Suicide is a serious public health problem that ranks as the 11th leading cause of death in the United States. Within the 15- to 24-year-old age group, it is the third leading cause of death.1 Many suicide victims have had contact with the mental health system before they died, and almost one fifth had been psychiatrically hospitalized in the year before completing suicide. A recent review found that psychiatric illness is a major contributing factor to suicide, and more than 90% of suicide victims have a DSM-IV diagnosis.
Suicide is a serious public health problem that ranks as the 11th leading cause of death in the United States. Within the 15- to 24-year-old age group, it is the third leading cause of death.1 Many suicide victims have had contact with the mental health system before they died, and almost one fifth had been psychiatrically hospitalized in the year before completing suicide.2,3 A recent review found that psychiatric illness is a major contributing factor to suicide, and more than 90% of suicide victims have a DSM-IV diagnosis.4,5 Because of these statistics, the assessment of suicide risk is a much needed core competency in psychiatric practice and has recently been recognized as a core curriculum requirement in the residency training of psychiatrists. The American Psychiatric Association (APA) has also recognized the importance of this clinical skill with its publication of a practice guideline for the assessment and treatment of patients with suicidal behaviors.6
Unfortunately, many clinicians take 1 of 2 approaches to suicide risk assessment.7 Some are overly cautious and assume that anyone who reports suicidal thoughts is at high risk for suicide. This approach leads to unnecessary deprivation of patients' liberties and rights as well as to the squandering of scarce clinical resources. Others underestimate suicidality through a dismissive attitude or inept assessment, thereby jeopardizing patient safety and increasing physician liability should there be a negative outcome.
Suicide is rare. Like an earthquake and other rare events, it is difficult to predict; attempts at prediction have led to a high number of false-positive predictions.8 For that reason, suicide prediction is not the goal of risk assessment. Rather, risk assessment serves a dual purpose.
First, more than any psychiatric intervention (other than successful treatment of an eating disorder) it has the potential to reduce patient mortality. While not aimed at predicting who will likely commit suicide, its purpose is to identify modifiable or treatable acute, high-risk suicide factors and available protective factors that will inform and guide patient treatment as well as safety management.9
Second, in the event of a malpractice suit, a thorough and documented risk assessment can help establish that a psychiatrist was not derelict in the duty to practice in a manner that adheres to a reasonable standard of care.
A survey of primary care physicians who lost a patient to suicide found that a risk assessment was only completed in 38% of cases.10 Studies of risk assessment documentation in psychiatric practice are lacking, although since 1998, suicide and attempted suicide account for 15% to 16% of malpractice claims by cause of loss in the United States.11,12 In lawsuits involving patient suicide, the determination of whether actions of the psychiatrist met the standard of care is made by a review of certain factors:
This article will outline the principles of suicide risk assessment from an evidenced-based and liability-prevention standpoint and provide a model for medical record documentation from a risk management perspective.
Evaluation of risk factors and protective factors
Suicide risk assessment begins with a thorough psychiatric interview with 3 goals: (1) to identify the specific factors that increase or decrease the risk for suicide; (2) to address the immediate safety of the patient so that treatment can take place in an appropriate setting; and (3) to identify psychiatric illnesses that can be targeted in treatment.
Numerous risk factors have been empirically linked to an increased risk of suicide. The APA Practice Guideline lists 56 specific risk factors divided into 10 headings: suicidal thoughts/behaviors, psychiatric diagnosis, physical illnesses, psychosocial features, childhood traumas, genetic and familial effects, psychological features, cognitive features, demographic features, and additional features.13 Some of these factors increase suicidal risk considerably more than others, and a clinical approach to risk assessment should be built around them. Risk factors to be examined in a fundamental suicide risk assessment are outlined in Table 1.
Several demographic factors have been associated with increased suicide risk: race (ie, white or Native American),13 male sex,14 marital status15 (ie, widowed, divorced, single), and age group (ie, adolescent or elderly).16 These variables should be documented. In addition, a detailed description of suicidal thoughts, plans, behaviors, and intent should be explored, as well as the specific suicide methods considered and their potential lethality.
