Introduction

Suicide is a major public health problem. The World Health Organization reported that self-inflicted injuries including suicide accounted for more than 800 000 deaths in 2001.1 If every suicide affects at least six family members or friends, then every year in the world there would be about 5 million new survivors. Suicide rates range from 3.4 per 100 000 in Mexico to 6.0 per 100 000 in the UK and 34.0 per 100 000 in the Russian Federation.1,2 In the US, suicide accounts for about 30 000 deaths per year.3,4

More than 90% of suicide victims have a diagnosable psychiatric disorder, and most individuals who attempt suicide have a psychiatric illness.4–7 The most common psychiatric conditions associated with suicide or serious suicide attempts are mood disorders, but personality disorders, alcohol and substance abuse, anxiety disorders, and schizophrenia are also frequently associated with suicidal behaviour. The assessment of suicidal behaviour is the topic relevant to all clinicians not just mental health professionals. For example, in England, around 80% of patients with psychiatric disorders are treated in primary care.8

The problem of predicting suicide

Suicide and suicidal behaviours are statistically rare, even in populations at risk.3,9 Clinicians deal with almost no other event as infrequent as suicide. For example, although suicidal ideation and attempts are associated with increased suicide risk, most individuals with suicidal thoughts or attempts will not die by suicide. Suicide attempts and ideation occur in approximately 0.7% and 5.6%, respectively, of the general US population.10 In comparison, in the US, the annual incidence of suicide in the general population is approximately 10.7 suicides for every 100 000 persons, or 0.0107% of the total population per year.11 Even under the best circumstances, the attempted prediction of an infrequent behaviour such as suicide inevitably generates a large number of false-positive and false-negative cases.12–15 In addition, the prediction of suicidal behaviour is based on inexact criteria that are relatively poor at predicting the behaviour of a given individual.16–18 Predicting suicide at the level of individual patient is not possible at the present time, even among high-risk groups of patients.

Implications for clinicians

The task of suicide prevention remains a major challenge for clinicians. The general public expects clinicians to foresee suicide attempts and to protect patients from death by suicide.18 However, the expectation that clinicians can always predict and prevent suicidal behaviour among their patients is unrealistic. Clinicians are expected to gather information about a patient's clinical features, to document these features, and to use this information to formulate decisions about a patient's dangerousness to self and the treatment plan.3,18 Clinicians are obligated to make every effort to prevent suicidal behaviour in their patients.

Primary prevention

The ideal method of protection against suicide is primary prevention, i.e. reduction of number of new cases.6,9 It requires effective treatment of psychiatric disorders, especially mood disorders.6,7,19 Modification of social, economic, and biological conditions, such as reduction of poverty, violence, divorce rates, and promotion of a healthy lifestyle may significantly contribute to primary prevention of suicide.9

Secondary prevention

The goal of secondary prevention is to decrease the likelihood of a suicide attempt in the high-risk patients.6 The continuum of suicidal behaviour ranges from ideas to gestures, to risky lifestyles, suicide plans, suicide attempts, and, finally, suicide completion.6,9 Suicidal acts can encompass indirect self-destructive behaviours, such as chronic alcohol and substance abuse, pathological gambling, risky sports, reckless driving, playing Russian roulette, and self-mutilation.9 However, not all partly or indirectly self-destructive behaviour should be interpreted as suicidal.

The management of suicidal patients includes diagnosis and treatment of existing psychiatric illnesses, assessment of suicide risk, and the reduction of access to highly lethal methods for committing suicide, such as guns.4,5 An appropriate suicide risk assessment should be based on known risk factors. Suicide is generally a complication of a psychiatric disorder, but it requires additional risk factors, because most psychiatric patients never attempt suicide.6,7,20,21 The objective severity of psychiatric disorders does not assist in identifying patients at high risk for suicide attempt. Risk factors for suicide can be organized according to whether their effect is on the threshold for suicidal acts, or whether they serve mainly as triggers or precipitants of suicidal acts.6,7,22,23 A predisposition to suicidal behaviour is a key element that differentiates patients who are at high risk versus those at lower risk. Risk factors affecting the diathesis for suicidal behaviour include marital isolation, not living with a child under age 18, family history of suicide, parental loss before age 11, childhood history of physical and sexual abuse, alcohol and/or substance abuse, tobacco smoking, cluster B personality disorders, hopelessness, impulsiveness, aggression, low self-esteem, low cerebrospinal fluid 5-hydroxyindolacetic acid levels, low blood cholesterol levels, and physical illnesses.6,7,22–24 While risk factors are usually additive (i.e. the patient's level of risk increases with the number of risk factors), they may also interact in a synergistic fashion.24 Suicide rates are highest among men who are older than 69 years.5 Women have much higher rates of attempted suicide but lower rates of completed suicide than men.5 Most common precipitants of suicidal acts include the onset or acute worsening of a psychiatric disorder, interpersonal losses or conflicts, financial troubles, and job problems.4,6,7

