Natalie H Polvi. Encyclopedia of Psychology & Law. Editor: Brian L Cutler. Volume 2, Sage Publications, 2008.
Prisoner suicide assessment and prevention is an area of active research and clinical involvement. Indeed, it is an important component of the forensic psychologist’s clinical responsibilities due to the disproportionately high incidence of prison suicide as compared with the general population. This high incidence is a consistent phenomenon across countries. In some countries, this translates into suicide being a major cause of death among prisoners. To address the topic of suicide assessment and prevention, first the research challenges in conducting prisoner suicide research and the relevant theory in conceptualizing the process of prisoner suicidality are briefly summarized. A review of relevant risk factors as evidenced by research are categorized into several domains and described. These domains include demographic factors, historical factors, criminality factors, and clinical factors. Following a review of risk factors, suicide prevention within the prison setting is summarized.
Research and Theory
In an effort to better understand and assess (i.e., predict) prisoner suicide, a large body of research has investigated the risk factors related to prisoner suicide. However, there are some intrinsic challenges in the prediction of suicide generally, and there are some methodological weaknesses in prisoner suicide research specifically. The major challenge in predicting suicide is that it is a relatively rare event (i.e., has a low base rate). Statistically, it is more difficult to predict a rare event than a frequent event. This creates significant challenges for researchers in designing good-quality predictive studies. For example, because suicide is a relatively infrequent event, a researcher would need to have a very large number of subjects at the beginning of a study for there to be a sufficient sample size of eventual suicides for analyses. For individual clinicians, this difficulty in predicting rare events creates the higher risk of false positives (i.e., predicting suicide where none occurs). False negatives (a suicide occurs when it was predicted that no suicide would occur) literally have life and death implications and represent what most clinicians and staff members want to avoid.
Some particular methodological weaknesses in prisoner suicide research include samples consisting of mixed populations of prisoners (e.g., remanded and sentenced prisoners), lack of control or comparison groups of nonsuicide or nonattempter prisoners, and reliance on descriptive studies, which generates difficulties in establishing a causal relationship between risk factors and outcome. Of note is that, in recent years, researchers have made efforts to address these methodological problems. There have been efforts to design more comparative studies, to use more sophisticated statistical analyses (e.g., logistic regression), and to even undertake matched control studies. This is a welcome shift in the research approach to prisoner suicide. Indeed, given the plethora of descriptive studies generated over the last 25 to 30 years, there are limitations in the value added by purely descriptive studies of prisoner suicide at this time. Designing more methodologically sound studies will more meaningfully build on current knowledge.
Overall, despite the challenges presented by researching prisoner suicidality, identifying relevant risk factors has facilitated the development of suicide assessment protocols and scales. Prior to reviewing relevant risk factors, it is important to note that, as useful as individual risk factors are, there has been a major contribution by researchers who have offered valuable conceptualizations of the suicidal process as one that occurs over time and is affected by multiple factors. In particular, Marti Heikkinen and colleagues have provided one of the most well-developed models. Their process model of suicide consists of risk factors, precipitating factors, vulnerability factors, and protective factors that contribute to the process of suicidality. In this model, the suicidal process is viewed as dynamic and affected by several categories of risk factors (i.e., biological, psychological, social, cultural). At the same time, the individual’s vulnerability and protective factors affect risk. For instance, a prisoner who has difficulty in coping with various areas of his life will likely experience greater risk. A prisoner with a strong social support system would likely experience some protection against risk. Finally, the model includes precipitating factors such as stressors and external events that contribute to triggering suicidality (e.g., loss of support, negative decision regarding release).
Research on prisoner suicidality reveals several domains of risk factors that are relevant to suicide assessment. These general domains include demographic factors, historical factors, criminality factors, and clinical factors.
During the last couple of decades, predominantly descriptive research has linked prisoner’s age with suicidal risk. Generally, the research has suggested that younger age groups (e.g., approximately under the age of 30 years) are overrepresented in suicide completer samples. However, a couple of recent and more methodologically sound research studies have challenged this conclusion, suggesting that either no relationship exists or that older age (i.e., 40 years and above) is predictive of prisoner suicide. More research will be required to clarify predictive relationships between age and prisoner suicide. A relatively new result provided by one of these recent studies was identifying homelessness as a predictive factor. It is worth mentioning that many descriptive studies have examined marital status as a correlate of prisoner suicide. These results have been equivocal and not uniformly supported by the recent better-quality research.
