Sarah M. Jacobs, MEd

Comment on this article

Completed suicide provokes a multitude of questions: What motivated it? What interventions could have diverted it? Could anyone or anything have prevented it? The question of who dies by suicide often overshadows the question of what lessons suicide attempt (SA) survivors can teach us. Their story does not end with the attempt episode. For these patients, we have ongoing opportunities for interventions to make a difference.

A history of SA strongly predicts eventual completion, so we must try to identify which survivors will reattempt and complete suicide. This article addresses what is known about the psychiatry of suicide survivors—suicide motives and methods, clinical management, and short- and long-term outcomes—from the perspective that suicidality in this population may be a trait, with SA or deliberate self-harm (DSH) as its state-driven manifestations. When viewed in this manner, it is not just a question of who survives a suicide attempt, but who survives suicidality.

CASE REPORT: End of the game

Ms. T, age 39, was admitted to the intensive care unit after an aspirin overdose. She had been living with a man in a southern state for 8 years since the demise of her first marriage, but kept deferring remarriage. She returned to Minnesota with her teenage daughter to visit her family and stayed 6 months. Her partner phoned Ms. T every day, telling her he wanted her to come back. One day he tired of the game and said, “Fine, don’t come back.” She immediately overdosed, then called him to tell him what she’d done. He called her daughter, telling her to go check on her mother and to call 911. When later asked why she did it, Ms. T said, “So he would know how much he loved me.”

Motive for self-harm

Ms. T’s suicide attempt was nonlethal, and she reported it immediately—characteristics of parasuicidal gesturing as a motive. A useful categorization of suicidal behavior divides it into discrete categories or narratives. Gardner and Cowdry describe 4: true suicidal acts, parasuicidal gesturing, self-mutilation, and retributive rage.¹ We modify this schema with 4 additional categories: altruism, acute shame, command hallucinations, and panic ( Table 1 ).^1-3 Categories are differentiated by affective state, motivation, and goal of behavior, but all involve situations in which the individual feels a lack of other options and resorts to maladaptive strategies.

Although this classification scheme helps clinicians understand a patient’s mindset, the specific motive underpinning DSH or SA is not consistently linked to its lethality. True suicidal acts frequently are marked by careful planning and high-lethality methods that increase the risk of completed suicide, but any motive can lead to a lethal act, whether or not death was intended.^2,3

Factors that increase the risk of SA and completed suicide include male gender, age (adolescent or age >60), low socioeconomic status, and alcohol or drug abuse.⁴ An underlying mood disorder accounts for 73% of attributable risk of suicide or medically serious SA in older adults.⁵ This connection between mood and suicidality highlights the concept that emotional pain can cause so much suffering that patients seek release from distress by ending their lives.

A useful model by Shneidman⁶ casts psychological pain as 1 dimension in a 3-dimensional system that includes press and perturbation. In this model:

• Pain refers to psychological pain (from little or no pain to intolerable agony).
  
• Press means actual or imagined events in the inner or outer world that cause a person to react. It ranges from positive press (good fortune, happy events, protective factors) to negative press (stressors, failures, losses, persecution), which in turn decrease or increase the likelihood of suicide.
  
• Perturbation refers to the state of being disturbed or upset.
  

Disruption in any 1 dimension tends to disturb the other 2 dimensions. When all 3 dimensions reach maximum distress, the stage is optimally set for suicide. DSM-IV diagnoses are examples of manifestations of high levels of perturbation. Perturbation is especially conducive to suicide when it involves constriction (a sense that the only viable option is death) and a penchant for self-harmful action.

Certain risk factors make SA simultaneously more likely to occur but less likely to be lethal. For example, parental discord, nonheterosexual orientation, and female gender have been found to increase non-fatal attempts among adolescents.⁷ Borderline personality disorder increases the reattempt rate out of proportion to completion among adults.⁸ One might interpret a pattern of repeated nonlethal attempts to mean the patient has no real intent to die, but this is not always the case.

 

⁸

Table 1

8 categories or narratives of suicidal behavior

Motive	Characteristics
True suicidal act	Release from intense baseline despair/hopelessness; self-nihilism as a permanent end to internal pain (entails highest intent to die and highest risk of completed suicide)
Self-mutilation	Relieving dysphoria or dissociation/depersonalization; acts of DSH designed to self-regulate or distract from emotional pain or other overwhelming affects
Retributive rage	Revenge; impulsiveness, vengefulness, and reduced capacity to conceive of other immediate options
Parasuicidal gesturing	Communication designed to extract a response from a significant other; often repetitive acts of DSH, strong dependency needs
Acute shame	Penance designed to escape from or to atone for a shameful act; often occurs within a short time after act is committed
Altruism	Relief of real or imagined burden on others; often occurs in setting of medical illness or substantial financial concerns
Command hallucinations	Acting in compliance with a command hallucination; often in setting of schizophrenia or depression with psychotic features
Panic	Driven by agitation, psychic anxiety, and/or panic attack; action intended as escape from real or imagined factor provoking agitation
DSH: deliberate self-harm
Source: References 1-3

CASE REPORT: Caught in the act

Mrs. L, age 35, works at a nail salon and took $12 from the cash register to buy gas so she could visit her husband in the next town. She’d never done anything like that before. She planned to return the money the next day, but her act was captured by a security camera and reported before she had a chance. Her boss said she had to go to the police.

