While the terms suicide and self-harm are often used interchangeably, it is important to be clear that they are conceptually different. Suicide is a self-inflicted act with the intention to cause death, whereas an act of self-harm may be the precedent to an act of suicide, but it can also exist as a concept in its own right as an intentional act of harm without suicidal intent. Self-harm is intentional, often repetitive, and often constitutes a coping mechanism to provide temporary relief from emotional distress. The crucial differentiation between suicidal behavior and self-harm lies in the nature of motivational intent underlying the action. This distinction makes the early identification of those who are at risk of self-harm or suicide more difficult and the screening of potential at-risk populations more complex. It has been argued that as the impact, treatment pathway, and even costs are similar for both groups, this difference is considered irrelevant for the screening of these behaviors. Skegg in 2005 suggested that self-harm and suicide should be considered as a continuum in the sense that self-harm covers a spectrum of behavior: The most serious form closely relates to suicide, while behaviors at the milder end of the spectrum range merge with other reactions to emotional pain. And it is this approach most commonly employed in the development of screening tools to identify those at risk of suicide and self-harm, whereby self-harm is defined as an act of self-poisoning or self-injury, irrespective of the purpose of the act. This article addresses the prevalence of suicide and self-harm in offender populations, lists risk factors for offender suicide and self-harm, and provides an overview of assessment tools.
Prevalence of Suicide and Self-Harm
Suicide and self-harm represent a significant global health challenge. In 2014, the overall rate of suicide in the United States was 12.93 per 100,000, a 20% increase since 2005, and the rate was almost double in males aged between 20 and 35 years. The rates of self-harm were estimated to be 12 times that of suicide, with women more likely to self-harm than men.
Offender populations are some of the most marginalized people in society, with higher rates of socioeconomic disadvantage and worse physical and mental health, so it comes as little surprise that the rates of self-harm and suicide are far greater in this population than the general public. Suicide is the most common cause of death in U.S. jails, with a rate far higher than the general population at 46 per 100,000, and this has increased 12% since 2009. Epidemiological studies suggest the rate of self-harm in offender populations is 30 times that of the general population with 30% of prison populations reporting self-harm behaviors; the rate rises to 50% for female offenders.
Risk Factors for Suicide and Self-Harm in Offenders
In order to develop screening tools for suicide and self-harming behavior, a number of authors have explored the risk factors associated with these behaviors in the offender population.
Demographic and Environmental
There is clear evidence that gender plays a role in whether an individual is likely to attempt suicide or engage in self-harm, with far more women likely to engage in self-harm and more men attempting suicide. Age is also a factor: Older offenders are more likely to attempt suicide and younger offenders engage in self-harm. Being Caucasian, greater socioeconomic disadvantage, having lower educational attainment, and not being in a relationship are also strong predictors of both self-harm and suicide in offender populations. First time incarceration, long sentence, and being on remand in prison are both associated with greater risk of self-harm and suicide irrespective of age and gender. In the United States, being incarcerated in the county jail system as opposed to the state prison system is associated with a higher risk of both self-harm and suicide.
Lifetime
Previous self-harm and suicide attempts are strong predictors of future self-harm and suicidal behavior. Childhood abuse, either physical or sexual, has been associated with greater risk, and childhood sexual abuse is particularly associated with higher risk in offender populations. Poor social networks, isolation, and the loss of significant others are also associated with increased risk.
Mental Health
Clinical diagnosis of depression, psychoses, or a history of substance abuse has been found to increase the risk of future self-harm and suicide in both male and female offenders. For women, in particular, a clinical diagnosis of personality disorder is associated with an increased risk of self-harm.
Coping Style
Studies in offender populations have identified weaker cognitive coping strategies, particularly the use of avoidant strategies, and poor problem-solving skills to be associated with increased risk of self-harm and suicide. Greater interpersonal conflict, internalization of conflict impulsive reaction, and resentment are all associated with greater risk of both self-harm and suicide.
Emotions
Increased risk has been found to be associated with higher self-rated aggression, greater impulsivity, and more severe feelings of shame. No increase risk has been associated with general dysphoria or sadness.
Screening Tools
The risk of suicide and self-harm presents a major challenge to managing the health of offender populations, and there is a clear need for efficient screening tools to identify those at risk. In addition to being diagnostically efficient, screening tools need to be simple to implement, complete, and interpret.
Interpreting the Tools
A screening test aims to identify those at increased risk in a timely manner in order to facilitate access to appropriate intervention and improve outcome. A screening tool will provide an indication whether an individual is positive or negative for increased risk. No screening tool is infallible, so those identified as being positive will include a proportion who are not truly positive and those identified as being negative will include a proportion who are not truly negative. The sensitivity of a screening tool is the proportion of those who screen positive who are truly positive (i.e., the proportion of those identified as being at increased risk who go on to engage in self-harm or attempt suicide) and the specificity is the proportion of those who screen negative who are truly negative (i.e., the proportion of those identified as being at no increased risk who do not later engage in self-harm or attempt suicide). The true positive and true negative rate is established by reference to a gold standard, sometimes a different test that is known to accurately identify positive and negative cases, or a behavioral outcome measured at some point in the future. While a hypothetical gold standard would be 100% sensitive and 100% specific, it will often be far more complex to administer than a simple screening test.