The research has been fairly convincing that both the presence of a psychiatric history (typically broadly defined in prisoner suicide research) and a history of substance abuse are connected to an increased risk of prisoner suicidality. Recent research using the matched control methodology and/or logistic regression analyses has supported these predictive relationships. Some research suggests that recent psychiatric contact or intervention may possess additional predictive power. Results from the larger body of suicide research have revealed the increased risk generated by a family history of suicide. In particular, a genetic component has been attributed as partly responsible for the relationship. Individuals with first-degree relatives (i.e., parents, siblings) who committed suicide are at greater risk for committing suicide. This risk becomes more elevated if the relative suffered from a mood disorder (i.e., depression, bipolar disorder). Given this research, it is important to consider this factor in assessing prisoner suicidality.
Several criminality factors are linked with greater risk of prisoner suicidality. These include sentence length, time served in sentence, security level, criminal history, and institutional adjustment. In terms of a prisoner’s sentence length, generally prisoners with lengthier sentences are disproportionately represented among prisoner suicides. In particular, those prisoners with life or indefinite sentences may be at higher risk. The amount of time served in one’s sentence is also linked to prisoner suicide but not as definitively as sentence length. Generally, prisoners who commit suicide do so earlier in their sentences (within approximately the first 2 years of being sentenced). A prisoner’s security level appears to be relevant to suicidal risk. A limited amount of recent research, some of which has used logistic regression, has revealed an overrepresentation of higher security prisoners (i.e., maximum security level) among suicide completers and attempters. Recent more methodologically sound research has indicated that several characteristics of prisoner criminal history are linked with suicidal risk. In one study, suicide attempters were more likely to have current convictions for homicide, break and enter, or robbery. Consistent with that result were two studies that found having a current violent offense was more predictive of suicide completers. In addition, prisoners with a history of prior criminal involvement (variously defined as prior offense[s], prior incarceration) were more likely to attempt or commit suicide. One study found that suicide completers and attempts were more likely to have had breach of trust offenses (i.e., escapes, violations of parole or probation). Finally, limited recent research using comparison groups and logistic regression found that both suicide completers and attempters had demonstrated negative institutional adjustment (e.g., institutional violence, contraband violations, substance abuse incidents, escape, requests for protective custody). Prisoners with a history of contraband-related incidents were three times more likely to attempt suicide. Those with a disciplinary history were 19 times more likely to engage in a suicide attempt. Both suicide completers and attempters participated in correctional programs less than nonattempters.
In addition to assessing the risk factors characteristic of suicidal prisoners, there is a fundamental role for the assessment of relevant generic clinical factors as part of the suicide assessment. In other words, a good assessment of prisoner suicidality is predicated on conducting a competent clinical assessment. In fact, there are several salient clinical factors that require particular attention. Clinical domains and factors important to the suicide assessment are described. The work of John and Rita Sommers-Flanagan has been used to lend some structure to the description of clinical factors. In addition, where relevant, prisoner suicide research related to that factor is summarized. The overarching clinical factors include the presenting problem, depression, suicidal ideation suicidal intent, suicidal plan, self-control, vulnerability, and coping.
Similar to a suicide assessment with a nonforensic client, the initial goal is to attempt to establish rapport and determine the nature of the presenting problem. Determining the prisoner’s level of distress and coping efforts will provide some indication of how to pace the remainder of the assessment. Identifying the precipitating factors and current stressor(s) provides some contextual and situational information.
There is a strong relationship between depression and suicidality as well as between depression and hopelessness. If a user combines alcohol and depression, then risk will further elevate. Therefore, it is important to determine the presence of depressive symptomatology. The diagnostic criteria and symptomatology of depression are not detailed here. Rather, relevant domains of functioning are reviewed as a means of suggesting some structure for the assessment process. These include emotional, physical, cognitive, behavioral, and social domains.
The emotional domain primarily refers to determining the presence of depressed mood and related factors such as frequency, intensity, and duration. Of particular concern is the presence of hopelessness. Research has established hopelessness as a strong predictive factor of suicide generally. Available research investigating this factor in prison populations has confirmed the predictive relevance of hopelessness. An additional emotional factor that warrants attention is the occurrence of a sudden and unexplained change in the individual’s mood and/or functioning. This is a salient clinical sign that has traditionally been interpreted as an indication of increased risk. Experts in suicide assessment suggest that the improvement may result from the individual making a decision about ending his or her emotional pain or result from an alleviation of mental illness. The suggested dynamic is that either of these occurrences reduces ambivalence, brightens moods, and frees up energy to act (and possibly carry out a plan for suicide).
The physical domain refers to determining the presence of physical symptomatology indicative of depression. Relevant factors include appetite, weight, sleep, energy level, concentration, psychomotor functioning, and self-care.
The cognitive domain involves assessing whether cognitive functioning is intact. For example, there may be the presence of thought distortions, disorganized thought, impaired judgment, or event psychotic symptoms. Research has also pointed to the relationship between depression and the presence of negative thinking about oneself, the world, and the future (referred to as the cognitive triad).