Mrs. L was so ashamed that she decided she wanted to die. She drove her car to a remote hunting area where she tried to shoot herself in the head. The gun bucked, however, and shot her in the shoulder instead. She climbed into the front seat and drove herself to the hospital.

Method of self-harm

Survival of a suicide attempt depends in part on the lethality of the suicide method. Although she survived, Mrs. L’s attempt was intended to be quite lethal and illustrates shame as a motive.

The method’s lethality does not always correlate with the intent to die.⁹ Attempters with the highest suicidal intent do not reliably choose the most lethal method, either because they overestimate the lethality of methods such as cutting or overdose or because less lethal methods were most accessible.

Physicians have a higher suicide rate than the general population,¹⁰ probably because of their knowledge of lethal means. Patients with greater access to information on method lethality may be more likely to match their method with their intent, thus raising the risk of suicide completion.¹¹ Patients who use a high-lethality method on the index SA tend to continue using high-lethality methods on reattempts, which makes eventual fatality likely.¹²

Firearms, which are both accessible and lethal, remain the most common and deadly method in the United States, with more suicides from gunshot than all other methods combined.¹³ Cultural factors also are involved, such as in India where poisoning (especially with readily available organophosphates) is more common than gunshot.¹⁴ Suicidality screening in psychiatric practice and in the emergency department should always include questioning about convenient access to lethal means, especially those commonly used among the local population.

Clinical management

Treatment goals for patients who have demonstrated suicidal behavior may include decreasing the occurrence of suicidal thoughts, plans, gestures, or attempts. At a population level, accepted management strategies include:

• psychotherapy (cognitive-behavioral therapy [CBT], dialectical behavioral therapy)
  
• contracts for safety (widely employed but lacking evidence of efficacy)
  
• medications that target underlying disorders (antidepressants, mood stabilizers, antipsychotics).
  

Acutely, benzodiazepines and even anti psychotics may play a role in calming patients who pose a danger to themselves or caregivers.¹⁵ Presenting symptoms can suggest appropriate pharmacologic treatment strategies ( Table 2 ).^16,17

Ineffective interventions? A study examining suicide trends since 1990 in the United States¹⁸ found disheartening evidence that although treatment dramatically increased, the incidence of suicidal thoughts, plans, gestures, or attempts did not significantly decrease ( Box 1 ).^18–26 Based on a systematic review of 15 randomized controlled trials, Arensman et al¹⁹ offered 2 explanations for why studies of various psychosocial and pharmacologic interventions showed no significant effect on suicidality compared with usual care:

• the intervention had a negligible effect on patient outcomes
  
• the sample size was too small to detect clinically important differences in reattempt rates.
  

We feel a third factor is at work and is all too often forgotten: suicide research operates at a population level, whereas suicidal phenomena are inherently individual.

 

Suicide research deploys a single intervention for a diverse group of subjects rather than tailoring the approach to each particular case. A certain intervention may be highly effective for 1 patient because it is well matched to the specific blend of issues driving that patient’s suicidality, yet ineffective for another because it fails to address that individual’s underlying issues. Thus, a single treatment program standardized for research can be simultaneously a success and a failure, depending on which patient is assessed. The overall outcome is statistical insignificance because success is lost in the noise of failure.

Treating the individual. To individualize your treatment approach, it may be useful to recast the case and treatment strategy into Shneidman’s cubic model.⁶ Identifying the uniquely personal drivers behind a patient’s thoughts and actions helps point toward the most effective management approach. Tailored pharmacologic treatments and psychotherapy can be used to help guide the patient away from maximum suicide risk.

A recent study by Brown et al²⁷ found a significantly lower reattempt rate and less severe self-reported depression and hopelessness in patients who received CBT for 18 months after a suicide attempt, compared with controls. Another recent trial found that SA patients who received 18 months of partial hospitalization, mentalization-based therapy, and 18 months of follow-up group therapy had a dramatically lower suicide completion rate at 5 years compared with patients who received treatment as usual.²⁸

Table 2

Symptom-targeted pharmacologic treatment of suicidal patients

Drug class	Impulsive-behavioral dyscontrol	Affective dysregulation	Psychotic features
SSRIs	Self-damaging behavior, impulsivity	Mood lability/mood crashes; anger; temper outbursts
Antipsychotics		Anger, temper outbursts	Cognitive symptoms; perceptual symptoms
Mood stabilizers (lithium, carbamazepine, valproic acid)	Self-damaging behavior, impulsivity	Mood lability/mood crashes, anger, temper outbursts
SSRI: selective serotonin reuptake inhibitor
Source: References 16,17

Box 1

Outcomes of self-harm

When considering outcomes of SA, it is important to separate the short-term outcome of a single SA from the long-term outcome of suicidality. Short-term outcome depends on the characteristics and management of the acute episode, whereas long-term encompasses ongoing management of suicidality as a trait.

In the short term, surviving a SA depends heavily on the lethality of method and access to acute treatment. It also depends on medical fitness to withstand injury, which may help account for the higher death rate among elderly suicide attempters. A frail or medically ill person is less likely to survive the bodily insult of a SA.