The efficiency of a screening tool involves a trade-off between sensitivity and specificity in order to avoid misclassification and aims to identify as many truly positive individuals and as many truly negative individuals as possible. As the potential implications of misclassifying an individual as negative for suicidal intent are high, sensitivity of the tool may be prioritized over the specificity, maximizing the identification of positive cases at the risk of misclassifying negative cases as positive. In addition, the predictive value of a screening tool is a useful concept to take into account. The positive predictive value is the proportion of those who are classified as positive who go on to commit suicide or engage in self-harming behavior, and the negative predictive value is the proportion of those who are classified as negative who do not go on to commit suicide or engage in self-harming behavior.
When assessing the efficiency of a screening tool, the tool must be compared with a gold standard, a criterion test that is established as a diagnostic test or benchmark. As noted earlier, a hypothetical gold standard test is considered to be 100% sensitive and 100% specific. In mental health research, gold standard tests are rare and where they do exist they are usually based on historical report of populations who are known to be at risk, and as such, the interpretation of any screening tools should always carefully consider the context of the gold standard comparison.
Review of Screening Tools
While a number of studies have been conducted to explore the utility of screening tools to identify those at increased risk of self-harm and suicide, there is a paucity of research focusing on offender populations. A major issue arises when researchers and clinicians simply adopt screening tools used in general or psychiatric populations to offender populations without careful consideration of the differential nature of the environment. Offenders often have higher rates of depressive symptoms; they have often experienced significant psychological trauma, have been removed from their usual environment, and have limited access to their social networks. Screening instruments that assess feelings of guilt, perceptions of punishment, and lack of control have limited validity in the offender population.
A comprehensive systematic review of screening tools focusing on suicide and self-harm in offender populations was conducted by Amanda Perry and colleagues in 2010. The review identified 404 research studies in the published literature and an additional 75 studies that had not been published. Only five studies assessing the screening properties of four instruments were considered to be of sufficient scientific quality to be considered in detail. Of these four instruments, two were found to have poor diagnostic properties and unacceptable predictive value: the Suicide Checklist and Suicide Probability Scale. Two instruments demonstrated acceptable levels of diagnostic properties and acceptable predictive value: Suicide Concerns for Offenders in Prison Environments and Suicide Potential Scale. In addition, a review by Simon Gilbody and colleagues 2014 identified the Beck Hopelessness Scale as being useful for screening offenders for risk of self-harm.
Suicide Concerns for Offenders in Prison Environments
This 28-item self-report instrument was developed specifically to identify offenders at increased risk of suicide and self-harm. The instrument assesses risk across two domains. The first relates to factors associated with a protective social network and assesses the presence of support, innate coping skills, and problem-solving abilities. The second relates to factors relating to depression, hopelessness, self-harming history, and suicidal ideation. Each question is rated on the Likert-type scale, scored 0–5, ranging from strongly disagree to strongly agree, with a score of 38 or more on the first factor and 30 or more on the second factor being indicative of a positive case. The instrument is quick and easy to complete and has demonstrated good psychometric properties in differentiating between offenders with a history of self-harm and those without. As a screening tool, it demonstrates acceptable levels of sensitivity at 81% and specificity at 71%. Suicide Concerns for Offenders in Prison Environments demonstrated excellent positive predictive value (94%) and acceptable negative predictive value (55%).
Suicide Potential Scale
The 9-item Suicide Potential Scale is a clinician-completed scale assessing previous suicide attempts, recent psychological or psychiatric interventions, recent loss of a significant other, previous use of alcohol and illicit substances, symptoms of depression, current major problems, current suicidal ideation, and the existence of a suicide plan. Each item is scored as being absent, score 0; or present, score 1. A total score of 3 or more has been reported of having high sensitivity, 86%, and high specificity, 80% (although there is no detailed evaluation of the positive and negative predictive value of this instrument).
Beck Hopelessness Scale
The 20-item self-completed instrument explores the individual’s negative perception of the future across three domains: feelings about the future, loss of motivation, and expectancy. Scoring requires access to the scoring template, and a score of 8 or more is considered indicative of moderate risk of self-harm and suicide. Meta-analytic reviews of the instrument suggest high levels of sensitivity for both suicide and self-harm, 80% and 78%, respectively, but modest levels of specificity for suicide and self-harm, both 42%.
Final Thoughts
While a great deal of work has been undertaken to understand how to identify those at high risk of suicide and self-harm in the general population, there has been little work undertaken with offender populations. Where work has been conducted, it needs to be interpreted with caution. From a methodological perspective, the majority of research has focused on identifying constructs in populations who are known to be at risk and then comparing these to those who are not considered at risk, but one needs to be wary that past behaviors may not be predictive of future behavior. In addition, since self-harm and suicide are more prevalent in offender populations, needs assessment and care planning are critical aspects of health-care provision, and more research is needed on efficient tools to identify those at risk of suicide and self-harm as part of a comprehensive assessment package.
References:
- Hawton, K., Linsell, L., Adeniji, T., Sarialan, A., & Fazel, S. (2014). Self-harm in prisons in England and Wales: An epidemiological study of prevalence, risk factors, clustering and subsequent suicide. Lancet, 382, 1147–1154.
- Perry, A., Marandos, R., Coulton, S., & Johnson, M. (2010). Screening tools assessing risk of suicide and self-harm in adult offenders: A systematic review. International Journal of Offender Therapy and Comparative Criminology, 54(5), 803–828.