Abstract
We conducted a comprehensive umbrella review including all published meta-analyses and systematic reviews on suicide-related outcomes among adults in custody. Eligible studies were searched for in major scientific databases following the PRISMA method up to February 2023. The quality of each study was assessed using the AMSTAR-2 protocol. Twenty studies were included, consisting of 11 meta-analyses and nine systematic reviews. Factors associated with suicide outcomes among individuals in custody were categorized into five domains: clinical status, historical, custodial, criminological, and prevention. However, all systematic studies reported considerable heterogeneity (I2 > 50%–85%). Most systematic literature reviews and/or meta-analyses showed moderate quality based on AMSTAR-2. This umbrella review highlights individual factors (in the clinical status and historical domain), contextual factors (in the custodial and criminological domains), and protective factors (in the prevention domain) to provide a clearer understanding of the prison-related domains involved in the development and maintenance of suicide-related outcomes.
Introduction
Worldwide, it is estimated that more than 10.77 million people are held in correctional institutions, either as persons in pretrial custody pretrial or as convicted felons (Fair & Walmsley, 2021). The burden of social and health problems among individuals in detention is significant. Indeed, this population commonly exhibits a higher prevalence of psychiatric and substance use disorders (Wainwright & Dawson, 2022; Zavattini et al., 2017), as well as somatic diseases (Tverborgvik et al., 2023). A history of incarceration also increases the risk of death from internal causes such as cardiovascular disease, cancer, infectious disease, and liver disease (Spittal et al., 2019). Increased alcohol, tobacco and drug use, homelessness, higher levels of infections, and low socioeconomic status are possible explanations for the elevated risk of death from internal causes among released prison populations (Binswanger et al., 2007).
However, the most troubling and pervasive cause of nonnatural mortality within correctional facilities is self-inflicted violent behavior (Fazel & Baillargeon, 2011), posing serious social, economic, and public health problems. Suicide and self-harm, both suicidal and nonsuicidal, rank among the most frequent causes of mortality risk among adults in custody (Favril et al., 2022; Fazel et al., 2016). Specifically, suicide is defined as the death resulting from intentional self-harm with the intent to die, while nonsuicidal self-injury is defined as intentional self-harm without the intent to die. On the other hand, suicide attempt is defined as intentional self-harm with the intent to die that did not result in death. In some cases, the intent to die may not have been determined: undetermined suicide-related behavior.
The risk factors for the development of suicide-related outcomes in prison are multifaceted and attributable to various domains, including the individual’s life history, presence of childhood trauma, clinical status, crime-related factors, and type of custody. Recent studies (Perugino et al., 2023) have revealed a notably alarming rate of suicides in prisons, underscoring the urgent need for intervention in this phenomenon.
The most common forms of self-inflicted violence among incarcerated individuals are nonsuicidal self-injury and suicide. Notably, self-injurious behavior often precedes suicide, with approximately 50% of those who died by suicide in prison having a history of nonsuicidal self-injury, showing that for every suicide attempt, there was an 8.4 times greater chance of having a suicide (Fazel et al., 2008).
A 5-year epidemiological study of adults in custody in prisons in England and Wales (Hawton et al., 2014) found that cutting and scratching were the most frequent methods of self-harm in both sexes. Moreover, concerning risk factors and triggering events for the self-harm episode, this study found associations with life sentences or unsentences (e.g., for individuals who are in the process of appealing their conviction or awaiting sentencing for additional charges).
Regarding clinical status in relation to risk factors for suicide-related outcomes in prison, studies have indicated that comorbidity with mental illness is associated with significantly higher rates of suicidal ideation and behavior. Severe psychopathologies such as posttraumatic stress disorder and major depression (Barry et al., 2020; Gooding et al., 2017), as well as borderline personality disorder (Favril, Indig, et al., 2020), and other psychiatric conditions like psychosis, anxiety, and substance dependence, influence suicidality in the prison population (Rivlin et al., 2010).
Furthermore, a recent landmark meta-analytic study investigated self-harm behaviors in prison, providing valuable insights for implementing prevention approaches and strategies (Favril, Yu, et al., 2020). Factors associated with self-harm serve as predictors of suicidal ideation (such as, thoughts or fantasies about suicide, including considering or planning one’s own death by suicide) and suicide attempt (an intentional, self-inflicted action that is taken with the explicit purpose of ending one’s own life). Among these factors, psychological aspects play a significant role, serving as risk factors for suicide (Iskric et al., 2020). However, the prison environment, along with custodial and criminological factors, coupled with other psychological factors, can exacerbate the risk of suicide beyond the presence of psychopathology. Indeed, self-harming behavior, considered a reliable predictor of suicide risk, can occur for various reasons, including attempts to influence the environment and as a response to symptoms of mental illness (Jeglic et al., 2005).