Long-term outcomes are harder to predict. Some patients’ index attempts result from a transient state—an isolated incident that never will be repeated. In others, suicidality is a trait—a chronic maladaptive pattern that is potentially lethal. After an index attempt, the most reliable predictors for eventual death by suicide are:

• diagnosed mental illness
  
• high-lethality method on the index SA
  
• number of reattempts.⁴
  

As time since the index attempt increases, the risk of repeat self-harm and of suicide completion both decrease.²⁹ This raises the tantalizing prospect that if patients can be effectively bridged across the first months and years after the index attempt, they may be more likely to survive their suicidality.

Mood disorders impact long-term outcome, yet only a limited number of studies have found a reduction in suicide rates in response to mood disorder treatment. In a 44-year follow-up study, long-term treatment of depression and bipolar disorder with lithium significantly reduced the suicide rate.³⁰ A meta-analysis of recurrent major affective disorder studies found that subjects on lithium maintenance treatment were 15 times less likely to commit suicidal acts, compared with those not on lithium.³¹

 

An important confounding factor in these findings is that effective lithium treatment requires long-term adherence, which implies a long-term doctor-patient relationship. As Cipriani et al³² noted, patients who can maintain an ongoing therapeutic relationship may be “less disturbed” than those who cannot, making them less likely to kill themselves regardless of pharmacologic treatment. Furthermore, patient interviews reveal that the therapeutic alliance created by a continuous relationship can be a protective support against further SA.³³

Clinical implications

Suicide survivors often continue to struggle with suicidality well beyond the index attempt. This suicidality is a maladaptive problem-solving method that functions as a chronic morbid illness. As such, it is not enough to analyze the phenomenon of surviving an SA; one must examine the ongoing process of surviving suicidality.

Consider 3 factors. Consider all 3 factors— motive, method, and management—when addressing suicide survivorship.

Motive does not cleanly predict completion of a given SA, but it can help predict the likelihood of reattempt, which in turn affects the likelihood of eventual completion. Motive also serves as a clinical tool because neutralizing the driving factors behind an individual’s suicide attempt can solve the acute problem and help avert a reattempt.

Method lethality significantly influences survival likelihood. In clinical practice, we have observed that the index attempt is a learning experience for some patients that will inform their choice of method on the next attempt. When interacting with a suicide survivor, carefully assess the reasoning behind their initial choice of method and whether it has evolved toward higher lethality since the index attempt.

Management recommendations after SA continue to evolve. Risk factor management—such as treating underlying mood disorders, home visits to reduce social isolation, and prioritized “green card” contact with psychiatrists—has been shown to decrease reattempt rates, but many other interventions have not shown the expected benefit. Increased intervention rates have not yielded proportional decreases in suicidal ideation, attempts, or completion.

Common themes among effective management techniques appear to be keeping lines of communication open and individualizing treatment. ( Box 2 ). Our group is studying SA patients in Olmsted County, Minnesota, over the past 40 years, seeking connections between motive, method, management, and outcome. We hope trends emerging from this data will inform clinical strategies tailored to subgroups of this patient population.

Box 2

Related resources

• American Association of Suicidology. www.suicidology.org.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• Mayo Clinic Patient/Family Education. Suicide: What to do when someone is suicidal. www.mayoclinic.com/health/suicide/MH00058.
  

Drug brand names

• Carbamazepine • Carbatrol
  
• Lithium • Eskalith, Lithobid
  
• Valproic acid • Depakene, Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Comment on this article

Completed suicide provokes a multitude of questions: What motivated it? What interventions could have diverted it? Could anyone or anything have prevented it? The question of who dies by suicide often overshadows the question of what lessons suicide attempt (SA) survivors can teach us. Their story does not end with the attempt episode. For these patients, we have ongoing opportunities for interventions to make a difference.

A history of SA strongly predicts eventual completion, so we must try to identify which survivors will reattempt and complete suicide. This article addresses what is known about the psychiatry of suicide survivors—suicide motives and methods, clinical management, and short- and long-term outcomes—from the perspective that suicidality in this population may be a trait, with SA or deliberate self-harm (DSH) as its state-driven manifestations. When viewed in this manner, it is not just a question of who survives a suicide attempt, but who survives suicidality.

CASE REPORT: End of the game

Ms. T, age 39, was admitted to the intensive care unit after an aspirin overdose. She had been living with a man in a southern state for 8 years since the demise of her first marriage, but kept deferring remarriage. She returned to Minnesota with her teenage daughter to visit her family and stayed 6 months. Her partner phoned Ms. T every day, telling her he wanted her to come back. One day he tired of the game and said, “Fine, don’t come back.” She immediately overdosed, then called him to tell him what she’d done. He called her daughter, telling her to go check on her mother and to call 911. When later asked why she did it, Ms. T said, “So he would know how much he loved me.”

Motive for self-harm

Ms. T’s suicide attempt was nonlethal, and she reported it immediately—characteristics of parasuicidal gesturing as a motive. A useful categorization of suicidal behavior divides it into discrete categories or narratives. Gardner and Cowdry describe 4: true suicidal acts, parasuicidal gesturing, self-mutilation, and retributive rage.¹ We modify this schema with 4 additional categories: altruism, acute shame, command hallucinations, and panic ( Table 1 ).^1-3 Categories are differentiated by affective state, motivation, and goal of behavior, but all involve situations in which the individual feels a lack of other options and resorts to maladaptive strategies.

