Revisiting Impulsivity in Suicide
Implications for Civil Liability of Third Parties
April R. Smith, M.S.,1 Tracy K. Witte, M.S.,1 Nadia E. Teale, M.S.,1 Sarah L. King, J.D.,2 Ted W. Bender, M.S.,1 and Thomas E. Joiner, Ph.D.1
Author information ► Copyright and License information ►
The publisher's final edited version of this article is available at Behav Sci Law
See other articles in PMC that cite the published article.
Go to:
Abstract
Previous research and popular conceptualizations of suicide have posited that many suicides are the result of impulsive, “on a whim” decisions. However, recent research demonstrates that most suicides are not attempted impulsively, and in fact involve a plan. Legally, suicide has historically been considered to be a superseding intervening cause of death that exonerates other parties from liability, but currently there are two general exceptions to this view. Specifically, another party may be found responsible for a suicide if that party either caused the suicide or failed in its duty to prevent the suicide from occurring. Both of these exceptions assume that the resulting suicide was foreseeable. Given that recent research has indicated that most suicides are planned, and thereby foreseeable to a certain extent under many circumstances, this article discusses issues of foreseeability as they pertain to litigation involving third party liability for the suicide of university students, prison inmates, and mental health patients. The authors contend that the surest way for universities, prison staff, and mental health practitioners to avoid being held liable for a suicide is to appropriately assess for suicidal intent.
Suicide is the third leading cause of death for people aged 15 to 24 and the 11th leading cause of death overall (Centers for Disease Control [CDC], 2004). Approximately 30,000 people die by suicide every year in the U.S. Moreover, four times that many people are hospitalized due to injuries sustained while attempting suicide (CDC, 2004). For the purposes of this paper we will use Silverman and colleagues’ (Silverman, Berman, Sanddal, O’Carroll, & Joiner, 2007) definition of a “suicide attempt,” which is “a self-inflicted, potentially injurious behavior with a nonfatal outcome for which there is evidence (either explicit or implicit) of intent to die.” Suicide is thus a serious mental health issue and public health problem. In the wake of a suicide, loved ones are often left looking for answers. When family members or friends feel something more should have been done for the decedent, they may ultimately try to settle things in a court of law. In fact, the most common type of lawsuit brought against psychiatrists is in regard to patient suicide (Simon, 2002). Furthermore, records from the American Psychological Association’s Insurance Trust indicate that of claims closed by 1990, patient suicide was the second most costly claim for psychiatrists and psychologists insured by the company (Bongar, Maris, Berman, & Litman, 1998). As our society has become more litigious, the number of malpractice suits filed against therapists has risen dramatically and will likely continue to rise (Berman, 1986).
Although the causes of suicide are complex, they are not mysterious, and in fact are becoming better understood thanks to decades of scientific research. One comprehensive theory of suicide is Joiner’s (2005) interpersonal-psychological theory. Importantly, at least 20 empirical studies on this theory have been conducted, and all were supportive (Van Orden, et al., 2008). According to this theory three proximal, jointly necessary, and sufficient causes must be present before a person will die by suicide; these are: 1) feelings of perceived burdensomeness, 2) a sense of thwarted belongingness, and 3) an acquired capability to lethally self-harm. Perceived burdensomeness occurs when a person believes his/her death is worth more than his/her life to others. In essence, a person experiencing burdensomeness feels that others would be better off if s/he were dead. Thwarted belongingness results when one of the basic human needs, to be connected to others (Baumeister & Leary, 1995), is not met. Both perceived burdensomeness and thwarted belongingness are theorized to contribute to the desire for suicide (c.f., suicidal ideation), and elevated levels of both perceived burdensomeness and thwarted belongingness have been found to significantly predict suicidal desire (Van Orden, Witte, Gordon, Bender, & Joiner, 2008).
The third necessary condition for death by suicide is the acquired capability to lethally self-harm (subsequently referred to as the acquired capability for suicide). Engaging in self-harm is something that usually involves great pain and is fear-inducing to most people. In order to overcome this ingrained fear, it is necessary to habituate in some way to stimuli associated with self-injury. With respect to self-injury, opponent process theory (Solomon, 1980; Solomon and Corbit, 1974) holds that exposure to painful stimuli eventually engages an opposite reaction, one likely to induce analgesia and calm. With repetition, the pain associated with these experiences decreases, and the more reinforcing aspects increase. Thus, Joiner’s (2005) theory proposes that the acquired capability for suicide develops after one has been repeatedly exposed to painful and/or provocative stimuli. A recent study has shown that these painful and/or provocative stimuli include but are not limited to past suicide attempts, self-injecting drug use, non-suicidal self-injury, and exposure to physical violence (Van Orden, et al., 2008). This same study also found evidence for the distinctiveness of the acquired capability for suicide from suicidal ideation. That is, one may have acquired the capability for suicide, but have no suicidal desire. Conversely, one may have suicidal desire, but lack the capability to die by suicide. The relative rarity of the intersection of these factors is consistent with the relative rarity of death by suicide (Joiner, 2005; Van Orden, et al., 2008).
Impulsivity has also been associated with death by suicide and is one of the most frequently implicated risk factors for engaging in maladaptive behaviors, such as serious self-injury (Anestis, Selby, & Joiner, 2007). However, recent research has shown that although people who attempt suicide tend to be more impulsive than those who do not, the actual act of suicide is generally not done impulsively. Despite these findings, there unfortunately continues to be a widespread misconception that the majority of suicides, particularly in adolescents, are impulsive in the moment (e.g., Carey, 2008). This misconception understandably leads to fear and anxiety on the part of mental health providers who treat suicidal clients. Moreover, this fear is heightened by the danger of being sued should one’s client die by suicide. However, the fact that the vast majority of suicides are not “impulsive,” but rather usually the result of extensive planning, has significant implications for the predictability of suicide and the related issue of third parties’ legal responsibilities to detect and protect against the risk that others could die by suicide. Thus, the main aim of this article is to explain and clarify the role of impulsivity in suicide, and discuss its potential legal implications for theories of third party liability.