Regarding the type of conviction, a study utilizing supranational data from the European Council’s annual criminal statistics suggested a significant association between suicide in prison and convictions for sexual offenses, sentences of more than 20 years, or life imprisonment. Notably, an interesting relationship appears to exist between charges for violent offenses (such as assault, battery, or murder) and rates of suicide in prison (Rabe, 2012). A recent literature review on the topic corroborated that suicide risk is highest during the early stages of incarceration and that assessing previous episodes of self-harm may predict suicidal ideation in new adults in prison (Gould et al., 2018). These findings highlight that within the prison context, the stage of incarceration and the type, presence, or absence of conviction can be identified as external risk factors related to the development of self-inflicted violent behaviors.
Specific elements of the prison environment can exacerbate the mechanisms underlying risk behaviors. Generally, characteristics such as overcrowding, high turnover of people in prison, and heightened security levels are associated with higher suicide rates and undermine the quality of relationships between staff, such as prison officers, and adults in custody, thereby contributing to increased self-harm and suicidal behaviors (van Ginneken et al., 2017). For instance, recent findings suggest that, relative to the custody domain, the type of sentence and prison regime may be risk factors for self-injurious violent behaviors. According to Boren and colleagues (2018), people just incarcerated for violent crimes exhibit a higher likelihood of completing suicide. This is attributed to past violence being established as a risk factor for self-inflicted violence among incarcerated individuals (Ammerman et al., 2015), and because this type of offense often carries a longer sentence, leading to increased isolation and reduced social support. Furthermore, maximum-security regimes can also be considered risk factors for self-inflicted violent behaviors compared with minimum- or medium-security regimes, both due to differences in monitoring and the conditions of close isolation and reduced opportunities for social contact (Boren et al., 2018).
It is important to underline that in most of the United States, there is a clear differentiation between “jails” (short-term detention facilities used to hold individuals who are awaiting trial, awaiting sentencing, or serving shorter sentences -usually less than a year-, typically operated by local governments) and “prisons” (long-term correctional facilities designed to hold individuals who have been convicted of crimes and sentenced to incarceration). However, this difference is not widespread in most European countries, and numerous systematic reviews that have examined variables relating to adults in custody have not taken this difference into consideration.
Overall, to date, several systematic reviews and meta-analyses have heterogeneously studied suicide-related outcomes among adults in custody in the prison setting. In light of these findings, a comprehensive understanding of individual and contextual risk and protective factors that may influence the risk of developing or maintaining suicide-related outcomes in prison can inform new treatment efforts to reduce morbidity and mortality in this high-risk population.
The purpose of this general review is to identify and systematically evaluate systematic reviews with or without meta-analysis that have examined the topic of suicide-related outcomes in the prison population.
Method
Search Strategy
This umbrella review summarizes the results of meta-analyses and systematic reviews examining suicide outcomes (including suicidal ideation, suicide attempts, and suicide risk) in the prison population. We adhered to the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines (Liberati et al., 2009). A comprehensive bibliographic search was conducted across PsycARTICLES, PsycINFO, PubMed, Scopus, Web of Science encompassing all available literature up to February 1, 2023. Our focus was on systematic literature reviews and meta-analyses exploring outcomes related to suicide and mental health within the context of detention.
The search strategy employed three key concepts (prison setting, suicide-related outcomes, review/meta-analysis) during the article identification phase for each database (see Supplemental Appendix A): “(Prison* OR Inmate* OR Jail* OR Felon* OR Detain* OR Custod* OR Remand OR Offender OR Institution OR Sentenced OR Gaol OR Correctional OR Incarcerated) AND (Suicid* OR Suicid* ideat* OR Suicid* thoughts OR Suicid* idea OR Suicid* thinking; Suicid* attemp* OR Self murder OR Self killing; Suicid* danger OR Suicid* behavior OR Suicid* tendencies OR Suicid* propensities) AND (Systematic review OR Meta-analysis OR meta-analy* OR meta-analysis OR pooled estimate OR pooled prevalence OR pooled effect).”
For databases such as PsycINFO, PsycARTICLES, MEDLINE, and PubMed, the search terms were applied to the title and abstract fields. However, for Web of Science and Scopus, no field codes were specified in search terms. During the screening phase, two authors (G.B. and S.B.) independently reviewed the titles and abstracts to assess their relevance to the topic. In cases where the content of the title or abstract was ambiguous, the full text was evaluated, and in instances of disagreement between the authors, a third author (PV) was consulted for resolution.
Eligibility Criteria
This umbrella review focuses on suicide outcomes within the context of incarceration, as presented in systematic literature reviews or meta-analyses where suicide-related outcomes are either the central topic or a significant component of a broader topic. No language restrictions were imposed on the selection of eligible studies. However, only English papers were found to be incorporated in the present umbrella study.
The main inclusion and exclusion criteria (see Table 1) pertain to the construct and type of data. Systematic literature reviews or meta-analyses that investigated at least one suicide-related outcome (such as suicide attempt, suicidal ideation, or suicide risk) among adults in custody were considered for inclusion. When a systematic review or meta-analysis examined multiple risk factors associated with suicide outcomes in prison, we extracted information related to clinical status, historical domain, custodial domain, clinical domain, and prevention domain.