Although this classification scheme helps clinicians understand a patient’s mindset, the specific motive underpinning DSH or SA is not consistently linked to its lethality. True suicidal acts frequently are marked by careful planning and high-lethality methods that increase the risk of completed suicide, but any motive can lead to a lethal act, whether or not death was intended.^2,3

Factors that increase the risk of SA and completed suicide include male gender, age (adolescent or age >60), low socioeconomic status, and alcohol or drug abuse.⁴ An underlying mood disorder accounts for 73% of attributable risk of suicide or medically serious SA in older adults.⁵ This connection between mood and suicidality highlights the concept that emotional pain can cause so much suffering that patients seek release from distress by ending their lives.

A useful model by Shneidman⁶ casts psychological pain as 1 dimension in a 3-dimensional system that includes press and perturbation. In this model:

• Pain refers to psychological pain (from little or no pain to intolerable agony).
  
• Press means actual or imagined events in the inner or outer world that cause a person to react. It ranges from positive press (good fortune, happy events, protective factors) to negative press (stressors, failures, losses, persecution), which in turn decrease or increase the likelihood of suicide.
  
• Perturbation refers to the state of being disturbed or upset.
  

Disruption in any 1 dimension tends to disturb the other 2 dimensions. When all 3 dimensions reach maximum distress, the stage is optimally set for suicide. DSM-IV diagnoses are examples of manifestations of high levels of perturbation. Perturbation is especially conducive to suicide when it involves constriction (a sense that the only viable option is death) and a penchant for self-harmful action.

Certain risk factors make SA simultaneously more likely to occur but less likely to be lethal. For example, parental discord, nonheterosexual orientation, and female gender have been found to increase non-fatal attempts among adolescents.⁷ Borderline personality disorder increases the reattempt rate out of proportion to completion among adults.⁸ One might interpret a pattern of repeated nonlethal attempts to mean the patient has no real intent to die, but this is not always the case.

 

⁸

Table 1

8 categories or narratives of suicidal behavior

Motive	Characteristics
True suicidal act	Release from intense baseline despair/hopelessness; self-nihilism as a permanent end to internal pain (entails highest intent to die and highest risk of completed suicide)
Self-mutilation	Relieving dysphoria or dissociation/depersonalization; acts of DSH designed to self-regulate or distract from emotional pain or other overwhelming affects
Retributive rage	Revenge; impulsiveness, vengefulness, and reduced capacity to conceive of other immediate options
Parasuicidal gesturing	Communication designed to extract a response from a significant other; often repetitive acts of DSH, strong dependency needs
Acute shame	Penance designed to escape from or to atone for a shameful act; often occurs within a short time after act is committed
Altruism	Relief of real or imagined burden on others; often occurs in setting of medical illness or substantial financial concerns
Command hallucinations	Acting in compliance with a command hallucination; often in setting of schizophrenia or depression with psychotic features
Panic	Driven by agitation, psychic anxiety, and/or panic attack; action intended as escape from real or imagined factor provoking agitation
DSH: deliberate self-harm
Source: References 1-3

CASE REPORT: Caught in the act

Mrs. L, age 35, works at a nail salon and took $12 from the cash register to buy gas so she could visit her husband in the next town. She’d never done anything like that before. She planned to return the money the next day, but her act was captured by a security camera and reported before she had a chance. Her boss said she had to go to the police.

Mrs. L was so ashamed that she decided she wanted to die. She drove her car to a remote hunting area where she tried to shoot herself in the head. The gun bucked, however, and shot her in the shoulder instead. She climbed into the front seat and drove herself to the hospital.

Method of self-harm

Survival of a suicide attempt depends in part on the lethality of the suicide method. Although she survived, Mrs. L’s attempt was intended to be quite lethal and illustrates shame as a motive.

The method’s lethality does not always correlate with the intent to die.⁹ Attempters with the highest suicidal intent do not reliably choose the most lethal method, either because they overestimate the lethality of methods such as cutting or overdose or because less lethal methods were most accessible.

Physicians have a higher suicide rate than the general population,¹⁰ probably because of their knowledge of lethal means. Patients with greater access to information on method lethality may be more likely to match their method with their intent, thus raising the risk of suicide completion.¹¹ Patients who use a high-lethality method on the index SA tend to continue using high-lethality methods on reattempts, which makes eventual fatality likely.¹²

Firearms, which are both accessible and lethal, remain the most common and deadly method in the United States, with more suicides from gunshot than all other methods combined.¹³ Cultural factors also are involved, such as in India where poisoning (especially with readily available organophosphates) is more common than gunshot.¹⁴ Suicidality screening in psychiatric practice and in the emergency department should always include questioning about convenient access to lethal means, especially those commonly used among the local population.