Go to:
ROLE OF IMPULSIVITY IN SUICIDE
As mentioned, much of the previous literature on suicidality has proposed a direct link between impulsivity and suicide. For example, Baumeister (1990) conceptualized suicide as a means of escape from aversive self-awareness. He proposed that when individuals experience aversive self-awareness they attempt to alleviate these feelings by achieving a state of “cognitive deconstruction,” which is characterized by a constricted, present focused time perspective and cognitive rigidity. This deconstructed state dulls self-awareness and leads to disinhibition. While disinhibited, individuals are less able to resist suicidal impulses and thus may end up attempting suicide impulsively. Other researchers (e.g., Mann, Waternaux, Haas, & Malone, 1999) have suggested a diathesis-stress model of suicide, in which the diathesis (or vulnerability) for suicide includes a tendency to act impulsively and experience more suicidal ideation. Thus, in the face of a stressor (i.e., a mental illness), an individual with this diathesis may impulsively attempt suicide.
There is clear support for some role for impulsivity in suicidal behavior. For example, research has consistently found links between impulsive personality characteristics and suicidal ideation and behavior (e.g., Conner, Meldrum, Wieczorek, Duberstein, & Welte, 2004; Hull-Blanks, Kerr, & Kurpius, 2004; Maser, et al., 2002; Pfeffer, Jiang, & Kakuma, 2000). Moreover, Dougherty and colleagues (Doughtery, Mathias, Marsh, Papageorgiou, Swann, & Moeller, 2004) found that higher levels of impulsivity were related to more previous suicide attempts. A strength of this particular study was its use of multiple measures of impulsivity; however, a limitation is the lack of explication regarding the mechanism for the relationship between suicide and impulsivity. Simon et al. (2001) reported that among 153 suicide attempters, 24% attempted impulsively. However, it is unclear whether these “impulsive” attempters were acting on a plan they had previously thought about or not; what is clear is that the majority of attempters — 76% in the Simon et al. study — did not attempt impulsively.
Go to:
ROLE OF IMPULSIVITY AND SUICIDE REVISITED
While it is clear that impulsivity is a significant risk factor for suicide, a compelling mechanism for the relationship between impulsivity and death by suicide has not been adequately documented. Furthermore, research indicates that of course not all impulsive individuals die by suicide, nor can all those who engage in suicidal behavior be characterized as impulsive. What, then, accounts for the fairly reliable association between impulsivity and suicide? The discrepancy may be due, in part, to the use of the term “impulsive.” In previous theories of suicidal behavior, state impulsivity (i.e., impulsivity experienced at a particular point in time) was thought to account for sudden, unplanned suicidal behavior. Joiner’s (2005) interpersonal-psychological theory calls the role of state impulsivity into question, and argues that impulsivity at the time of self-harm is less important than impulsivity exhibited throughout life leading up to suicidal behavior. Here, impulsivity is conceptualized as a trait or personality variable that influences involvement in risky behavior (e.g., substance use, reckless driving, fighting, etc.). According to this theory, impulsive people do not die on a whim due to state impulsivity, rather, they engage in experiences, or risky behaviors, that over time instill in them the capability to enact serious self-harm (should they develop the desire to do so).
As discussed earlier, the interpersonal-psychological theory of suicide holds that in order for people to die by suicide, they must have a diminished sense of belonging, perceive themselves to be a burden to loved ones, and have acquired the capability to enact lethal self-injury. More specifically, the theory suggests that individuals advance along a trajectory of escalating capability for self-injury by engaging in activities that foster fearlessness of and competence for suicide. The most direct path along this trajectory is deliberate self-injury or suicide attempts. However, certain painful or provocative experiences, such as those promoted by trait impulsivity, may also serve this function indirectly. In fact, research indicates that impulsive people are injured in accidents more than others, and are more prone to engage in substance use (Cherpitel, 1993). Painful and provocative experiences, like injuries and substance use, create familiarity with pain and potentially life-threatening situations, thereby creating opportunities for habituation to self-harm.
Mental disorders that have impulsivity as a core feature may confer risk for painful and provocative experiences as well. For example, a study comparing individuals with Borderline Personality Disorder (BPD) to other diagnostic groups in an outpatient community clinic, found that those with BPD were more likely to endorse a variety of painful and provocative experiences (Stellrecht, Selby, Bender, & Joiner, 2008). Endorsing more of these experiences partially mediated the relationship between BPD diagnosis and likelihood of past suicide attempt.
Consistent with Joiner’s (2005) interpretation of impulsivity’s distal role in suicide, the results of other studies indicate that impulsive suicides (i.e., dying on a whim) are vanishingly rare. For example, Baca-Garcia and colleagues (2005) simultaneously assessed attempt (i.e., state) impulsivity and attempter (i.e., trait) impulsivity in an inpatient population. They found that impulsive traits did not predict attempt impulsivity (i.e., attempting suicide without prior planning), and that non-impulsive attempts (i.e., those that involved prior planning) were more lethal compared to impulsive attempts (see also Baca-Garcia et al., 2001). These findings highlight that plans and preparations for suicide may be critically involved in lethality. Similarly, Wyder & De Leo (2007) surveyed a community sample regarding past suicidal behavior. Of those who reported a past suicide attempt, only one quarter were classified as impulsive (i.e., little planning was involved). Moreover, similar to Simon et al. (2001), these authors did not assess for prior planfulness; thus it is not clear how many “impulsive” suicides actually involved plans that had been developed weeks and months and before, were impulsively enacted. Notably, no differences were found on a measure of trait impulsivity between “impulsive” attempters and non-impulsive attempters. Additionally, “impulsive” suicide attempters were less likely to believe that their attempt would be lethal.