The frequency, intensity, and duration of suicidal thoughts should also be documented. Having a suicidal plan is much more predictive of completed suicide than mere suicidal desire.17 In a plan involving a weapon, the clinician should inquire about the accessibility of the weapon and should note if there has been a recent movement of the weapon (particularly if a firearm) within the home. In addition to current suicidal ideation and plan, a thorough history of past suicide attempts and their seriousness (potential lethality), past suicidal thoughts, and the circumstances in which they occurred (eg, while alone, while drinking, after a job loss, during/ after a divorce) should be documented.
One guiding principle of risk assessment is that the current presentation should be compared with previous situations in which a serious suicide attempt occurred. If the situations are similar, a patient may be at increased risk. There is evolving evidence that patients who have a history of multiple attempts are at elevated risk when compared with those who have a history of suicidal ideation or only one attempt.18 The combination of 2 factors-having a plan and a history of multiple attempts-is particularly worrisome.19
Among precipitants and stressors, the clinician should inquire about recent significant losses (eg, financial, interpersonal, identity), acute or chronic health problems, and family instability.20 Studies have repeatedly shown that people who completed suicide were more likely to have established a treatment relationship with a primary care provider and to have frequently visited their provider before the suicidal act.21-23
All psychiatric diagnoses, with the exception of mental retardation, have been shown to increase suicide risk.24 Among psychiatric diagnoses, major depression leads to a 20-fold increase in lifetime risk, and suicide risk in patients with bipolar disorder has been found to be similar to that of patients with major depression. Mood disorders, primarily during depressive phases, are the most frequent diagnoses in completed suicide.25,26 Of all the symptoms of depression, hopelessness has been identified as being greater in suicide attempters versus nonattempters, even when the severity of depressive symptoms was rated the same.27 For this reason, clinicians should assess not only the presence of hopelessness but also its severity and duration.
The risk of suicide in patients with schizophrenia is estimated to be 8.5-fold higher than in the general population. Drug abuse, depressive symptoms, and a higher number of total hospitalizations are associated with increased suicide risk in people with schizophrenia.28 Studies have not clearly established the role of command auditory hallucinations in suicide risk, and this warrants further study.29,30 Contrary to common belief, command hallucinations account for a small percentage of suicides in persons with schizophrenia. In fact, suicide in patients with schizophrenia may be more likely to occur during periods of improvement after a psychotic episode, when insight into the implications of having a schizophrenic illness induces hopelessness or depressive symptoms.31,32
Finally, personality disorders are common in suicide attempters, with overall rates of about 40%.26 Borderline and antisocial personality disorders are the most common types associated with suicide attempts.33
In a person with suicidal ideation, alcohol use and intoxication can be disinhibiting and thereby lead to suicide attempts. Alcoholism alone increases suicide risk 6-fold.34 Persons with alcoholism and major depression are more prone to suicide than individuals with major depression or alcoholism alone, and major depression has been found in 50% to 75% of persons with alcoholism who commit suicide.35 Suicide has also been associated with the quantity of alcohol consumed per day but not with drinking frequency.36 Abuse of other substances has also been associated with an increased suicide rate, especially among younger individuals.37 In an analysis of completed suicides in 13 states, 33.3% tested positive for alcohol and 16.4% tested positive for opiates.38
A positive family history of suicide increases suicide risk, probably through both genetic and environmental effects. Genetic studies have shown that identical twins have a higher concordance of suicide and suicide attempts than do fraternal twins.39 Other studies have consistently shown that there is an approximate 4.5-fold increased relative risk of suicide completions or suicide attempts in close relatives of suicidal patients compared with relatives of nonsuicidal individuals.40 Thus, a family history of suicide and suicide attempts should be obtained.
Finally, a clinician should inquire about psychosocial risk factors in a patient expressing suicidal ideation, including the patient's employment status and current support systems. Unemployment is associated with increased rates of suicide, and suicide rates are higher in socioeconomically deprived geographical areas.41 In estimating suicide risk, both the patient's social supports and the patient's perception of available social supports should be assessed. Living alone as well as social isolation may increase the risk of suicide.42,43
While the above risk factors (Table 1) of suicide are not comprehensive, they serve as a basis for a suicide risk assessment. However, in addition to factors that increase suicide risk, there are known protective factors. These factors are listed in Table 2. It is useful to ask suicidal patients about their reasons for living. Also, the Reasons for Living Inventory, a self-report instrument measuring patient beliefs that may contribute to the inhibition of suicidal behavior, can be used to formally identify protective factors.44 The presence of children and the number of children in the home appear to decrease suicide risk in women.45 In addition, religious belief is protective. In the United States, Catholics have the lowest suicide rate, followed by Jews and Protestants.46 However, it appears that the strength of religious beliefs is more protective than a specific religious belief system per se.47 Finally, the presence of good social supports is a protective factor, regardless of whether support comes from family members or others.48 Social supports can also be used in the management of suicidality, as outlined in the next section.