Clinical management

Treatment goals for patients who have demonstrated suicidal behavior may include decreasing the occurrence of suicidal thoughts, plans, gestures, or attempts. At a population level, accepted management strategies include:

• psychotherapy (cognitive-behavioral therapy [CBT], dialectical behavioral therapy)
  
• contracts for safety (widely employed but lacking evidence of efficacy)
  
• medications that target underlying disorders (antidepressants, mood stabilizers, antipsychotics).
  

Acutely, benzodiazepines and even anti psychotics may play a role in calming patients who pose a danger to themselves or caregivers.¹⁵ Presenting symptoms can suggest appropriate pharmacologic treatment strategies ( Table 2 ).^16,17

Ineffective interventions? A study examining suicide trends since 1990 in the United States¹⁸ found disheartening evidence that although treatment dramatically increased, the incidence of suicidal thoughts, plans, gestures, or attempts did not significantly decrease ( Box 1 ).^18–26 Based on a systematic review of 15 randomized controlled trials, Arensman et al¹⁹ offered 2 explanations for why studies of various psychosocial and pharmacologic interventions showed no significant effect on suicidality compared with usual care:

• the intervention had a negligible effect on patient outcomes
  
• the sample size was too small to detect clinically important differences in reattempt rates.
  

We feel a third factor is at work and is all too often forgotten: suicide research operates at a population level, whereas suicidal phenomena are inherently individual.

 

Suicide research deploys a single intervention for a diverse group of subjects rather than tailoring the approach to each particular case. A certain intervention may be highly effective for 1 patient because it is well matched to the specific blend of issues driving that patient’s suicidality, yet ineffective for another because it fails to address that individual’s underlying issues. Thus, a single treatment program standardized for research can be simultaneously a success and a failure, depending on which patient is assessed. The overall outcome is statistical insignificance because success is lost in the noise of failure.

Treating the individual. To individualize your treatment approach, it may be useful to recast the case and treatment strategy into Shneidman’s cubic model.⁶ Identifying the uniquely personal drivers behind a patient’s thoughts and actions helps point toward the most effective management approach. Tailored pharmacologic treatments and psychotherapy can be used to help guide the patient away from maximum suicide risk.

A recent study by Brown et al²⁷ found a significantly lower reattempt rate and less severe self-reported depression and hopelessness in patients who received CBT for 18 months after a suicide attempt, compared with controls. Another recent trial found that SA patients who received 18 months of partial hospitalization, mentalization-based therapy, and 18 months of follow-up group therapy had a dramatically lower suicide completion rate at 5 years compared with patients who received treatment as usual.²⁸

Table 2

Symptom-targeted pharmacologic treatment of suicidal patients

Drug class	Impulsive-behavioral dyscontrol	Affective dysregulation	Psychotic features
SSRIs	Self-damaging behavior, impulsivity	Mood lability/mood crashes; anger; temper outbursts
Antipsychotics		Anger, temper outbursts	Cognitive symptoms; perceptual symptoms
Mood stabilizers (lithium, carbamazepine, valproic acid)	Self-damaging behavior, impulsivity	Mood lability/mood crashes, anger, temper outbursts
SSRI: selective serotonin reuptake inhibitor
Source: References 16,17

Box 1

Outcomes of self-harm

When considering outcomes of SA, it is important to separate the short-term outcome of a single SA from the long-term outcome of suicidality. Short-term outcome depends on the characteristics and management of the acute episode, whereas long-term encompasses ongoing management of suicidality as a trait.

In the short term, surviving a SA depends heavily on the lethality of method and access to acute treatment. It also depends on medical fitness to withstand injury, which may help account for the higher death rate among elderly suicide attempters. A frail or medically ill person is less likely to survive the bodily insult of a SA.

Long-term outcomes are harder to predict. Some patients’ index attempts result from a transient state—an isolated incident that never will be repeated. In others, suicidality is a trait—a chronic maladaptive pattern that is potentially lethal. After an index attempt, the most reliable predictors for eventual death by suicide are:

• diagnosed mental illness
  
• high-lethality method on the index SA
  
• number of reattempts.⁴
  

As time since the index attempt increases, the risk of repeat self-harm and of suicide completion both decrease.²⁹ This raises the tantalizing prospect that if patients can be effectively bridged across the first months and years after the index attempt, they may be more likely to survive their suicidality.

Mood disorders impact long-term outcome, yet only a limited number of studies have found a reduction in suicide rates in response to mood disorder treatment. In a 44-year follow-up study, long-term treatment of depression and bipolar disorder with lithium significantly reduced the suicide rate.³⁰ A meta-analysis of recurrent major affective disorder studies found that subjects on lithium maintenance treatment were 15 times less likely to commit suicidal acts, compared with those not on lithium.³¹

 

An important confounding factor in these findings is that effective lithium treatment requires long-term adherence, which implies a long-term doctor-patient relationship. As Cipriani et al³² noted, patients who can maintain an ongoing therapeutic relationship may be “less disturbed” than those who cannot, making them less likely to kill themselves regardless of pharmacologic treatment. Furthermore, patient interviews reveal that the therapeutic alliance created by a continuous relationship can be a protective support against further SA.³³

Clinical implications

Suicide survivors often continue to struggle with suicidality well beyond the index attempt. This suicidality is a maladaptive problem-solving method that functions as a chronic morbid illness. As such, it is not enough to analyze the phenomenon of surviving an SA; one must examine the ongoing process of surviving suicidality.