Lastly, Witte et al. (2008) compared three groups of suicidal adolescents in a large epidemiological study: those who planned a suicide attempt, but did not actually attempt; those who did not plan an attempt, but did attempt (i.e., “impulsive” attempters); and those who both planned for and attempted suicide. Given that planful attempts are considered a more serious form of suicidal behavior (e.g., Baca-Garcia et al., 2005), the authors predicted that the individuals who had habituated to fear and pain associated with self-injury (i.e., those who had engaged in the largest number of impulsive behaviors and had the highest trait impulsivity) would be more likely to have made a planful attempt (i.e., non-impulsive attempt). This prediction stands in contrast to the more traditional view, which would predict that impulsive people would tend to make impulsive suicide attempts. Results were consistent with the authors’ prediction and Joiner’s (2005) theory; the so-called “impulsive” suicide attempters (i.e., those who attempted suicide without prior planning) were actually less likely to engage in other impulsive behaviors than those who made a planful attempt. Furthermore, those who engaged in planned suicide attempts were more likely to have required medical attention for their injuries. Finally, less than 10% of the adolescents who had attempted suicide in their sample had done so impulsively. Again, these findings are consistent with the idea that suicide attempts “on a whim” are quite rare.
Thus it appears that impulsivity does play an important, but distal role, in suicidal behavior. Research has demonstrated that impulsive individuals are more likely to engage in painful and provocative experiences and that these experiences appear to make them less fearful about death. Given their greater acquired capability for suicide, if these individuals go on to experience perceived burdensomeness and thwarted belongingness, they will be at high risk for death by suicide, according to the interpersonal-psychological theory.
Go to:
BACKGROUND ON SUICIDE AND THE LAW
Historically, suicide has been considered an illegal act for which the decedent bore sole responsibility. The associated legal consequences included, for example, violation of the decedent’s body, harsh restrictions regarding burial, and forfeiture of property. In light of this theory of responsibility, in the wake of a suicide there was no avenue for family members to seek legal recourse against other parties. By the mid 1700s, however, there was a shift in this legal theory of responsibility. Rather than being viewed as criminals responsible for their death, individuals who died by suicide began to be seen as acting under psychological duress. Finally, by 1961, all English penal statues against suicide had been repealed, and currently suicide is decriminalized in every state (Berman, 1990).
Suicide has often been considered to be a “superseding intervening cause” that breaks the causative chain of events leading to death, thereby making it impossible to establish that any party other than the decedent could be liable for a suicide. However, more recently, courts have recognized certain instances in which civil lawsuits, referred to as tort actions, may be filed following a suicide. A tort is “a civil wrong alleged to have caused injury” (Maris, Berman, & Silverman, 2000). Specifically, in order for an individual making a tort claim to win his or her case, the following four criteria must be established. First, it must be established that the defendant had a duty to prevent harm to the plaintiff, which could either stem from a special relationship between the defendant and plaintiff or from the defendant possessing knowledge that would be key in preventing harm. Second, there must be evidence that the defendant failed to fulfill the established duty. Third, there must be evidence that the plaintiff was actually injured. Finally, there must be evidence that the harm to the plaintiff was caused by the defendant. Under the theory that suicide is a “superseding intervening cause,” the last criterion for third party liability would never be satisfied, because no third party would be deemed to have caused a suicide.
More recently, there has been a shift in view regarding suicide, and courts have begun to recognize the possibility that a suicide can be caused by forces other than the decedent’s own impulses and that third parties may be the source of these causes. Specifically, courts have recognized the following instances in which another party may be held liable for causing a suicide: the suicide was the result of tortious acts, the suicide resulted from prior physical injury, or the suicide resulted from the use of intoxicating substances. Moreover, courts have found exception to the “superseding intervening cause,” if the suicide was made possible by the negligence of a custodian (Lake & Tribbensee, 2002).
Notably, these exceptions assume that the suicide is “foreseeable,” or extremely likely to happen—either someone (or some institution) did something to cause the subsequent suicide or someone failed to adequately assess for suicide or failed to act on the knowledge of the high likelihood that a suicide would occur. Unfortunately, neither causation nor duty to prevent can be determined by a bright-line test, as human behaviors are influenced by a complex web of circumstances and pre-existing factors. Moreover, the difficulty of predicting suicide is compounded by the difficulty arising from predicting low base rate phenomena.
The number of people who die by suicide in the United States is approximately 11 in 100,000 (American Association of Suicidology, 2004). However, this rate differs depending on the population. For example, among individuals with mood disorders, there is a 2% suicide rate for outpatients, 6% suicide rate for inpatients hospitalized for suicidal symptoms, and 4.1% suicide rate for inpatients hospitalized for other reasons (Bostwick & Pankratz, 2000). Because it is not possible to be 100 percent accurate in predicting any human behavior, attempts to predict low base rate phenomena like suicide must be based on an algorithm that either errs on the side of over-prediction (false positives) or under-prediction (false negatives). Mathematically, with very low base rate phenomena, one would be more accurate predicting that the incident in question will never occur than that it will. However, when the phenomena involve life or death, it is necessary to try to predict the actual occurrence of the phenomena (true positives) with as much accuracy as possible. There are multiple predictive risk factors for suicide, including depression, substance abuse, suicidal ideation, and past suicide attempts. However, while mental health professionals could maximize their positive predictive power by attempting to protect anyone with these risk factors (perhaps through hospitalization [or at least discussion thereof] or warning family members), these are not often viable solutions as often there is no legal method to force hospitalization and the disclosure of information predicting suicide can violate ethical (and sometimes legal) standards. Therefore, mental health practitioners must walk a fine line between protecting themselves from civil litigation, maintaining patient confidentiality and trust, and recognizing when there is a high likelihood of danger.