Eligibility Criteria
Unsystematic literature reviews (such as narrative reviews) were excluded based on the type of data. In addition, systematic reviews or meta-analyses that did not investigate aspects associated with suicidal outcomes among persons in custody, as well as systematic reviews or meta-analyses that considered suicide-related outcomes but not in the prison setting (e.g., people released from captivity) were excluded based on construct.
Outcomes
The main outcomes considered pertain to factors that may be associated with the risk of suicide, defined as the likelihood or probability that an individual may engage in suicidal behaviors (such as suicide attempts or suicidal ideation) or in completing suicide. It should be noted that in different studies, this risk is assessed through various assessment procedures (such as evaluating the presence of suicide attempts or suicidal ideations or assessing suicidal risk by comparing the number of suicides among the prison population to a normative sample). From our umbrella review, it emerged that factors associated with the risk of suicide can be grouped into two broad categories: conditions related to the prison setting and suicide outcomes.
Condition Related to Prison Setting
We included meta-analyses and/or systematic reviews of the literature that examined suicide-related outcomes in the prison population. Due to the heterogeneity observed in the studies, we defined specific domains of interest: clinical status, historical domain, custodial domain, criminological domain, and prevention (for more details please see Supplemental Appendix C).
Suicide Outcomes
We separately considered all suicide outcomes as reported by the authors of the original studies: suicidal ideation (thinking about, considering, or planning suicide), suicidal attempts (potentially injurious self-inflicted behaviors with a nonfatal outcome and intention to die, current or past), and the general measure of suicidal risk (which includes both thoughts and behaviors related to suicide).
Data Extraction
The title and abstract of each systematic review and meta-analysis were screened for eligibility, following which the full texts were examined to determine compliance with inclusion and exclusion criteria. Data extraction was conducted by a single author (S.B.), with cross-verification by a second author (G.B.). In cases of discrepancy, a final decision was reached through consultation with a third author (PV).
A protocol was established to extract and code data across the following categories: (a) Author and year of publication; (b) Journal name; (c) Country; (d) Number of studies included in the systematic literature review/meta-analysis; (e) Number of subjects; (f) Subject characteristics (percentage male, mean age); (g) Variables related to detention; (h) Suicide-related outcomes; (i) Main results.
In addition, if quantitative synthesis was performed, data on effect size along with corresponding confidence interval, significance levels, and measures of heterogeneity (using the I-square index) were also extracted.
Quality Assessment
To conduct the quality assessment, two authors (S.B. and G.B.) independently evaluated the quality of each included systematic review or meta-analysis using the AMSTAR-2 protocol (Lu et al., 2020). This protocol comprises a list of 16 questions (for details, see Table 2) related to the methodology established for the review, inclusion/exclusion criteria, paper synthesis strategy, independent data selection and extraction, risk of bias, and appropriate methods for conducting the meta-analysis. Each question can be answered as “Yes,” “No,” or “Partially Yes” to assign an overall score. Depending on the score assigned to the critical items (Q2, Q4, Q7, Q9, Q11, Q13, Q15) and noncritical items, each systematic literature review and meta-analysis was rated as “high” overall if it had no or one noncritical weakness; “moderate” if it had more than one noncritical weakness; “low” if it had one critical flaw with or without noncritical weaknesses; and "critically low" if it had more than one critical flaw with or without noncritical weaknesses.
Quality Assessment of the AMSTAR2.
Note. Y = Yes; N = No; PY = Partial yes
Q1: Do the research issue and inclusion standard reviewed include component parts of PICO?
Q3: Do the authors describe the selection of inclusion research design?
Q5: Whether two authors independently screen articles?
Q6: Do two authors extract data independently?
Q8: Do the authors in detail describe the brought into studies?
Q10: Have the authors reported the financial sources for the research listed?
Q12: If conducted meta-analysis, does the author evaluate the potent influence of study bias on meta-analysis results or other comprehensive evidence?
Q14: Does the author offer a reasonable interpretation and discussion for the heterogenicity observed in the outcome?
Q16: Has the author make known any possible conflicts of interest, any funds they get to conduct the study?
Results
Search Results
A total of 381 systematic literature reviews and meta-analyses were identified from the databases. After removing 205 duplicates, we screened 176 citations by title and abstract excluding 123 citations that did not meet the eligibility criteria. Subsequently, we evaluated 53 full-texts and excluded 33 due to data type or construct investigated. Finally, after a thorough review of the full texts, the authors agreed that 20 met the eligibility criteria (see Supplemental Appendix B).
See Figure 1 for a detailed flowchart illustrating the inclusion and exclusion process.

Flowchart of the Literature Search and Evaluation
Features of Systematic Studies
The included systematic studies were published from 2008 to 2022 in several leading journals in the field, detailed in Supplemental Appendix B. Among the 20 systematic studies included, eleven were meta-analyses and nine were systematic literature reviews, primarily conducted in England (45%) with the remaining studies conducted in the United States, Canada, Australia, Africa, China, Italy, and Belgium. These studies were organized by domains related to the topics analyzed in the systematic literature reviews or meta-analyses on suicide outcomes among adults in custody.