|Children in the home and sense of family responsibility|
Management and documentation
The results of a formal suicide risk assessment should be organized and well documented, especially the identified risk factors and protective factors. Risk factors can be divided into 2 types: static and dynamic. Static risk factors are those factors that cannot be changed and therefore are not used as a target for treatment interventions. They include race, age, gender, marital status, history of suicide attempts, and family history of suicide. Dynamic risk factors, on the other hand, can be targeted for treatment intervention. These include current suicidal thinking (including plan and intent), current psychiatric illness and symptomatology, current interpersonal or life crises, alcohol or illicit drug use, unemployment, and lack of social support. A list of the dynamic risk factors with a corresponding intervention designed to address each risk factor indicates that the clinician has successfully identified the significant risk factors and has taken steps to develop a comprehensive approach to treatment.
The first and most important step in the management of suicidality is to decide in what setting treatment may safely occur. Consideration should be given to hospitalization for patients with significant intent or who have a history of attempts in similar circumstances. Adjustments in medication dosage, changes in medication, more frequent appointments, temporary relocation with a family member or increased family involvement in the patient's treatment, removal of access to weapon(s) or changing weapon storage practices,49 attendance at Alcoholics Anonymous or Narcotics Anonymous meetings, referral to vocational rehabilitation, and other strategies may be considered as the clinical presentation warrants. The use of a no-suicide contract is controversial and cannot take the place of a formal suicide risk assessment.50 Furthermore, these contracts do not protect a clinician from malpractice liability.51 Finally, suicide risk assessment is not a one-time event. Serial assessments should be done as warranted by the clinical picture and when significant changes occur in the patient's treatment.
Documentation is the cornerstone to the defense of a potential lawsuit resulting from a patient suicide. While failing to document a suicide risk assessment is not usually by itself a cause of patient suicide, the quality of documentation can determine whether a malpractice attorney accepts or declines a suicide malpractice case52 (although failing to document is a deviation in the standard of care). The common belief among attorneys is, "If it isn't written down, it didn't happen." Good documentation does several things: it records clinician competence at identifying suicide risk factors and designing interventions, it makes a challenge to the physician's standard of care much more difficult, and it helps prevent suicide by creating a record that the physician can review and use at future patient encounters. In the process of documentation, patient comments that indicate decreased suicide risk, such as "I would never kill myself," should be quoted in the record; why the patient may be unlikely to attempt suicide should also be noted. Patient quotes can carry significant weight with jurors.
In general, it is more important to document the details of decisions that increase suicide risk rather than those that decrease it. For example, more documentation of suicide risk assessment is indicated when a patient is discharged from a hospital following a suicide attempt than when a decision has been made to hospitalize a potentially suicidal patient from an emergency department. When a clinician consults with a colleague in the course of risk assessment, that consultation should also be documented. Consultation with another psychiatrist before the discharge of a potentially suicidal patient is highly recommended.
Liability prevention in the event of a completed suicide
Postsuicide entries in a medical chart are looked upon with great suspicion and will be framed by plaintiff's counsel as purely self-serving.53 Also, never alter or destroy parts of a patient record after an adverse incident. Reaching out to the family of a patient who committed suicide can be an important clinical and compassionate response, and risk management strategies usually support such a response.54 However, psychiatrists have an obligation to maintain patient confidentiality after death, even in the face of family members' understandable desire for information.55 Therefore, treatment information should not be released without valid authorization from a person legally authorized to give consent after death. Such requests should be placed in writing and be accompanied by written documentation of the requester's legal authorization.
Finally, losing a patient to suicide can have a significant emotional impact on the treating psychiatrist. Attendance at the patient's funeral may occur after consideration of several factors: the relationship with the patient's family, the family's current attitude towards the psychiatrist, and the psychiatrist's own feelings about the patient's suicide. Should the psychiatrist need to process his or her feelings, this should be done in the context of a therapeutic relationship and not with family members, colleagues, or others.