The following is an illustration of the difficulty of predicting suicide even with a risk assessment tool that predicts suicide accurately 99% of the time (and to be clear, we know of no such instrument); that is, 99% of the people who will eventually die by suicide test positive on this instrument. Let us also assume that approximately 20% of a clinical sample will test positive on this instrument. Bayes’ theorem states that the probability of A, given B (e.g., the probability of suicide, given a positive test result) is equal to the probability of B, given A (e.g., the probability of a positive test result, given that someone died by suicide) times the probability of A (e.g., probability of suicide), the product of which is divided by the probability of B (e.g., probability that someone will have a positive test result).
Probability of suicide∣Positive test result = (Probability of positive test result∣suicide) × (Probability of Suicide) ∕ Probability of a positive test result
Because the overall base rate of suicide is so low (11 in 100,000 or 0.011%) and is in the numerator of the equation, it dramatically reduces the probability of suicide given that someone has a positive test result. Even if we assume that 99% of people who die by suicide had a positive test result , only 0.011% of all people die by suicide, the product of which is 1.09% (99% × 0.011% = 1.09%). Dividing this number by the probability of a positive test result (20%) gives us 5%, which is the probability of suicide given that someone has a positive test result (based on the above parameters). This means that 95% of the individuals testing positive for suicide risk will not die by suicide. This example demonstrates that even with risk assessment tools that are far more sensitive than those currently in use, it is still not feasible to be able to predict suicide without a large number of false positives. In this example, if we decided to err on the side of caution and hospitalize everyone with a positive test result, we would be hospitalizing 95 out of 100 people who are not actually at risk.
And yet despite the difficulty of predicting human behavior and the low base rate of suicide, the research presented above indicates that suicide is not completely unpredictable, because most suicidal people plan for their deaths. Notably, the clinical conception of “prediction,” which refers to one’s ability to predict behavior based on information obtained through an assessment, differs from the legal concept of “forseeability,” which refers to the reasonable anticipation of the possible consequence of a particular action. Generally speaking clinicians do not engage in long term risk assessment, rather they conduct short term risk assessments frequently and regularly. These short term risk assessments do, however, factor in more distal risk factors, such as a suicide attempt made several years prior.
Given the increased scientific understanding of suicide and its subsequent decriminalization, tort actions based on suicide are likely to increase. Typically, suicide litigation is constrained to four areas: malpractice claims against therapists, claims against penal institutions, life insurance claims, and workmen’s compensation cases (Maris, 1992). The discussion below will focus on the first two areas because they most clearly involve a duty to prevent harm when a suicide is foreseeable, and the research reviewed above addresses this issue most clearly. Specifically, the following section explores the relevance of impulsivity to court cases pertaining to forensic settings and mental health professionals, and we will also discuss university student suicide, as this represents a special case of the duty to warn.
Suicide attempts that involve a greater degree of prior planning also tend to be more medically serious. This implies that people who make medically serious attempts, including fatal attempts, likely had planned to do so ahead of time. Given that one key component of determining liability for the death of another individual by suicide is foreseeability, the degree of planfulness of a given suicide attempt is an important piece of information in liability suits. It is important to note, however, that although prior planning is associated with lethality, there is no guarantee that an individual will share his or her suicide plan with another person before attempting. For example, a recent study examining death by suicide in an incarcerated population found that 40% of the inmates who died by suicide over a ten-year period of time did not communicate their suicidal intent to anyone ahead of time (Daniel & Fleming, 2006). Additionally, Isometsa, Heikkinen, Marttunen, and Henriksson (1995) studied individuals who had died by suicide within four weeks of interacting with a health care professional and found that the majority (more than 75%) of the decedents had not discussed suicide intent with the health care provider.
Go to:
THE ROLE OF IMPULSIVITY IN THE FORESEEABILITY OF SUICIDE
A key legal case hinging on the foreseeability of suicide is Bogust v. Iverson (1960; Cohen, 2007). Iverson was the director of student personnel services and a professor of education at Stout State College. In his role as director of student personnel services, Iverson counseled a student named Jeannie Bogust. Although there were indications that she was psychologically distressed, Iverson was not aware that Bogust was suicidal. Six weeks after Iverson stopped counseling Bogust, she died by suicide. Bogust’s parents sued Iverson, stating that he failed to secure psychiatric treatment for their daughter and to notify them of her condition, and was, therefore, responsible for her death. The Wisconsin Supreme Court held that Iverson could not be held responsible for Bogust’s death because her suicide would not have been foreseeable to a reasonable person in his situation. Furthermore, even if Iverson had ensured that Bogust received psychiatric treatment or had notified her parents, it was the court’s opinion that there was not sufficient reason to believe that this necessarily would have prevented her suicide.