The data extracted for each domain related to preexisting or contextual conditions exacerbated by the experience of incarceration. The following domains were highlighted:
Clinical status (11 contributions): These contributions explored the association or comorbidity between mental disorders and suicide-related outcomes. They encompassed studies analyzing both general psychiatric status and specific psychopathological clusters. In addition, contributions analyzing the effect of psychiatric treatment during incarceration, psychological distress, and psychological risk factors for suicide-related outcomes such as substance abuse and high impulsivity were included within the clinical status domain.
Historical domain (2 contributions): These contributions explored the association between individual history (e.g., presence of childhood trauma) and suicide-related outcomes in incarcerated individuals.
Custodial domain (6 contributions): These contributions investigated the association or effect of factors related to the type of custody, such as segregation or confinement, physical or sexual victimization, the presence of threats, lack of social support, absence of prison employment, absence of contact/visits, and exposure to self-harm.
Criminological domain (3 contributions): These contributions examined the association or effect of crime-related factors, such as length of sentence, life sentence, type of crime committed, previous incarceration, being a detainee, or on remand.
Prevention domain (6 contributions): These contributions focused on prevention programs and screening tools useful in preventing suicide-related outcomes in prison. For more details, refer to Supplemental Appendix C.
The systematic studies related to clinical status included between 4 and 28 primary studies on the topic and analyzed the association between the adults in custody’s clinical status and suicide-related outcomes in the prison setting. Specifically, the suicide-related outcomes investigated were suicidal behaviors and attempts (Facer-Irwin et al., 2019; Fanizza et al., 2019; Favril et al., 2022; Fazel et al. 2013; Sirdifield et al., 2009; Tsegaye et al, 2021), suicidal ideation (Casiano et al., 2013; Sirdifield et al. 2009), and suicide risk (Fazel et al., 2008, 2013; Skarupski et al., 2018; Zhong et al., 2021), associated with factors related to clinical status, including Posttraumatic Stress Disorder (Facer-Irwin et al., 2019; Zhong et al., 2021), Suicidal Ideation (Favril et al., 2022; Zhong et al., 2021), Psychiatric Treatment (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021), Previous suicide attempts (Favril et al., 2022; Zhong et al., 2021), Current psychological distress (Favril et al., 2022), Psychiatric diagnoses (Favril et al., 2022; Fazel et al., 2008, 2013; Sirdifield et al., 2009; Skarupski et al., 2018), Drug/Alcohol Dependence and Abuse (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021), Bipolar disorder (Fazel et al., 2013), Major Depression (Fazel et al., 2013; Tsegaye et al. 2021; Zhong et al., 2021), Schizophrenic spectrum (Fazel et al., 2013), and general clinical status (Casiano et al., 2013; Fanizza et al., 2019).
Historical domain studies (Angelakis et al., 2020; Favril et al., 2022) included between 4 to 16 primary studies on the topic and analyzed the relation between childhood trauma (sexual abuse, physical abuse, emotional abuse, neglect), life history related to suicide and self-injurious behavior, and outcomes related to suicide attempts over the life course among persons in custody.
Systematic studies on the custodial domain included between 3 and 21 primary studies on the topic. They investigated the impact of segregation (Brown, 2020) and confinement (Luigi et al., 2020), threats of violence and physical and sexual victimization, poor social support, disciplinary infractions, absence of prison employment, single cell accommodation (Fazel et al., 2008), and exposure to self-harm (Favril et al., 2022), as well as custodial status in general (Livanou et al., 2019), in relation to several suicide-related outcomes such as current (Brown, 2020) and past (Livanou et al., 2019) suicidal ideation and attempts, and suicide risk (Favril et al., 2022; Fazel et al., 2008; Luigi et al., 2020; Zhong et al., 2021).
Systematic studies related to the criminological domain involved 4 to 30 primary studies on the topic. The variables related to the domain investigated were sentencing period (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021), type of crime such as violent crime, murder, burglary, robbery, or theft offense, drug offense (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021), previous incarceration (Favril et al., 2022; Zhong et al., 2021), remand status (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021) in association with suicide risk.
Finally, systematic studies that analyzed prevention programs and tools included between 6 and 38 primary studies. The programs investigated were prevention programs in correctional facilities (Stijelja & Mishara, 2022; Winicov, 2019), management programs (Barker et al., 2014), and reduction programs (Carter et al., 2022) for suicidal behaviors. In addition, two contributions analyzed the effectiveness of screening tools for suicide-related outcomes (Casiano et al., 2013; Gould et al., 2018). These prevention programs/tools were investigated in association with suicide risk (Barker et al., 2014; Carter et al., 2022; Gould et al., 2018), suicide attempt (Carter et al., 2022; Stijelja & Mishara, 2022), and suicidal ideation (Carter et al., 2022; Casiano et al., 2013).