Useful questions that should be considered in any evaluation for suicidal risk can be formulated as follows:25 How has the patient reacted to stress in the past, and how effective are his or her typical coping strategies? Has the patient contemplated or attempted suicide in the past? If so, how frequently and under what circumstances? What are the patient's current social circumstances, and how similar are they to past situations when suicide was attempted? Is the patient hopeless, helpless, powerless, angry? Does the patients have psychotic symptoms such as hallucinations or delusions?

The best predictor of suicidal behaviour is a history of a suicide attempt and current suicidal thoughts.4–6 Clinicians should regularly inquire about current depression, hopelessness, and suicidal ideation. It is sometimes surprising to find out how frankly patients may talk about their suicidal thoughts, when given an opportunity. The risk of suicide should be considered imminent if the patient reports the intention to die, has a suicidal plan, and has lethal means available.5 Expressions of despair and hopelessness also suggest an imminent risk. Comorbid depression and alcoholism may indicate the short-term risk of suicide, even in the absence of suicidal behavior.5,26 It is important to note that except for patients that actually admit to having suicide plans, it is extremely difficult to predict which patients in the high-risk population will actually kill themselves.

Tertiary prevention

Tertiary prevention is aimed at diminishing the consequences of suicide attempts.6 Improvement in professional education how to assess and treat suicidal persons may help rapidly detect and limit the damage that has occurred. Tertiary interventions include the assessment of family members who may be influenced by the suicide attempt to attempt suicide themselves.

Universal, selective and indicated interventions

In 1994, the Institute of Medicine Committee on Prevention of Mental Disorders recommended that prevention of mental disorders should be divided into three categories: universal preventive interventions, selective preventive interventions, and indicated preventive interventions.27,28 Suicide prevention efforts can also classified as either universal, selective, or indicated.29 A universal approach is designed for everyone in a defined population regardless of their risk for suicide, such as a health care system, or a county, or a school district. Ideally, cost per individual is low, and the intervention is effective. A selective approach is for subgroups at increased risk, for example, due to age, gender, ethnicity or family history of suicide. An indicated approach is designed for individuals who, on examination, have a risk factor or condition that puts them at very high risk, for example, a recent suicide attempt. Indicated preventive interventions may be reasonable even if intervention costs are high.

An individual who is determined to commit suicide will prevail despite the best efforts of health care professionals.5 However, most of the people who desire to kill themselves at one time will feel different after improvement in their psychiatric disorder and/or after receiving help with other problems.

It is important to note that our concern about prediction of suicide is related to our failure to prevent all suicides.15 Absence of suicide generates no data. If suicide is difficult to predict, its prevention is even more difficult to detect. It is likely that many suicidal individuals are recognized and successfully treated.

In summary, careful evaluations and appropriate treatments of patients with psychiatric illnesses, and social improvements may reduce suicide rates. Further research is needed to develop new approaches to treating and preventing suicidal behaviour.

References

1
World Health Organization. Burden of mental and behavioral disorders. In:
The World Health Report 2001. Mental Health: New Understanding, New Hope
 . Geneva, World Health Organization,
2001
:
19
–45.
2
Chishti P, Stone DH, Corcoran P, Williamson E, Petridou E. Suicide mortality in the European Union.
Eur J Public Health
 
2003
;
13
:
108
–14.
3
Hughes DH. Can the clinician predict suicide?
Psychiatr Serv
 
1995
;
46
:
449
–51.
4
Mann JJ. A current perspective on suicide and attempted suicide.
Ann Intern Med
 
2002
;
136
:
302
–11.
5
Hirschfeld RM, Russell JM. Assessment and treatment of suicidal patients.
N Engl J Med
 
1997
;
337
:
910
–15.
6
Sher L, Oquendo MA, Mann JJ. Risk of suicide in mood disorders.
Clin Neurosci Res
 
