Suicide risk assessment: Questions that reveal what you really need to know
How to quickly identify high-risk and protective factors in patients with suicidal behavior.
You can make more-informed decisions about a patient’s acute suicide risk—such as over the phone at 3 AM—if you know what to ask the psychiatry resident or crisis worker. For suicide risk assessment—especially when you have not seen the patient—you need specific, high-yield questions to draw out danger signals from large amounts of data.
We are not suggesting that a short list of questions is sufficient for this extremely difficult task. Rather—because we recognize its complexity—we offer the questions we find most useful when evaluating patients with suicidal behaviors.
American Psychiatric Association practice guidelines1 provide a comprehensive discussion of assessing suicide risk. In addition, we teach clinicians we supervise to probe for high-risk and less-commonly explored “protective” factors.
Mental health clinicians are more experienced in probing for high-risk factors than for protective factors. Because population studies offer limited help (Box 1),2 we ask clinicians these questions to evaluate the seriousness of a suicide attempt:
Why research offers limited help in assessing acute suicide risk
- Most studies of suicide risk factors focus on medium- to long-range risk
- Population-based risk factors (such as being Caucasian, over age 65, or depressed) apply to so many patients that their clinical usefulness is limited1
- Population-based risk factors often have high sensitivity but low specificity (recent loss is an important risk factor for attempting suicide, for example, but very few persons with a recent loss attempt suicide)
- In an acute situation, the positive predictive value of suicide risk factors—alone or in combination—is not known
3 important questions to ask in overdose cases
What method was used? Methods other than taking an overdose or cutting the wrists may be more dangerous.
What amount was used? (in overdose or poisoning cases)
What treatment was needed? If the patient took an overdose of opioids and needs intubation, this indicates a relatively serious attempt. On the other hand, the psychiatric seriousness of an acetaminophen overdose depends on whether the patient realized the danger in taking high doses of acetaminophen. Additional questions can help determine the seriousness of suicide attempts by overdose (Table).
Was the attempt impulsive or planned? Planned attempts tend to be more serious.
What is the ‘risk-rescue ratio’? The potential lethality of the attempt and the likelihood of being saved must be evaluated together. Where did the attempt occur? In a setting where others were likely to intervene? Was the patient alone? Attempts in the presence of others may be considered less alarming.
What did the patient do immediately afterward? Did he tell anyone? How did he get to the hospital? Did he seek help on his own? Who called the ambulance?
How does the patient feel about the attempt now? Is she glad or disappointed she didn’t die? Does she regret the attempt?
Have there been past attempts? Does the patient have a history of suicide attempts or significant selfmutilation? If so, what was the most serious incident? Past attempts tend to predict future attempts.
Other considerations for patients who have attempted suicide and those who have not but are being evaluated for possible suicide risk include:
Mental state. To estimate a patient’s mental state and depth of negative affect, without having seen her yourself, three helpful questions are:
- Does she still look upset, depressed, or angry? Anger and agitation tend to increase risk.
- Can she smile or relax, even briefly?
- Does she feel things are likely to improve?
Access to firearms. Suicide by firearms accounts for 55% of all suicides. 3 Does the patient have access to a gun and bullets? If so, get details. Does he now keep the gun on his person instead of in a locked cabinet, as he did earlier? We find that questions about guns are all too frequently omitted.
Contract for safety. Can the patient reliably contract not to harm himself and to call for help in a crisis? Although contracts for safety have limited value—as will be discussed—a patient’s refusal to contract for safety may indicate a higher suicide risk.
Some patients may refuse to contract for safety in order to be hospitalized for other reasons. They may say they can’t be sure what they will do if not admitted or declare that the clinician will be blamed for their death.
A patient’s high-risk clinical features must be balanced against factors that may reduce suicide risk.
How much social support? Can family or friends constantly stay with the patient, watch him closely, and get help if the situation worsens? This is the simplest and most relevant method to assess the availability of protective support.
How much mental health support? Does the patient have a good relationship with a psychiatrist or therapist who can see the patient soon?
Have circumstances changed? Has the stressor that led to suicidal behavior resolved, at least in part? For example, if a patient’s fight with her boyfriend led to her taking an overdose, have they made up?
Four common myths. Clinicians assessing patients for acute suicide risk often overestimate the protective value of some factors. They may tell you:
- The patient only tried to harm himself while he was intoxicated. He’s not intoxicated now and therefore is not at high risk. The patient will likely get intoxicated again, despite his protestations to the contrary. Substance intoxication and withdrawal tend to worsen depression and diminish inhibitions, making suicide more—not less—likely.
- The patient contracts for safety. A contract for safety may have some value, 4 but its clinical and legal merits in suicide risk assessment are overestimated. 5 We are concerned about how often we see clinicians judge that a contract for safety overrides numerous high-risk factors.
- The patient was only trying to get attention. It is difficult for mental health professionals—and even for patients—to reliably ascertain what motivates someone to attempt suicide. Multiple motivations and ambivalence are common.
- The patient is ‘just a borderline.’ Because patients with borderline personality disorder tend to make repeated suicide gestures, clinicians may not take their suicide attempts seriously. This statement reveals ignorance about the suffering of persons with borderline personality disorder; their rate of completed suicide is approximately 10%. 6
Trouble assessing high-risk and protective factors? Write them down
- When a suicide assessment is inconclusive, draw two columns on a sheet of paper. List the patient’s high-risk factors on one side and protective factors on the other
- Seeing the information in black and white often helps clarify the assessment
- Consider both the number of factors in each column and your clinical sense of each factor’s importance and intensity. Place a check mark next to particularly important factors
- This balance sheet can remind you of further questions to ask and often reveals that either the high-risk or protective factors far outweigh the others in number and/or intensity
Distinguish short-term vs long-term risk. Based on the questions above, we often conclude that a patient is at high long-term risk of suicide, but the immediate risk is much lower. Acute hospitalization is unlikely to alleviate the long-term risk (though sometimes is the only way to get the patient into psychiatric treatment).
Consider the source. Never disregard the “gut feeling” of the person who interviewed the patient, but also factor in your assessment of that clinician’s judgment. Sometimes inexperienced staffs’ intuitions may derive more from countertransference than from objective assessment.
Write it down. In cases where suicide risk seems unclear, it may help to list a patient’s risk and protective factors (Box 2). We have found this technique to be a useful teaching tool as well.
Be flexible. Because no method for assessing shortterm suicide risk is foolproof, be ready to re-evaluate your assessment and—if you are unsure—to take action to protect the patient.
1. American Psychiatric Association. Practice guideline for the assessment and treatment of patients with suicidal behaviors. Am J Psychiatry. 2003;160(11 suppl):1-60.
2. Fawcett J. Suicide risk factors in depressive disorders and in panic disorder. J Clin Psychiatry. 1992;53(suppl):9-13.
3. National Institute of Mental Health. Suicide facts. Available at: http://www.nimh.nih.gov/SuicidePrevention/suifact.cfm. Accessed June 3, 2004.
4. Stanford EJ, Goetz RR, Bloom JD. The no harm contract in the emergency assessment of suicidal risk. J Clin Psychiatry 1994;55(8):344-8.
5. Simon RI. The suicide prevention contract: clinical, legal, and risk management issues. J Am Acad Psychiatry Law 1999;27(3):445-50.
6. Paris J. Chronic suicidality among patients with borderline personality disorder. Psychiatr Serv 2002;53(6):738-42.