Consider 3 factors. Consider all 3 factors— motive, method, and management—when addressing suicide survivorship.

Motive does not cleanly predict completion of a given SA, but it can help predict the likelihood of reattempt, which in turn affects the likelihood of eventual completion. Motive also serves as a clinical tool because neutralizing the driving factors behind an individual’s suicide attempt can solve the acute problem and help avert a reattempt.

Method lethality significantly influences survival likelihood. In clinical practice, we have observed that the index attempt is a learning experience for some patients that will inform their choice of method on the next attempt. When interacting with a suicide survivor, carefully assess the reasoning behind their initial choice of method and whether it has evolved toward higher lethality since the index attempt.

Management recommendations after SA continue to evolve. Risk factor management—such as treating underlying mood disorders, home visits to reduce social isolation, and prioritized “green card” contact with psychiatrists—has been shown to decrease reattempt rates, but many other interventions have not shown the expected benefit. Increased intervention rates have not yielded proportional decreases in suicidal ideation, attempts, or completion.

Common themes among effective management techniques appear to be keeping lines of communication open and individualizing treatment. ( Box 2 ). Our group is studying SA patients in Olmsted County, Minnesota, over the past 40 years, seeking connections between motive, method, management, and outcome. We hope trends emerging from this data will inform clinical strategies tailored to subgroups of this patient population.

Box 2

Related resources

• American Association of Suicidology. www.suicidology.org.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• Mayo Clinic Patient/Family Education. Suicide: What to do when someone is suicidal. www.mayoclinic.com/health/suicide/MH00058.
  

Drug brand names

• Carbamazepine • Carbatrol
  
• Lithium • Eskalith, Lithobid
  
• Valproic acid • Depakene, Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Comment on this article

Completed suicide provokes a multitude of questions: What motivated it? What interventions could have diverted it? Could anyone or anything have prevented it? The question of who dies by suicide often overshadows the question of what lessons suicide attempt (SA) survivors can teach us. Their story does not end with the attempt episode. For these patients, we have ongoing opportunities for interventions to make a difference.

A history of SA strongly predicts eventual completion, so we must try to identify which survivors will reattempt and complete suicide. This article addresses what is known about the psychiatry of suicide survivors—suicide motives and methods, clinical management, and short- and long-term outcomes—from the perspective that suicidality in this population may be a trait, with SA or deliberate self-harm (DSH) as its state-driven manifestations. When viewed in this manner, it is not just a question of who survives a suicide attempt, but who survives suicidality.

CASE REPORT: End of the game

Ms. T, age 39, was admitted to the intensive care unit after an aspirin overdose. She had been living with a man in a southern state for 8 years since the demise of her first marriage, but kept deferring remarriage. She returned to Minnesota with her teenage daughter to visit her family and stayed 6 months. Her partner phoned Ms. T every day, telling her he wanted her to come back. One day he tired of the game and said, “Fine, don’t come back.” She immediately overdosed, then called him to tell him what she’d done. He called her daughter, telling her to go check on her mother and to call 911. When later asked why she did it, Ms. T said, “So he would know how much he loved me.”

Motive for self-harm

Ms. T’s suicide attempt was nonlethal, and she reported it immediately—characteristics of parasuicidal gesturing as a motive. A useful categorization of suicidal behavior divides it into discrete categories or narratives. Gardner and Cowdry describe 4: true suicidal acts, parasuicidal gesturing, self-mutilation, and retributive rage.¹ We modify this schema with 4 additional categories: altruism, acute shame, command hallucinations, and panic ( Table 1 ).^1-3 Categories are differentiated by affective state, motivation, and goal of behavior, but all involve situations in which the individual feels a lack of other options and resorts to maladaptive strategies.

Although this classification scheme helps clinicians understand a patient’s mindset, the specific motive underpinning DSH or SA is not consistently linked to its lethality. True suicidal acts frequently are marked by careful planning and high-lethality methods that increase the risk of completed suicide, but any motive can lead to a lethal act, whether or not death was intended.^2,3

Factors that increase the risk of SA and completed suicide include male gender, age (adolescent or age >60), low socioeconomic status, and alcohol or drug abuse.⁴ An underlying mood disorder accounts for 73% of attributable risk of suicide or medically serious SA in older adults.⁵ This connection between mood and suicidality highlights the concept that emotional pain can cause so much suffering that patients seek release from distress by ending their lives.

A useful model by Shneidman⁶ casts psychological pain as 1 dimension in a 3-dimensional system that includes press and perturbation. In this model:

• Pain refers to psychological pain (from little or no pain to intolerable agony).
  
• Press means actual or imagined events in the inner or outer world that cause a person to react. It ranges from positive press (good fortune, happy events, protective factors) to negative press (stressors, failures, losses, persecution), which in turn decrease or increase the likelihood of suicide.
  
• Perturbation refers to the state of being disturbed or upset.
  

Disruption in any 1 dimension tends to disturb the other 2 dimensions. When all 3 dimensions reach maximum distress, the stage is optimally set for suicide. DSM-IV diagnoses are examples of manifestations of high levels of perturbation. Perturbation is especially conducive to suicide when it involves constriction (a sense that the only viable option is death) and a penchant for self-harmful action.