The behavioral domain refers to behavioral symptoms of depression that can be observed. These may include decreased pleasure in one’s usual activities, decreased physical activity, restlessness, poor concentration, and poor problem solving. Changes in self-care and other negative behavior may be present.
The social domain refers to interpersonal and social functioning. Some examples can include social withdrawal, rejecting support, interpersonal conflict, and decline in social skills.
Suicide Ideation and Suicidal Intent
Suicidal ideation and suicidal intent are related to increased risk for suicide. Ideation does not necessarily result in high risk. Expressing suicidal intent generally presents a greater risk than ideation. Inquiring directly about ideation and intent is important. Questions regarding frequency, duration, and intensity can provide additional information. In addition, collateral information and/or behavioral observations can be useful. If the prisoner commits or contracts, it is suggested that the commitment be made for both self-harm and suicide rather than assuming the commitment for one act will generalize to the other.
Having a suicide plan can present a serious level of risk. Determining the details is crucial. Relevant domains of functioning to assess include prior suicide attempts, specificity, lethality, availability, and proximity.
A history of prior suicide attempts increases the risk for suicide. A suicide attempt within the past year elevates risk even further. Obtaining details about the prior attempts can help identify any patterns or past precipitants that may be relevant to the current situation.
Specificity of the suicide plan needs to be determined. Generally, the more detailed (i.e., high specificity) a plan, the greater the risk of suicide.
Lethality of a suicide plan is defined as the amount of time that passes between initiating the suicidal act and dying. High lethality is a plan that results in a quick death (e.g., hanging) and translates into high risk. In addition to the lethality presented by a specific method, there is also an impact resulting from how the method will be used. Research on prison suicide has revealed that hanging, a very lethal method, is the primary method of committing suicide.
By examining a prisoner’s behavioral history (e.g., history of impulse control difficulties) and obtaining information about previous suicide attempts, the prisoner’s degree of self-control can be assessed. Another factor affecting self-control is the use of alcohol or substances.
Coping and Vulnerability
A prisoner’s poor ability to cope is a risk factor of suicide. Research has identified a component of the prison population that is particularly vulnerable and poor copers. These individuals tend to have difficulty coping across time and situations.
Psychosocial isolation (e.g., emotional and social support) increases the likelihood of suicide. The prisoner’s access to emotional and social support resources should be assessed.
Physical isolation of suicidal prisoners can have a detrimental effect. Research addressing this issue had revealed that placing suicidal prisoners in some form of isolation (e.g., constant observation) is quite detrimental and can actually contribute to increased suicidal risk.
Suicide prevention is typically conceptualized as an institutional or organizational approach to preventing prisoner suicides. Among the preventive strategies, there can be policies and procedures that specify the management of identified suicidal prisoners (e.g., type, frequency, and/or location of observations; mental health referrals). More broadly, there can be policies, procedures, and programs designed to improve identification of suicidal prisoners. Screening for suicidal risk at intake can be a valuable preventive strategy. During the last decade, there has been an appreciable amount of research directed toward developing effective screening instruments. Some scales are designed to be administered by nonclinical frontline staff at intake (e.g., correctional staff), while other scales require mental health training to administer. Some screening instruments have been designed for specific settings (e.g., remand centers vs. prisons), while other scales have been designed for use across a variety of settings. Implementing a screening instrument can be influenced by human resource and cost factors. A brief scale that requires no mental health training to administer is typically less resource intensive than a scale that may require clinically trained staff (e.g., nurse) to administer. However, the administrative costs must be weighed against the effectiveness of the scale to accurately identify prisoners as potential suicide risks.
An important component of suicide prevention programs includes training programs for prison staff. These programs can vary in scope. Some training programs target improved knowledge about the indicators of suicidality, while others target skill building in detection and basic intervention. Indeed, some institutions or jurisdictions have developed programs that provide training to prisoners in an effort to improve knowledge and/or provide skills that facilitate detection and peer support. Yet another suicide prevention strategy can involve changes to the physical environment that essentially decrease opportunity for a suicidal prisoner. For example, changes in cell location (e.g., observation cell) may improve visibility of a suicidal prisoner. Improvements to the physical structure of a cell may include installing tamper-proof fixtures and eliminating structures that provide opportunities to implement a suicide plan (e.g., a noose attached to a pipe).
Overall, research supports the conclusion that suicide prevention programs can reduce the incidence of prisoner suicides. Wider implementation of suicide prevention policies, procedures, and programs is necessary to further advance prevention efforts. Indeed, more comprehensive suicide prevention programs would also be beneficial.