Quality Assessment
The methodological quality assessed through AMSTAR-2 is presented in Table 2. Five systematic reviews were rated as “high,” ten as “moderate,” and five as “low.” The methodological quality of the studies, following the AMSTAR-2 protocol, is displayed in Table 2 as well. Overall, the most prevalent weaknesses in the areas of risk of bias included independently performed article selection and data extraction, detailed and comprehensive description of included and excluded studies, assessment of bias risk, evaluation of publication bias influence on results obtained, and discussion of observed heterogeneity in the findings.
Risk Factors Associated With Clinical Status and Suicide Outcomes Among Incarcerated Individuals in Prison
Eleven contributions (55%) focused on studying the clinical status of adults in custody associated with suicide outcomes in prison (Casiano et al., 2013; Facer-Irwin et al., 2019; Fanizza et al., 2019; Favril et al., 2022; Favril, Indig, et al., 2020; Fazel et al., 2008, 2013; Sirdifield et al., 2009; Skarupski et al., 2018; Tsegaye et al., 2021; Zhong et al., 2021).
Three contributions reported suicide attempt as the main outcome (Fanizza et al., 2019; Favril et al., 2022; Tsegaye et al., 2021); two contributions focused on suicidal ideation (Casiano et al., 2013; Favril, Yu, et al., 2020), three contributions focused on suicide risk (Fazel et al., 2008; Skarupski et al., 2018; Zhong et al., 2021), and the remaining contributions analyzed mixed outcomes related to suicidal behavior/ideas and suicide risk (Facer-Irwin et al., 2019; Fazel et al., 2013; Sirdifield et al., 2009).
Regarding suicide attempts among people in prison, the results show that clinical factors are closely associated with suicide attempts in prison. Indeed, results report a higher prevalence of suicide in individuals in custody with clinical psychiatric diagnoses than in nonclinical incarcerated population (Fanizza et al., 2019). A positive and significant association was observed between the onset of depression among adults in custody and a history of suicide attempts (Tsegaye et al., 2021).
Furthermore, recent suicidal thoughts appear to be a cross-sectional risk factor together with a previous attempt or history of self-harm and parasuicidal behavior. Similarly, a large effect has also been observed in those with a current or historical psychiatric diagnosis in prison (Favril et al., 2022).
Regarding suicidal ideation, contributions have shown high rates of suicidal ideation (lifetime or current) among adults in custody associated with several mental disorders, including borderline personality disorder, affective disorders, substance use disorder, posttraumatic stress disorder, social phobia, anxiety, and attention deficit hyperactivity disorder (Casiano et al., 2013). Self-harm is also considered among the clinical risk factors closely associated with the development of lifetime and current suicidal ideation (Favril, Yu, et al., 2020).
Furthermore, concerning suicide risk, contributions have shown that clinical factors related to thoughts of suicide, a history of suicide attempts, having a psychiatric diagnosis, being on psychotropic medication, and having problems with alcohol abuse are closely associated with suicide risk among individuals in prison (Fazel et al., 2008). In addition, anxiety and depression also emerge as clinical factors associated with suicide risk (Skarupski et al., 2018). In fact, the recent contribution of Zhong et al. (2021) shows that, in the clinical domain, suicidal ideation while in prison, having a history of past suicide attempts or self-harm, and being prescribed psychotropic medication are factors associated with suicide risk.
In addition, risk factors for suicide related to individual disorders have been highlighted, including a current psychiatric diagnosis (Sirdifield et al., 2009), depression diagnosis, and alcohol abuse (Zhong et al., 2021).
Finally, several contributions have reported associations between clinical status and mixed suicide-related outcomes (e.g., including suicide attempts, suicidal ideation, measures of suicide risk, or self-injurious behavior). For instance, posttraumatic stress disorder reports a significant and positive association with suicidal and parasuicidal outcomes (Facer-Irwin et al., 2019). Other findings concern the association between suicidal outcomes and mood disorders (bipolar and major depressive disorder), as well as the spectrum of schizophrenic disorders associated with increased suicide risk among adults in custody (Fazel et al., 2013).
Risk Factors Associated With Historical Domain and Outcomes Related to Suicide Among Incarcerated Individuals in Prison
Two contributions (10%) reported data regarding the historical domain associated with outcomes related to suicide in prison (Angelakis et al., 2020; Favril et al., 2022). Both contributions analyzed aspects related to different forms of abuse: sexual, physical, emotional, neglect (Angelakis et al., 2020; Favril et al., 2022). In addition, one contribution also included family history of suicide and self-harm, as well as out-of-home care, in the historical domain (Favril et al., 2022). Regarding the outcome related to suicide attempt among adult in prison, studies that investigated the association with the historical domain found a strong relation between childhood maltreatment and suicide attempts in prison. In particular, the likelihood of committing suicide attempts was positively associated with a childhood history characterized by sexual, emotional, and physical abuse, as well as emotional and physical neglect (Angelakis et al., 2020).
