Let's face it. With our most troubled, desperate patients, each visit is an exercise in doing something that may very well be impossible: predicting whether they are going to attempt suicide.

The famous "risk factors" for suicide are helpful but can do little to tell you if self-destruction is imminent, because they refer to long-term statistical probabilities. In fact, most people with these risk factors will never attempt suicide. Nonetheless, you should memorize them for two reasons: first, they serve as red flags during your assessment of potentially suicidal patients, prompting you to dig more deeply for serious suicidal ideation; and second, they are important to document from a medico-legal standpoint (see related article in this issue).

A new research finding sheds light on the usefulness of depression as a predictor of suicide.

The best way to memorize these risk factors is to use the famous "SAD PERSONS" mnemonic for the 10 major suicide risk factors (Psychosomatics 1983; 24:343-349). These include: Sex (more men complete suicides, more women attempt it), Age (elderly are at highest risk), Depression and bipolar disorder (each increases lifetime suicide risk 15-fold), Previous suicide attempt (strongest risk factor of all), Ethanol abuse (up to half of attempters used alcohol just before the attempt), Rational thinking loss (this refers to psychosis, but most studies to look at command hallucinations have found no significant association with an attempt), Social supports lacking, Organized plan, No spouse or child (especially for women, having a child decreases suicide risk), and Sickness (especially those leading to chronic pain and functional impairment).

Aside from these risk factors, clinicians might find it helpful to tuck away some "hard figures" about the rates of suicide in different populations. The lifetime risk of suicide in the general population is about 1%, and the lifetime risks for selected psychiatric populations are displayed in the following table (figures from Am J Psychiatry Practice Guidelines 2003; 160:11):

Psychiatric diagnosis Lifetime	Suicide Rate
General Population	1 %
Patients with a previous suicide attempt	27.5 %
Major Depression	15 %
Bipolar Disorder	15 %
OCD, Panic Disorder	7-8 %
Schizophrenia	6 %

As you can see, a prior suicide attempt is the strongest predictor, followed by a history of depression or bipolar disorder. However, knowing that depression is a risk factor for suicide doesn't help us much, since just about all of our patients have flirted with this diagnosis at one time or another. A recent study from a well-known group at Columbia University provides some intriguing information on how to translate depression into a more useful assessment of risk (Am J Psychiatry 2004; 161:1433-1441).

The researchers enrolled 308 patients with major depression, most of whom were inpatients at study entry. The patients were evaluated with a large number of psychiatric symptom scales, and then were reevaluated after 3 months, 1 year, and 2 years. Two years after the 308 patients were enrolled, four had committed suicide and 38 had attempted suicide.

Not surprisingly, by far the strongest predictor of a suicide attempt was a history of a prior suicide attempt; patients in this group were 4.4 times more likely to attempt suicide over the 2-year follow-up than those without such a history. What about the predictive value of depression? Oddly enough, a higher score on the Hamilton Depression Scale was not related to increased suicide risk; however, a higher score on a self-administered scale, the Beck Depression Inventory, increased the risk of a suicide attempt by a hefty factor of three. According to the authors, the take-home point of this finding is that as clinicians, we need to pay more attention to the patient's subjective sense of depression, pessimism, and hopelessness, and less attention to our standard survey of neurovegetative symptoms.

What about anxiety? In 1989, a famous article in the New England Journal of Medicine reported that 20% of all patients with panic disorder had attempted suicide, over twice the rate of other psychiatric disorders (N Engl J Med 1989; 321:1209-1214). Recent research has implicated anxiety and agitation as one of the more powerful acute predictors of a suicide attempt. For example, in a chart review study of 76 patients who had committed suicide while in the hospital, researchers found that fully 79% of these patients had met criteria for "severe or extreme anxiety and/or agitation." Ominously, 78% of these patients denied suicidal ideation during their last communication with a clinician (J Clin Psychiatry 2003; 64:14- 19). The implication is clear: highly anxious, agitated patients, regardless of the underlying diagnosis, are at higher risk of imminent suicide, and should be offered a robust dose of anxiolytics to help keep them safe.

Of course, the best way to assess for suicidality is to come right out and ask patients if they are planning a suicide, or have recently been thinking about it. It's not always easy to get a straight answer to such questions, however. Patients who are hell-bent on suicide may not want to reveal their plan, and others who are on the fence may be reluctant to share their thoughts for fear of being "committed." Dr. Shawn Shea, who is interviewed in this issue, provides some real nuggets of wisdom for how to gracefully extract sensitive information that may otherwise remain hidden.

TCR VERDICT:  Agitation + hopelessness + prior attempt = lethality.