it be part of the discharge plan? Yes.23,26

Depression and Hopelessness

The presence of major depression increases the risk of suicide by 20 fold. For patients with depression, hopelessness can factor into why some choose suicide and some do not.34

Why is this?

Hopelessness is a cognitive end point of ’no other way out’ and an indicator of depression severity. The more severe the depression, the higher the risk. The less severe the depression, as in a patient you’d more likely classify as dysthymic, the lower the risk. Not zero risk, but lower.11,16,40


What is it about having a medical illness that increases suicide risk?

It’s not the illness itself. It’s the secondary effects like severe pain and functional limitations that lead to depression. Aspects to look for are functional impairment, chronic pain, and increased dependence on others.17,36

The medical diagnoses most associated with increased suicide risk are: epilepsy, CNS disorders, and malignancy, but the details of which particular diagnoses do or do not increase risk is generally beyond the realm of emergency medicine. If the disease affects the patient’s ability to work, function in society, see, hear, or live independently, consider it a risk factor.17


A patient’s communication of suicidal ideation or intent is a heavily weighted warning sign and risk factor. Pay close attention to the intensity of the suicidal thoughts.11 Are they fleeting? Are they unrelenting and obtrusive? Has the ideation led to formulation of plans?

Suicidal ideation is a major precursor to attempted or completed suicide, but if a patient denies SI, are they automatically low risk? No, this is only one risk factor out of many. Denial of suicidal ideation by itself doesn’t erase suicide risk. It has to be taken in the context of each presentation.7

Frequent, intense thoughts of suicide often precede an attempt. If ideation and future intent are present, risk increases. Here’s the caveat, and this comes from a 2003 study in the Journal of Clinical Psychiatry looking at inpatient suicide: 78% of inpatients who committed suicide either in the hospital or shortly after discharge denied suicidal ideation in their last psychiatric interview before suicide.5

If you’re seeing a discrepancy with what you perceive as the patient’s risk level and their denial of SI, try saying something like this: “You’ve told me that you aren’t thinking about suicide and that you don’t want to die, but your behavior and what your family says suggests otherwise. It almost seems like you’re telling me one thing but thinking another. Help me make sense out of this.”

Lack of Social support/Isolation

Itʼs hard to objectively measure social support. A patient can come in with a family that expresses concern and empathy, but the patient can feel that they are isolated and donʼt have any support. Patients who perceive that they have supportive interpersonal relationships are at lower risk for suicide. The absence, or perceived absence, of support increases risk. In the elderly, family discord is a red flag.20

The complement to social support is collateral information. Mental health patients in the ED are not always forthcoming with information and may withhold key pieces of the history. Collateral history from friends, family, coworkers, etc, is essential. If itʼs not possible; no oneʼs coming in; no oneʼs available or wants to talk by phone, thatʼs a red flag that there may be a lack of social support.

Organized or serious attempt

Should you ask your patients details about suicidal ideation or plans? Yes. Can it plant the idea in their head and increase chance of suicide? No. There’s no evidence that talking about suicide increases the likelihood of subsequent attempts.15,23

The intent to die can be more important than the potential lethality of the method. 26 For example, a patient takes 5 grams of ibuprofen thinking that itʼs going to kill them. Lethal overdose? Probably not. High suicide risk? Absolutely. Another patient takes 5 grams of diltiazem because that was in the closest bottle they grabbed when they were mad at their spouse for watching football. Lethal overdose? One of the worst. High suicide risk? Much less so than example one.

Start off with the details of the attempt. Was there an attempt or are they just thinking about it? The more detailed and specific the plan, the greater the risk. Plans to avoid detection or use lethal methods are very high risk.12 Some questions to ask:

“What did you think would happen?” Did they want to go to sleep, get a reaction out of somebody, want to die? “What thoughts were you having beforehand that led up to this attempt?” This gives an idea of their mental state and premeditation.

“How do you feel now?” Are they relieved that theyʼre not dead, are they disappointed that they didnʼt die? If they didn’t carry out the attempt, ask them what kept them from doing it. This can give insight into potential protective factors.

Before seeing the patient, I’ll try to get as much possible information from sources such as the social worker and medical records. By the time I start taking the history, several of the TRAAPPED SILO points already done. Areas I want to address with the patient are mental status, agitation, previous attempts, and, if they came in following an attempt, their intended lethality.

Are some factors more heavily weighted than others? Statistically, some risk factors like previous attempts, alcohol, ideation, and depression could be seen as more important, but a risk factor or destabilizing force is going to have a unique effect on each patient. That’s why there isn’t a score at the end of this. Risk factors need to be taken in the context of each patient presentation.

Protective Factors [SAFE]

One challenge with suicide assessments is that the risk factors correlating with suicide risk are very common, but suicide itself is relatively uncommon. This leads to a high false positive rate when identifying suicide risk. So why do some attempt and others don’t? The reasons are multifactorial, but if there are factors that increase risk, it follows that there are factors that decrease risk.34 Look at depression: compared to the general population, the rate of suicide is 20 times higher, but 99.9% of depressives don’t attempt suicide.16,26 What differentiates attempters from non attempters? The last part of the mnemonic, SAFE, examines protective factors.


A sense of support and belonging has tremendous importance. The essential ingredient is having a feeling of connection to a community or network such as friends, family, religious organizations, etc. Religion can play several roles as a protective factor. There’s the social support piece, but a moral objection to suicide is also protective. Being married and having children in the home can be protective, while being single, divorced, or widowed are risk factors.1,13,14,34


Good insight, coping skills, and decision making.

Future oriented

Thinking about and looking forward to future events. Patient has ‘reasons for living.’


Not in the marital sense, although marriage is a protective factor, but engaged in the therapeutic process and not withholding information or resisting the evaluation. A willingness to participate

Suicide Risk Stratification

Related posts:

About Rob Orman

Deceptively tall.
Podcaster to the stars, father of many, lover of few

Awesome article, I know - please share your erudite thoughts...