The study by Favril et al. (2022) confirmed that traumatic life history events, such as sexual, physical, and emotional abuse, are relevant risk factors for suicide attempts in prison, as well as having a family history of suicide.
Risk Factors Associated With Custodial Domain and Outcomes Related to Suicide Among Incarcerated Individuals in Prison
Six contributions (30%) investigated the relationship between the custodial domain and suicide-related outcomes among adults in custody (Brown, 2020; Favril et al., 2022; Fazel et al., 2008; Livanou et al., 2019; Luigi et al., 2020; Zhong et al., 2021). Most studies reported suicide attempt as the main outcome (Brown, 2020; Favril et al., 2022; Livanou et al., 2019), while others investigated outcomes related to suicide risk (Fazel et al., 2008; Zhong et al., 2021) and suicidal ideation (Brown, 2020).
Regarding suicide attempts among incarcerated individuals, studies have reported that segregation may shape up as a risk factor for both suicidal behaviors in prison and the development of suicide-related ideation (Brown, 2020). In addition, solitary confinement and single-cell accommodation have been identified as factors with the strongest relations to suicide attempts in prison, followed by physical or sexual victimization. Poor social support during incarceration, threats of violence, disciplinary infractions, absence of prison labor, and lack of social contact and visitation are factors moderately associated with suicide attempts among adult in prison (Favril et al., 2022). The effects of solitary confinement are also confirmed regarding the outcome related to suicide risk. In fact, results have shown that solitary confinement, single-cell accommodation, and not having access to social visits are institutional custodial factors that increase the risk of developing suicide-related thoughts and behaviors (Luigi et al., 2020; Zhong et al., 2021).
Risk Factors Associated With the Criminological Domain and Outcomes Related to Suicide Among Incarcerated Individuals in Prison
Three contributions (15%) reported findings about the criminological domain and suicide-related outcomes among adults in custody (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021). Two contributions investigated suicide risk as a suicide outcome (Fazel et al., 2008; Zhong et al., 2021), while Favril et al. (2022) analyzed the relationship between crime-related domain aspects and suicide attempts among adults in custody.
With regard to suicidal behavior, aspects related to violent crimes and a previous incarceration showed positively related association with a 60% increase in the odds of suicide attempts among incarcerated individuals. In terms of length of the sentence, serving a sentence of 5 years and more, compared with less than 5 years, and a life sentence, compared with determinate sentences, significantly increased the likelihood of committing suicide attempts in prison (Favril et al., 2022).
In addition, criminological factors also seem to have a strong impact on suicide risk. A sentence length of 18 months or more, the crime of manslaughter, and violent crimes in general have shown a positively related association with suicide risk. Despite significant heterogeneity among primary studies, increased sentence length and life sentences are factors associated with a significant increase in suicide risk (Fazel et al., 2008).
Consistently, the systematic study by Zhong et al. (2021) also showed that criminological factors appear to be strongly associated with suicide risk among adults in custody, particularly being in prison or in pretrial detention, serving a life sentence. In contrast, conviction for a drug-related offense showed an inverse relation with suicide risk compared h being in prison or on remand. The offense category related to conviction for criminal homicide and sex offenses was also associated with increased suicide risk. In general, violent crimes have been associated with suicide risk among individuals in custody; however, primary studies show considerable heterogeneity (Zhong et al., 2021).
Screening/Prevention Programs and Tools for Outcomes Related to Suicide in Prison
Six contributions (30%) investigated preventive programs and screening tools regarding outcomes related to suicide in prison (Barker et al., 2014; Carter et al., 2022; Casiano et al., 2013; Gould et al., 2018; Stijelja & Mishara, 2022; Winicov, 2019).
Regarding suicide risk-related screening at the time of entry, the systematic study by Casiano et al. (2013) states that it hovers around 84% among U.S. facilities that reported screening at-risk individuals (80%) and that these facilities differed in the level of suicide training by identifying that less than 38% of facilities where an incarcerated individual suicide occurred provided annual prevention training to direct care staff. In addition, about half of the primary studies showed that screening was not conducted by either mental health professionals or trained counselors, and that levels of agreement on suicide risk by staff were weak, underscoring a general ineffectiveness in screening systems useful for suicide risk prevention programs in prisons.
Consistent with the levels of screening related to suicide risk found, the instruments that should be considered for screening are also found to be scarce and not very robust. Of the most widely used instruments, the Suicide Risk Assessment Scale and the Depression, Hopelessness and Suicide Scale show the most promise in terms of ease of implementation and prediction of suicide (Gould et al., 2018).
In contrast, the results of systematic studies investigating prevention programs to reduce suicide-related outcomes in prison have shown the effectiveness of multifactorial programs. These programs include screening and assessment of adults in custody upon entry, improved staff training, post-entry observation for suicide risk, psychological monitoring and treatment of at-risk incarcerated individuals, limited use of solitary confinement, and increased social support (Barker et al., 2014).