Several aspects of this case are worth noting. First, Iverson was neither a mental health professional nor a medical doctor, and as such, he did not necessarily have any specific training in conducting suicide risk assessments. Because of this, the court did not hold him to the same standards as an individual in the field of psychiatry or psychology would be held (Mallanda, 2006). Given that most suicides are not impulsive, but rather planned, perhaps with the appropriate training, Iverson would have been able to recognize Bogust’s suicide risk and to probe for resolved plans and preparations, thus making her death foreseeable (and possibly preventable). Second, this ruling occurred at a time when universities were deemed to act in loco parentis with respect to their students (Cohen, 2007). This means that the universities were believed to be acting in the place of the students’ parents, and as such had certain rights and responsibilities for their wards, including the responsibility to protect their students from harm. Because Iverson was not held responsible for Bogust’s death even under a theory imposing a heightened legal duty, this case set a precedent suggesting that universities do not have a special duty to prevent suicide in their students. Although state court decisions are not binding precedent on the courts of other states, landmark cases such as this nevertheless have significant influence on later jurisprudence addressing the issue.
Over 1,000 college students die by suicide each year (The Jed Foundation, n.d.), and even though young adults enrolled in college have lower suicide rates than members of their cohort who are not enrolled in college, suicides among university students are on the rise (Gray, 2007). Despite abandonment of the theory that universities act in loco parentis with respect to their students, there is a growing expectation that college and university administrators take action to prevent student suicides (Mallanda, 2006). Although space restrictions prevent a lengthy discussion here, there are several legal and ethical issues that are specific to college student suicide. As illustrated by the case of Jain v. State (2000), one main legal difficulty is the application of the Family Education Rights and Privacy Act (FERPA, 1974), which prohibits schools from releasing information from the record of a student over age 18 (without the written consent of the student) unless the disclosure would be instrumental in preventing harm to the student (Blanchard, 2007). Sanjay Jain was a college student who had experienced many discipline and emotional problems beginning during his first semester. After an argument with his girlfriend, the staff at his residence hall became aware of his plan to kill himself using the exhaust from his moped. Although Jain promised to go to counseling, he ended up killing himself shortly after this incident (Blanchard, 2000). His family sued the university, stating that if university staff had notified them of Jain’s distress, they could have prevented his death by ensuring that he received the mental health treatment he needed. The Iowa Supreme Court ruled, however, that Jain’s suicide was caused by his own superseding intervening act (Blanchard, 2007), such that the university was not legally obligated to prevent suicide and was not responsible for his death. This ruling indicates that universities are not obligated to notify parents of a known suicide risk faced by students, although this case is only directly applicable to court cases in the state of Iowa.
Given the research reviewed regarding impulsivity’s role in suicide, we take some issue with this precedent. Jain represents an example of a “non-impulsive” suicide, as Jain had a plan for his suicide, which he communicated to others. Although counseling was recommend for Jain, stronger action may have been warranted. As discussed, the degree of planfulness involved in a suicide attempt is an extremely important piece of information that should be dealt with appropriately. Resolved plans and preparations, like those evidenced by Jain, place one at an elevated risk for suicide, and action commensurate with this risk should be taken. For example, in addition to urging an at-risk student to seek counseling, universities could take action to secure the student’s permission to notify his/her parents of his/her condition. Although certainly not all suicidal students will grant such permission, it is important for school administrators to at least make several attempts at doing so and to document these attempts. Simply assuming that the student would not grant this permission is not adequate. Building rapport with a suicidal person’s family not only has the potential to save his or her life, but it also could prevent the family from taking legal action should a suicide occur.
One example of a more comprehensive approach to preventing arguably foreseeable suicides is the policy adopted by the University of Illinois, which requires any student who attempts or threatens suicide to attend four sessions with a mental health professional, the first of which must occur within one week of the suicidal incident. If a student refuses, he/she is dismissed from the university (Joffe, 2003). Possibly as a result of this, the University of Illinois has a suicide rate that is half that of most universities, and only one student has been dismissed because of refusal to attend counseling (Gray, 2007). This policy has been in place since 1984. Since that time, none of the students who have died by suicide at this school participated in this suicide prevention program (Joffe, 2003). Therefore, those who have died by suicide since that time did so without making any university officials aware of their risk, and thus these deaths might be considered “unforeseeable” by university officials, at least in a legal sense.
Despite the fairly convincing evidence that this type of program is effective in reducing suicide rates in a university setting, this is not the tactic taken by most institutions of higher education. Unfortunately, many college and university administrators take what is considered to be an “extreme hands off approach” rather than helping the student obtain psychiatric treatment (Gray, 2007). Under these policies at the first hint of a student’s suicide risk, the university automatically dismisses the student, in an effort to reduce liability in the event that the student eventually dies by suicide. In their desperation to not appear as though they are not in a “special relationship” with a student and thus in a position where they would be considered to have a duty to prevent suicide, universities may actually expose themselves to a greater risk of liability. There are several examples of students (and their parents) who have sued universities for discrimination under such policies, since these dismissals appear to be a violation of the Americans with Disabilities Act (Cohen, 2007; Gray, 2007). Moreover, being dismissed from a university could increase a student’s sense of thwarted belongingness, thereby placing the student at heightened risk for suicide. Furthermore, it could be argued that dismissals themselves increase a university’s liability for the suicide of a dismissed student, on the grounds that dismissal for risk of suicide necessarily indicates that the university had knowledge of that risk, but did not take any appropriate preventative measures. Given the evidence indicating foreseeability of suicides, and the effectiveness of mandatory counseling policies such as the University of Illinois’s, university administrators would better protect themselves from liability for student suicide by monitoring this risk and taking active preventative measures than by simply dismissing at-risk students.