Certain risk factors make SA simultaneously more likely to occur but less likely to be lethal. For example, parental discord, nonheterosexual orientation, and female gender have been found to increase non-fatal attempts among adolescents.⁷ Borderline personality disorder increases the reattempt rate out of proportion to completion among adults.⁸ One might interpret a pattern of repeated nonlethal attempts to mean the patient has no real intent to die, but this is not always the case.

 

⁸

Table 1

8 categories or narratives of suicidal behavior

Motive	Characteristics
True suicidal act	Release from intense baseline despair/hopelessness; self-nihilism as a permanent end to internal pain (entails highest intent to die and highest risk of completed suicide)
Self-mutilation	Relieving dysphoria or dissociation/depersonalization; acts of DSH designed to self-regulate or distract from emotional pain or other overwhelming affects
Retributive rage	Revenge; impulsiveness, vengefulness, and reduced capacity to conceive of other immediate options
Parasuicidal gesturing	Communication designed to extract a response from a significant other; often repetitive acts of DSH, strong dependency needs
Acute shame	Penance designed to escape from or to atone for a shameful act; often occurs within a short time after act is committed
Altruism	Relief of real or imagined burden on others; often occurs in setting of medical illness or substantial financial concerns
Command hallucinations	Acting in compliance with a command hallucination; often in setting of schizophrenia or depression with psychotic features
Panic	Driven by agitation, psychic anxiety, and/or panic attack; action intended as escape from real or imagined factor provoking agitation
DSH: deliberate self-harm
Source: References 1-3

CASE REPORT: Caught in the act

Mrs. L, age 35, works at a nail salon and took $12 from the cash register to buy gas so she could visit her husband in the next town. She’d never done anything like that before. She planned to return the money the next day, but her act was captured by a security camera and reported before she had a chance. Her boss said she had to go to the police.

Mrs. L was so ashamed that she decided she wanted to die. She drove her car to a remote hunting area where she tried to shoot herself in the head. The gun bucked, however, and shot her in the shoulder instead. She climbed into the front seat and drove herself to the hospital.

Method of self-harm

Survival of a suicide attempt depends in part on the lethality of the suicide method. Although she survived, Mrs. L’s attempt was intended to be quite lethal and illustrates shame as a motive.

The method’s lethality does not always correlate with the intent to die.⁹ Attempters with the highest suicidal intent do not reliably choose the most lethal method, either because they overestimate the lethality of methods such as cutting or overdose or because less lethal methods were most accessible.

Physicians have a higher suicide rate than the general population,¹⁰ probably because of their knowledge of lethal means. Patients with greater access to information on method lethality may be more likely to match their method with their intent, thus raising the risk of suicide completion.¹¹ Patients who use a high-lethality method on the index SA tend to continue using high-lethality methods on reattempts, which makes eventual fatality likely.¹²

Firearms, which are both accessible and lethal, remain the most common and deadly method in the United States, with more suicides from gunshot than all other methods combined.¹³ Cultural factors also are involved, such as in India where poisoning (especially with readily available organophosphates) is more common than gunshot.¹⁴ Suicidality screening in psychiatric practice and in the emergency department should always include questioning about convenient access to lethal means, especially those commonly used among the local population.

Clinical management

Treatment goals for patients who have demonstrated suicidal behavior may include decreasing the occurrence of suicidal thoughts, plans, gestures, or attempts. At a population level, accepted management strategies include:

• psychotherapy (cognitive-behavioral therapy [CBT], dialectical behavioral therapy)
  
• contracts for safety (widely employed but lacking evidence of efficacy)
  
• medications that target underlying disorders (antidepressants, mood stabilizers, antipsychotics).
  

Acutely, benzodiazepines and even anti psychotics may play a role in calming patients who pose a danger to themselves or caregivers.¹⁵ Presenting symptoms can suggest appropriate pharmacologic treatment strategies ( Table 2 ).^16,17

Ineffective interventions? A study examining suicide trends since 1990 in the United States¹⁸ found disheartening evidence that although treatment dramatically increased, the incidence of suicidal thoughts, plans, gestures, or attempts did not significantly decrease ( Box 1 ).^18–26 Based on a systematic review of 15 randomized controlled trials, Arensman et al¹⁹ offered 2 explanations for why studies of various psychosocial and pharmacologic interventions showed no significant effect on suicidality compared with usual care:

• the intervention had a negligible effect on patient outcomes
  
• the sample size was too small to detect clinically important differences in reattempt rates.
  

We feel a third factor is at work and is all too often forgotten: suicide research operates at a population level, whereas suicidal phenomena are inherently individual.

 

Suicide research deploys a single intervention for a diverse group of subjects rather than tailoring the approach to each particular case. A certain intervention may be highly effective for 1 patient because it is well matched to the specific blend of issues driving that patient’s suicidality, yet ineffective for another because it fails to address that individual’s underlying issues. Thus, a single treatment program standardized for research can be simultaneously a success and a failure, depending on which patient is assessed. The overall outcome is statistical insignificance because success is lost in the noise of failure.