The systematic review by Carter et al. (2022) examined the effectiveness of interventions for reducing suicidal thoughts and behaviors in individuals who have had contact with the criminal justice system. The results provided evidence for the feasibility of some approaches for preventing suicide-related outcomes among adults in custody. In particular, the results recommended several primary and secondary prevention interventions for self-inflicted violent behaviors in prison, such as group treatment, peer support, tailored treatment, and multicomponent treatment.
Consistently, the meta-analysis by Stijelja and Mishara (2022) pointed out that most primary studies proposed multicomponent suicide prevention strategies applied through adults in custody screening, staff training in cardiopulmonary resuscitation and crisis intervention, supervision of high-risk incarcerated individuals, adequate communication between prison staff and individuals in custody, debriefing of staff, improved clinical and suicide review procedures, limited access to media, and provision of mental health care and support for adults in custody.
In addition, one of the most popular psychotherapeutic approaches can be found in cognitive and dialectical behavioral therapy (CBT, DBT) to reduce the frequency and severity of suicidal behavior. Topics of therapy and psychoeducation include identifying triggers, recognizing resulting thoughts and feelings, identifying thinking distortions, modifying distorted thoughts, and adopting new, more rational thinking and coping skills.
In addition, peer prevention programs, in which fellow incarcerated individuals are trained on topics such as effective and active listening, suicide prevention and intervention, nonverbal communication, and the nature of mental illness, have also been found to be effective (Winicov, 2019).
Discussion
This umbrella review focused on the study of suicide-related outcomes among adults in custody in the prison setting and included 20 studies, 11 meta-analyses and nine systematic reviews of the literature.
The included studies were organized into different domains to focus on the analyzed topics: (a) clinical status, (b) historical, (c) custodial, (d) criminological, and (e) prevention programs/screening. The contributions that analyzed the association or prevalence of suicide-related outcomes among incarcerated individuals and clinical status (Casiano et al., 2013; Facer-Irwin et al., 2019; Fanizza et al., 2019; Favril et al., 2022; Favril, Yu, et al., 2020; Fazel et al., 2008, 2013; Sirdifield et al., 2009; Skarupski et al., 2018; Tsegaye et al., 2021) have shown that within prison, psychopathological onset of mental illnesses such as major depression acts as risk factors for suicide attempts (Tsegaye et al., 2021). In addition, self-injurious behaviors and thoughts related to suicide have been found to be strongly associated with the development of suicide-related behaviors (Favril et al., 2022). With regard to suicidal ideation, the ideation-action perspective of suicide (Klonsky et al., 2018; Klonsky & May, 2015; O’Connor & Kirtley, 2018; Van Orden et al., 2010) posits that the development of suicidal thoughts and the transition from ideation to behavior are distinct processes preceded by different risk factors.
In recent years, this paradigm has been adopted to study suicide risk among the incarcerated individuals (Favril & O’Connor, 2021; Favril, Yu, et al., 2020; Larney et al., 2012; Stoliker, 2021; Stoliker et al., 2021). These studies have reported preliminary evidence suggesting that factors related to suicidal ideation may be different from those regulating the transition to the suicidal act (Stoliker & Abderhalden, 2023). Consistent with this evidence, the present umbrella review also collected findings from recent systematic studies that emphasize this distinction, even though most of them do not take it into account. In fact, regarding the association between clinical status and the development of suicidal thoughts, significant associations were found with several mental disorders such as borderline personality disorder, affective disorders, substance use disorder, and posttraumatic stress disorder (Casiano et al., 2013). We can interpret these findings by considering that certain factors related to the development and/or maintenance of suicide-related thoughts and behaviors may be exacerbated within the prison context, particularly among adults in custody with a history of childhood maltreatment, specifically sexual and emotional abuse (Angelakis et al., 2020), and familiarity with suicide (Favril et al., 2022). In fact, the prison environment and the custodial experience may emerge as additional risk factors influencing the development of suicide-related outcomes (Boren et al., 2018). Among the custody-related aspects, several contributions included in this umbrella review (Brown, 2020; Favril et al., 2022; Fazel et al., 2008; Livanou et al., 2019; Luigi et al., 2020; Zhong et al., 2021) have pointed out that conditions associated with solitary confinement-such as the segregation regime, single-cell accommodation, and the scarcity or absence of social contact-can result in suicidal behavior (Luigi et al., 2020; Zhong et al., 2021). Indeed, the condition of solitary confinement and the resulting feeling of loneliness, both from the outside world and within the prison, can have a determining effect on the development of suicide-related outcomes. Furthermore, additional contributions have investigated the association between the criminological domain (Favril et al., 2022; Fazel et al., 2008; Zhong et al., 2021) and suicide-related outcomes, highlighting that the length of the sentence-particularly life sentences-is significantly associated with the occurrence of suicidal behaviors among individuals in prison. In line with this evidence, the study by Stoliker and colleagues (2021) highlighted the factor of hopelessness. Especially within the prison population, which may perceive incarceration as an insurmountable barrier to a better future, the presence of hopelessness may increase the likelihood of developing self-inflicted violent behaviors (Stoliker et al., 2021). Linked to these factors, including the conditions of solitary confinement and the length of the sentence, the Motivational-Volitional Integrated Model of Suicidal Behavior (O’Connor, 2011) posits that hopelessness arises from feelings entrapment—whether external or internal—that may play a crucial role, particularly in the development of suicidal ideation (Höller et al., 2022).