The foreseeability of some suicides also has implications for the legal liability of prison administrators. Suicide is the fifth leading cause of death in prison and jail settings, with only heart disease, cancer, liver disease, and AIDS killing more people (Mumola, 2007). Thus, although forensic administrators are not technically mental health service providers, the residents of prisons and jails are clearly at elevated risk for suicide, and staff at these facilities should be properly trained in recognizing suicide risk. Unfortunately, jails and prisons are not legally obligated to screen for suicide risk, as Burns v. City of Galveston (1990, 5th Circuit) determined that psychological screening is not a necessary component of the medical care required by the 1983 Civil Rights Act (Franks, 1993). Without this screening requirement, it is much more difficult to establish that a suicide was foreseeable by a prison staff, unless the inmate (of his or her own volition) reveals his/her suicide plans to another individual. Furthermore, in forensic suicides, a jailer’s liability (at least under a Constitutional claim) may depend on showing that he or she displayed “deliberate indifference” to the risk of an inmate’s suicide (Franks, 1993). This deliberate indifference standard was set in Estelle v. Gamble (1976) by the U.S. Supreme Court, which found that medical treatment that is simply inadequate is not necessarily in violation of the Eighth Amendment, which forbids “deliberate and unusual punishment.” Rather, the Eighth Amendment would be violated, the Estelle Court held, only if the prison staff were deliberately cruel in their disregard of the decedent’s medical needs. Notably, unlike tort issues, Supreme Court cases which apply an interpretation of the US Constitution are binding everywhere. Even the knowledge of past suicidal behavior coupled with a staff failure to take precautionary measures is not enough to establish liability for a government entity, as seen in Freedman v. City of Allentown (1988, 3rd Circuit).
One desirable solution is proposed by Franks (1993), who recommended that prison staffs should screen their inmates for suicide risk even if not legally required to do so, given the potential that this screening has to save lives and to curb litigation risk should future case law determine that suicide screening is a right of prisoners. Nevertheless, screening alone is not sufficient to prevent suicide; it is also necessary to take appropriate action to prevent a suicide attempt once risk has been assessed. Ninety-two percent of the suicide decedents in Daniel and Fleming’s (2006) ten-year study of suicide deaths in a state prison system had been screened for suicide; clearly, this was not enough to prevent their deaths. A large percentage of these suicides were arguably foreseeable by the prison staffs: Sixty percent of the inmates who died by suicide expressed suicidal intent to someone before dying, and 82% of these individuals communicated their intent within one week of their fatal attempt. Sixty-five percent of the inmates had made prior attempts, and of these, 71% had previously attempted while incarcerated. Tragically, none of the suicide decedents were on suicide watch at the time of their death. Although the failure of prison administrators to take action in response to knowledge of factors indicating suicide risk might not fit the standard of “deliberate indifference” that would constitute a Constitutional violation, it may nevertheless be grounds for civil liability. This is not to say that anyone who attempts suicide should be placed on suicide watch indefinitely; but rather that a reasonable way to protect oneself from liability, as well as to increase suicide prevention, is to provide adequate care for inmates at high risk for suicide. In Daniel and Fleming’s (2006) study, it seems reasonable to assume that at least some of these deaths could have been prevented if only prison officials had recognized and dealt with the increased risk status of people who had made recent suicide attempts.
It is important to note that more than a tenth of jail suicides occur within 24 hours of arrest (Shaw, Baker, Hunt, Moloney, & Appleby, 2004), which might suggest that these suicides were enacted impulsively. Although it is possible that some individuals (including non-inmates) impulsively decide to die by suicide, the evidence points to the contrary. As Conrad and colleagues report, inmates are at higher risk for suicide throughout the course of their lives, even before they are first arrested (2007). This increased risk is due in part to their greater likelihood to have a mental illness and/or substance use problem, which are known risk factors for suicide. Moreover, the prison environment itself has been implicated as increasing suicide risk (Liebling, 1994). Thus the fact that many inmates die by suicide shortly after they are arrested may speak to the fact that they are at a higher risk for suicide (compared to non-inmates) at the time of their arrest and that this diathesis interacts with the potent stressor of being arrested to increase the likelihood of suicide. Another plausible explanation is that these inmates may have previously thought about or made a plan for their suicide, but did not decide to use this plan until they were arrested. People oftentimes make advance plans for certain situations (e.g., carrying an umbrella in the car in case of a rainy day), and then do not necessarily think about their plan until the situation arises (e.g., a rainy day). Thus, although the decision to use a suicide plan can be made so quickly the suicide itself looks impulsively enacted, the act itself likely involved some planning. Therefore, just because an individual dies shortly after being arrested does not mean that this individual died impulsively, but rather that being arrested pushed one or more of the three proximal and necessary causes for suicide to the intersection point.
A final category of suicide liability cases impacted by new evidence indicating the predictability of suicide pertains to malpractice cases against the mental health professionals who treated a client who died by suicide. As discussed above, tort law theory specifies that an individual may be held liable if, among other requirements, he or she has a duty to prevent harm to another individual and he/she failed in that duty. It seems relatively straightforward that a mental health professional does have a duty to prevent the suicide of one of his/her clients. Indeed, this theory is supported by Kockelman v. Segal (1998), in which the California Court of Appeals held that psychiatrists have a duty to attempt to prevent suicide, and this duty holds even for outpatient psychiatrists (Packman, Pennuto, Bongar, & Orthwein, 2004). However, the actions required by that duty depend on the foreseeability of the suicide. Again, although most people who attempt or die by suicide do not do so impulsively, it is another matter whether they are willing to share their plans, should they have them, with another person, including their therapists.