2001
;
1
:
337
–44.
7
Mann JJ. Neurobiolgy of suicidal behavior.
Nat Rev Neurosci
 
2003
;
4
:
819
–28.
8
Paton J, Jenkins R, Scott J. Collective approaches for the control of depression in England.
Soc Psychiatry Psychiatr Epidemiol
 
2001
;
36
:
423
–8.
9
Maris RW. Suicide.
Lancet
 
2002
;
360
:
319
–26.
10
Crosby AE, Cheltenham MP, Sacks JJ. Incidence of suicidal ideation and behavior in the United States,
1994
.
Suicide Life Threat Behav
 
1999
;
29
:
131
–40.
11
Minino AM, Arias E, Kochanek KD, Murphy SL, Smith BL. Deaths: final data for 2000.
Natl Vital Stat Rep
 
2002
;
50
:
1
–119.
12
MacKinnon DR, Farberow NL. An assessment of the utility of suicide prediction.
Suicide Life Threat Behav
 
1976
;
6
:
86
–91.
13
Pokorny AD. Prediction of suicide in psychiatric patients. Report of a prospective study.
Arch Gen Psychiatry
 
1983
;
40
:
249
–57.
14
Murphy GE. On suicide prediction and prevention.
Arch Gen Psychiatry
 
1983
;
40
:
343
–4.
15
Murphy GE. The prediction of suicide: why is it so difficult?
Am J Psychother
 
1984
;
38
:
341
–9.
16
Fawcett J, Scheftner WA, Fogg L, Clark DC, Young MA, Hedeker D, Gibbons R. Time-related predictors of suicide in major affective disorder.
Am J Psychiatry
 
1990
;
147
:
1189
–94.
17
Fawcett J, Clark DC, Scheftner WA. The assessment and management of the suicidal patient.
Psychiatr Med
 
1991
;
9
:
299
–311.
18
Cornelius JR, Clark DB, Salloum IM, Bukstein OG, Kelly TM. Management of suicidal behavior in alcoholism.
Clin Neurosci Res
 
2001
;
1
:
381
–6.
19
Oquendo MA, Kamali M, Ellis SP, Grunebaum MF, Malone KM, Brodsky BS, Sackeim HA, Mann JJ. Adequacy of antidepressant treatment after discharge and the occurrence of suicidal acts in major depression: a prospective study.
Am J Psychiatry
 
2002
;
159
:
1746
–51.
20
Mann JJ, Waternaux C, Haas GL, Malone KM. Towards a clinical model of suicidal behavior in psychiatric patients.
Am J Psychiatry
 
1999
;
156
:
181
–9.
21
Sher L. On the role of neurobiological and genetic factors in the etiology and pathogenesis of suicidal behavior among immigrants.
Med Hypotheses
 
1999
;
53
:
110
–11.
22
Sher L, Mann JJ. Neurobiology of suicide. In: Soares JC, Gershon S, eds.
Textbook of Medical Psychiatry
 . New York, Marcel Dekker,
2003
:
701
–11.
23
Van Heeringen K. The neurobiology of suicide and suicidality.
Can J Psychiatry
 
2003
;
48
:
292
–300.
24
American Psychiatric Association.
Practice Guideline for the Assessment and Treatment of Patients with Suicidal Behaviors
 . Washington DC, American Psychiatric Association,
2003
.
25
Goldman HH.
Review of General Psychiatry
 , 5th edn. New York, Lange Medical Books/McGraw-Hill,
2000
.
26
Sher L, Oquendo MA, Galfalvy HC, Grunebaum MF, Burke AK, Mann JJ. Suicidality in depressed patients with and without a history of alcoholism. In:
The 157th Annual Meeting of the American Psychiatric Association, New York, May 1–6, 2004
 . New Research Abstracts,
2004
, pp.
5
–6.
27
Mrazek PJ, Haggerty RJ.
Reducing Risks for Mental Disorders: Frontiers for Preventive Intervention Research
 . Washington DC. National Academy Press,
1994
.
28
Muños RF, Mrazek PJ, Haggerty RJ. Institute of Medicine Report on Prevention of Mental Disorders. Summary and Commentary.
Am Psychologist
 
1996
;
51
:
1116
–22.
29
National Strategy for Suicide Prevention: Goals and Objectives for Action
 . Washington DC, US Public Health Service.
2002
.