In the prison context, external factors related to the so-called “pain of imprisonment,” such as overcrowding, the level of prison security, and social disconnection, contribute to mental conditions of psychic entrapment. These conditions develop through perceptions of defeat and humiliation, defining themselves as premotivational factors that can lead to the development of self-injurious and suicidal ideas and intentions (Stoliker, 2018).
Moreover, the type of crime—particularly violent crimes, including homicide—is also significantly associated with the risk of developing suicide-related thoughts and behaviors (Zhong et al., 2021). In addition to findings concerning individual and contextual risk factors associated with the development of suicide-related outcomes, the present umbrella review also encompassed systematic studies addressing screening tools and suicide risk reduction programs within the prison setting, focusing on prevention (Barker et al., 2014; Carter et al., 2022; Casiano et al., 2013; Gould et al., 2018; Stijelja & Mishara, 2022; Winicov, 2019). These contributions have highlighted a general difficulty related to the timely screening of incarcerated individuals at risk of suicide (Casiano et al., 2013), associated with a lack of specialized training and a shortage of mental health professionals who can promptly intervene in cases of moderate or high risk.
Moreover, the instruments used also appear to lack robustness, being difficult to implement or not very versatile, and therefore challenging to adapt in various prison settings. In fact, they often fail to reflect the specific aspects of prison life, containing potentially defective or unclear questions. In addition, they take too long to complete, are impractical at intake, may be administered too long after entry into the prison, and have been validated with insufficient sample sizes. It is argued that many of these screening instruments lack sufficient sensitivity to detect a high percentage of individuals at risk. As a result, the generalizability of these instruments across multiple jurisdictions is unproven (Gould et al., 2018). Instead, during development, programs that aim to provide support and monitoring for incarcerated individuals at risk of suicide through the training of prison staff can help prevent suicide and mitigate the negative effects of solitary confinement (Barker et al., 2014). Systematic studies have revealed that the most effective preventive interventions were those that provided more accurate intake screening (Barker et al., 2014). However, evidence of effective interventions to reduce the rate of self-inflicted violent behaviors in individuals released from prison remains a critical gap in the literature (Carter et al., 2022). Among the approaches analyzed, the multicomponent or multifactorial approach appears to have the highest level of effectiveness in reducing risk behaviors. This aligns with the understanding of suicide as a multifaceted phenomenon. Therefore, prevention programs aimed at reducing suicide-related behaviors in prison should address multiple risk factors by incorporating several strategies with solid evidence of effectiveness (Stijelja & Mishara, 2022). Despite the small number of studies eligible for inclusion, we believe that this umbrella review has evaluated systematic and organized studies related to suicide outcomes among the incarcerated population. However, limitations are present; indeed, the limited number of contributions for each domain did not allow us to reanalyze the data, thereby preventing control for heterogeneity among studies. In addition, few primary studies adopted the ideation-to-action perspective (Klonsky et al., 2018), which limited our ability to systematize and discuss the results by distinguishing the identification of risk/protective factors related to specific elements that can precipitate the transition from thinking to acting. This hindered our ability to outline the points at which to interrupt this risk trajectory.
Conclusion
This umbrella review confirms that the incarcerated population is at high risk of developing suicide-related outcomes. It highlights individual factors (clinical status, historical domain), contextual factors (custodial, criminological domains), and protective factors (prevention program) that need consideration in prison intervention programs. However, to draw more definitive conclusions, further primary studies need to be conducted. These studies should consider the interdependent effect of variables associated with the analyzed domains for assessing suicide risk among adult in custody. In addition, adopting an empirical-theoretical perspective that considers the development of suicidal thoughts and the transition from ideation to behavior as distinct processes, preceded by different risk factors, would enhance our understanding.
Supplemental Material
sj-docx-1-cjb-10.1177_00938548241260806 – Supplemental material for Suicide-Related Outcomes Among Adults in Custody: An Umbrella Review of Systematic Literature Reviews
Supplemental material, sj-docx-1-cjb-10.1177_00938548241260806 for Suicide-Related Outcomes Among Adults in Custody: An Umbrella Review of Systematic Literature Reviews by Serena Bruno, Silvia Andreassi, Giulia Ballarotto, Valeria Carola, Silvia Cimino, Giacomo Ciocca, Barbara Cordella, Michela Di Trani, Federica Galli, Carlo Lai, Viviana Langher, Erika Limoncin, Manuela Tomai and Patrizia Velotti in Criminal Justice and Behavior
Footnotes
Supplemental Material
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