Gross (2005) discusses three ways a therapist might be held liable for a client’s suicide. The first is for negligence, which involves not using reasonable care to prevent another person’s injury. The second is deliberate indifference, defined above in reference to suicide in forensic settings. The third is malpractice, which involves not meeting professional standards. Notably, malpractice differs from general negligence liability in that general negligence refers to what a reasonable person would or would not do to protect another individual from foreseeable harm, whereas malpractice refers to behavior that is culpable because it falls short of what the professional standards for a mental health practitioner call for to protect a client or patient from foreseeable harm. Issues related to reasonable care and deliberate indifference are addressed above as they relate to suicides in university and forensic settings; thus, the following discussion focuses on issues relevant to malpractice.
Simon (2002) discusses a proposed standard of professional care for psychiatrists in terms of suicide risk assessment. Although suicide risk assessment is considered to be vital for preventing suicide attempts and completions, many psychiatrists (and likely, many psychologists and other mental health professionals) do not conduct adequate risk assessments. Reasons for this omission are numerous, but include a perceived lack of time to do them, anxiety about suicidal behavior in general, the mistaken belief that documentation of risk assessment can make one more vulnerable to lawsuits, and inadequate training in suicide risk assessment (Simon, 2002). In Stepakoff v. Kantar (1987), the Massachusetts Court of Appeals discussed the method used to establish a professional standard of care and held that it requires actions consistent with what most similar professionals do in their everyday practice. One difficulty with this approach, however, is that it is based on an evaluation of status quo professional standards rather than ideal professional standards. For example, many psychiatrists do not conduct appropriate risk assessments for suicide, although it could be argued that they should be expected to do so. In other words, it is not safe to assume that one is not liable for malpractice simply because one is similarly negligent to other people in one’s profession (Simon, 2002).
The better approach for mental health professionals is to strive for what the ideal therapist would do rather than what the “average” therapist would do; this will increase client safety and reduce exposure to legal liability. Part of this should include keeping up-to-date with research on impulsivity, as impulsivity is a well-documented risk factor for various mental disorders and maladaptive behaviors. Recent research on impulsivity has demonstrated that most people are not likely to attempt suicide impulsively. In light of this, clinicians need to conduct regular, systematic suicide risk assessments, document these assessments, take actions commensurate with the degree of risk, and document these actions. It is also important to seek information for the risk assessment from sources other than the patient (e.g., past mental health providers, family members) — with proper consent that is — because the patient may be motivated to conceal suicidal intent (Simpson & Stacy, 2004). Specifically, clinicians should make several attempts to secure permission from the client to speak to these other sources and appropriately document this effort. More importantly, conducting these risk assessments can save lives if an individual is determined to be at risk for suicide and appropriate preventative measures (e.g., inpatient hospitalization, close monitoring) are taken. If a clinician does all of the above, this will not necessarily protect him or her from being sued (although Gross [2005] suggests that most plaintiffs will not pursue litigation if the therapist has taken these measures), but it will likely ensure that the therapist is not found to be responsible for the client’s death.
With regard to risk assessment, it is not enough to simply ask a client about the presence of suicidal ideation (Gross, 2005; Joiner et al., 1999; Simon, 2002); other factors must be considered as well and included in the clinician’s risk assessment. For example, Beck has shown that future acts of violence are better predicted by an individual’s previous behavioral pattern than by verbal threats of violence (1998). Similarly, one of the most potent predictors of future suicidal behavior is previously engagement in suicidal behavior (Joiner et al., 2005). Thus, the risk assessment should combine client self-report with past history from at least one other source (e.g., a family member, mental health records). Given impulsivity’s positive association with suicidal behavior, it stands to reason that an individual’s degree of impulsivity should be one component of the risk assessment. In line with our discussion of impulsivity’s indirect and distal relationship with suicide, the rationale for this is that impulsivity increases the likelihood that an individual will acquire the capability for suicide, and according to Joiner’s (2005) theory of suicide, an individual must have acquired this capability before s/he will die by suicide. Therefore, clinicians should not be overly focused on an individual’s level of impulsivity per se; rather, more time should be spent determining whether the individual’s level of impulsivity has in fact led to a lifestyle fraught with painful and provocative experiences, which should be included in the risk assessment as well. Also, given the evidence that impulsivity tends to decrease with age (Okun, 1976), prudent clinicians should consider whether an individual used to be impulsive and as such, was exposed to many painful and provocative experiences that have increased his/her enduring acquired capability for suicide.
Inpatient providers are generally considered to be more responsible for preventing harm to clients than outpatient providers, which is due partially to the greater severity of the illness of these patients and also because the patients are under constant care. One example of an inpatient psychiatrist being held liable is the Weatherly v. State of New York (1981) case, which involved a patient who was released from suicide watch in the hospital eight days prior to jumping out of a hospital window. The day after being removed from suicide watch, the patient’s mental status began to deteriorate notably, yet he was not placed back on suicide precaution. The psychiatrist was held responsible for this death because it could reasonably be considered “foreseeable.” In this example, it could be considered irrelevant whether the patient attempted “impulsively,” as a reasonable caretaker should have been aware that the individual was at elevated risk, thus necessitating closer monitoring.