Treating the individual. To individualize your treatment approach, it may be useful to recast the case and treatment strategy into Shneidman’s cubic model.⁶ Identifying the uniquely personal drivers behind a patient’s thoughts and actions helps point toward the most effective management approach. Tailored pharmacologic treatments and psychotherapy can be used to help guide the patient away from maximum suicide risk.

A recent study by Brown et al²⁷ found a significantly lower reattempt rate and less severe self-reported depression and hopelessness in patients who received CBT for 18 months after a suicide attempt, compared with controls. Another recent trial found that SA patients who received 18 months of partial hospitalization, mentalization-based therapy, and 18 months of follow-up group therapy had a dramatically lower suicide completion rate at 5 years compared with patients who received treatment as usual.²⁸

Table 2

Symptom-targeted pharmacologic treatment of suicidal patients

Drug class	Impulsive-behavioral dyscontrol	Affective dysregulation	Psychotic features
SSRIs	Self-damaging behavior, impulsivity	Mood lability/mood crashes; anger; temper outbursts
Antipsychotics		Anger, temper outbursts	Cognitive symptoms; perceptual symptoms
Mood stabilizers (lithium, carbamazepine, valproic acid)	Self-damaging behavior, impulsivity	Mood lability/mood crashes, anger, temper outbursts
SSRI: selective serotonin reuptake inhibitor
Source: References 16,17

Box 1

Outcomes of self-harm

When considering outcomes of SA, it is important to separate the short-term outcome of a single SA from the long-term outcome of suicidality. Short-term outcome depends on the characteristics and management of the acute episode, whereas long-term encompasses ongoing management of suicidality as a trait.

In the short term, surviving a SA depends heavily on the lethality of method and access to acute treatment. It also depends on medical fitness to withstand injury, which may help account for the higher death rate among elderly suicide attempters. A frail or medically ill person is less likely to survive the bodily insult of a SA.

Long-term outcomes are harder to predict. Some patients’ index attempts result from a transient state—an isolated incident that never will be repeated. In others, suicidality is a trait—a chronic maladaptive pattern that is potentially lethal. After an index attempt, the most reliable predictors for eventual death by suicide are:

• diagnosed mental illness
  
• high-lethality method on the index SA
  
• number of reattempts.⁴
  

As time since the index attempt increases, the risk of repeat self-harm and of suicide completion both decrease.²⁹ This raises the tantalizing prospect that if patients can be effectively bridged across the first months and years after the index attempt, they may be more likely to survive their suicidality.

Mood disorders impact long-term outcome, yet only a limited number of studies have found a reduction in suicide rates in response to mood disorder treatment. In a 44-year follow-up study, long-term treatment of depression and bipolar disorder with lithium significantly reduced the suicide rate.³⁰ A meta-analysis of recurrent major affective disorder studies found that subjects on lithium maintenance treatment were 15 times less likely to commit suicidal acts, compared with those not on lithium.³¹

 

An important confounding factor in these findings is that effective lithium treatment requires long-term adherence, which implies a long-term doctor-patient relationship. As Cipriani et al³² noted, patients who can maintain an ongoing therapeutic relationship may be “less disturbed” than those who cannot, making them less likely to kill themselves regardless of pharmacologic treatment. Furthermore, patient interviews reveal that the therapeutic alliance created by a continuous relationship can be a protective support against further SA.³³

Clinical implications

Suicide survivors often continue to struggle with suicidality well beyond the index attempt. This suicidality is a maladaptive problem-solving method that functions as a chronic morbid illness. As such, it is not enough to analyze the phenomenon of surviving an SA; one must examine the ongoing process of surviving suicidality.

Consider 3 factors. Consider all 3 factors— motive, method, and management—when addressing suicide survivorship.

Motive does not cleanly predict completion of a given SA, but it can help predict the likelihood of reattempt, which in turn affects the likelihood of eventual completion. Motive also serves as a clinical tool because neutralizing the driving factors behind an individual’s suicide attempt can solve the acute problem and help avert a reattempt.

Method lethality significantly influences survival likelihood. In clinical practice, we have observed that the index attempt is a learning experience for some patients that will inform their choice of method on the next attempt. When interacting with a suicide survivor, carefully assess the reasoning behind their initial choice of method and whether it has evolved toward higher lethality since the index attempt.

Management recommendations after SA continue to evolve. Risk factor management—such as treating underlying mood disorders, home visits to reduce social isolation, and prioritized “green card” contact with psychiatrists—has been shown to decrease reattempt rates, but many other interventions have not shown the expected benefit. Increased intervention rates have not yielded proportional decreases in suicidal ideation, attempts, or completion.

Common themes among effective management techniques appear to be keeping lines of communication open and individualizing treatment. ( Box 2 ). Our group is studying SA patients in Olmsted County, Minnesota, over the past 40 years, seeking connections between motive, method, management, and outcome. We hope trends emerging from this data will inform clinical strategies tailored to subgroups of this patient population.

Box 2

Related resources

• American Association of Suicidology. www.suicidology.org.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• Mayo Clinic Patient/Family Education. Suicide: What to do when someone is suicidal. www.mayoclinic.com/health/suicide/MH00058.
  

Drug brand names

• Carbamazepine • Carbatrol
  
• Lithium • Eskalith, Lithobid
  
• Valproic acid • Depakene, Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.