Courts have not typically held outpatient mental health professionals to the same standard as inpatient mental health professionals, presumably because the former have less control over and access to their clients (Bongar, Maris, Berman, & Litman, 1998; Kjervik, 1984). This standard was set by Speer v. United States (1981, 5th Circuit), which involved an outpatient psychiatrist prescribing a month’s worth of medicine, which was used by a patient to make a fatal overdose. The court ruled that the psychiatrist was not responsible for this suicide because as an outpatient, the suicide victim could not be directly monitored and kept under his therapist’s control. Still, there is indication that the number of lawsuits against outpatient providers is increasing (Packman, Pennuto, Bongar, & Orthwein, 2004), and some plaintiffs are successful in proving liability. In Bellah v. Greenson (1978, California Court of Appeals), an outpatient died by suicide via overdose. Her psychiatrist had documented that he perceived her to be at risk; however, there was no indication of preventative measures being taken; as such, he was found responsible for her death. Interestingly, most tort cases are resolved out-of-court because defendants tend to be too risk-averse to go to trial, regardless of whether or not they could prevail on legal grounds, due to the notoriously unpredictable nature of juries (Gantler & Cahill, 1994). Thus, an increasing body of precedents holding third parties liable for suicide could also increase the likelihood that defendants will feel pressured to settle out of court, since these precedents may make the defendants even more apprehensive about being held liable if the suit goes to court.
In sum, a review of the relevant literature did not uncover the use of impulsivity per se in a determination of liability. Nevertheless, the breakdown of the causal link between impulsivity and suicide means that courts will continue to have to recognize that a suicide can be caused by forces other than the decedent’s own impulses, and that third parties may be the source of these causes. Moreover, the breakdown of the link between impulsivity and suicide and the resulting implication that suicides are often foreseeable to a third party has a great deal of impact on the third party’s duty to prevent the suicide. As we have argued, it is not reasonable to expect a third party to accurately predict all suicides, but it is reasonable to expect that adequate risk assessments be conducted and preventative measures be taken to ensure the safety of suicidal individuals. Given the evidence that most suicidal behavior is preceded by a plan, caretakers should make an effort to determine whether a plan for suicide exists; the prevailing evidence does not support the notion that suicides are by definition unforeseeable (e.g., that people usually attempt “impulsively”).
The ultimate point is that at least some suicides can be prevented if some person with a degree of responsibility for the victim (e.g., school counselor, therapist, prison guard, etc.) has the knowledge and training to probe more carefully to determine whether a suicide plan has been devised. Although Daniel and Fleming (2006) and Isometsa and colleagues (Isometa, Heikkinen, Marttunen, & Henriksson, 1995) showed that a substantial proportion of suicide decedents appear to die by suicide unexpectedly, these individuals are not necessarily systematically screened for suicidal thoughts. As such, it is possible that if university and prison staffs and mental health professionals carefully probe for a suicide plan in their students, wards, and patients some lives can be saved. A failure to probe for such a plan could be a determining factor in a malpractice case filed for negligence on the part of the caregiver. It also is a good idea to probe for the existence of prior specific plans for suicide. For example, Joiner et al. (2003) found that the only predictor of eventual death by suicide among suicidal psychiatric outpatients was their “worst point” (i.e., most severe in the individual’s lifetime) resolved plans and preparations for suicide. Resolved plans and preparations are distinct from desire for suicide insofar as they are indicative of an individual feeling capable of, fearless about, and intent on using a specific type of method (to which they have access) to make a suicide attempt, whereas suicidal desire is more vague and consists of general thoughts about suicide and wishes for death. In Joiner et al.’s (2003) study, current or “worst point” suicidal desire was not predictive of subsequent death, nor were current resolved plans and preparations; it was the “worst point” resolved plans and preparations that were key in predicting suicide risk.
Go to:
CONCLUSION
Given that suicide is a leading cause of death for prisoners and young adults, and that over half of practicing psychiatrists and close to a third of practicing psychologists will lose a patient to suicide at some point during their career (Berman, 1990), suicide litigation will continue to affect families, universities, prison staff, and mental health practitioners in a significant way. Thus, in the wake of this increased litigation, which is costly both emotionally and financially for all parties involved, it is important that people are aware of new research in the field of suicidology, as this research has the potential to inform legal theories of liability for suicide.
Recent research has begun to clarify the role of impulsivity in suicidal behavior. Though people who attempt suicide tend to have higher levels of trait impulsivity, the majority of these people do not attempt suicide “impulsively,” and in fact have prior plans regarding their attempts. This implies that the majority of suicide attempts have an element of foreseeability to them, and thus it is critical that knowledgeable people not only routinely assess for suicide plans, but also take action appropriate to the level of risk (for a review of suicide assessment and appropriate action plans, see “The suicide assessment decision tree,” Joiner, Walker, Rudd, & Jobes, 1999.) Sound suicide assessment has the potential not only to save lives, but it will also protect individuals from liability should legal action ensue. This is not to say that an individual will not die by suicide even if preventative efforts have been made. However, if a person who has denied the existence of a plan for suicide dies by suicide, it would likely be quite difficult to find another party liable for failing to prevent this suicide from occurring if that party checked for other risk factors (e.g., past attempts, “worst point” resolved plans, etc.) and made appropriate recommendations in light of all known risk factors. Moreover, in the worst case scenario, where a suicide occurs despite appropriate assessment and action, the knowledge that the standard of care was rigorously followed will hopefully ease feelings of guilt or inadequacy that often follow a death by suicide, as well as form a defense to civil liability for the suicide.
To summarize, the main aim of this article is to correct misunderstandings and assuage some of the fears regarding suicide that many people, including mental health practitioners, hold. Another aim is to provide recommendations for meeting (and preferably, exceeding) the standard of care when encountering suicidal individuals in one’s profession. Not all suicidal individuals will broadcast their intentions to harm themselves; thus, it is important to make a concerted effort to uncover them if they do exist. A recent article in the New York Times stated that suicide is “an intimate, often impulsive decision that has defied scientific understanding” (Carey, 2008). As the above review indicates, suicide is rarely an impulsive decision and we know this precisely because of our increased scientific understanding resulting from decades of research.
