A Systematic Review of Risk Factors Implicated in the Suicide of Police Officers

Suicide has long been considered as nearing ‘epidemic levels’ in law enforcement populations. Nevertheless, despite the argued scale of the problem, no review has yet systematically examined the evidence base to elucidate the risk factors or predictors implicated in the suicidal behaviours of police officers. The current review aims to do this, by considering a final sample of 20 papers that met inclusion criteria. Findings from this qualitative review revealed five superordinate risk factors (i.e., problematic substance use close to, or at the time of death; presence of depression and previous suicide attempts; differences in trauma response; exposure to excessive and prolonged job-related stress, including dissatisfaction; absence of a stable intimate relationship), which when taken in isolation each incrementally contributed to suicide risk, but when found to be comorbid, appeared to markedly increase the likelihood of completed suicide. Implications for suicide prevention, policy design, and treatment formulation are discussed, along with limitations and directions for future research.


Introduction
Mental health problems play a significant role in the noncommunicable disease burden estimated to affect approximately 300 million people worldwide (Institute for Health Metrics and Evaluation 2017). Of this broader class of 'mental health problems', suicidal behaviours and/or ideation pose a considerable concern for general populations (Di Nota et al. 2020), with more than 700,000 deaths attributed to completed suicide annually (WHO Report 2021). Despite the scale of this issue, in small communities and closed occupational groups (e.g. law enforcement, emergency care workers, medical practitioners, etc.), the immediate and long-term impacts of suicide are arguably exacerbated given the depressive 'ripple' effect that may eventuate as a product of witnessing first-hand trauma (Barron 2010).
Much has been theorised and written about the incidence and prevalence of suicide in law enforcement over the past five decades (Violanti et al. 1998). Data from the most recent Office for National Statistics (UK) bulletin on suicides by occupation show that there has been a total of 169 serving police officers who have committed suicide between 2011 and 2019, with this averaging to approximately 21 deaths per year in the UK (ONS 2019). Given that elevated prevalence estimates as compared to normal populations have been consistently observed across time periods and jurisdictions (Violanti 1995a, b), some researchers have described suicide in law enforcement populations as an 'epidemic' (Loo 2001).
Despite these strong assertions, other authors have cast doubt over suicide classification as the leading cause of death among law enforcement officers (Chopko et al. 2014). Methodological problems in the assessment and collection of suicide statistics, including inadequate sample sizes, limited measurements of specific police populations, regional variations, and inconsistent reporting periods, make accurate estimation of the true problem increasingly difficult (Barron 2010;Chopko et al. 2014). Irrespective of whether police officers experience higher rates of suicide compared to the general population (Violanti 1995a, b), police agencies are typically drawn from populations where mental and physical illness are minimal (i.e., psychological and medically tested), officers are employed and engaged in meaningful occupational work, and afforded adequate support via training, health benefits, and counselling (Barron 2010). It therefore stands that rates equivalent to the general population mortality figure, which considers both the young and old, those with mental and physical illnesses, or who are unemployed, uneducated, or from lower socio-economic backgrounds (Barron 2010;Violanti 2001), warrants significant concern.
Consequently, researchers have shifted away from quantifying rates of suicide in favour of investigating the risk factors and predictors that underpin suicide in this occupational group (Barron 2010;Berg et al. 2003). While general expert consensus holds that suicide causation is multidimensional (Barron 2010), the literature reports three particular classes or levels of stressors thought to be implicated in the completed suicide of law enforcement personnel: (1) personal or individual stressors, (2) occupational factors, and (3) organisational issues (Hart et al. 1995). Personal or individual stressors refer to factors internal to the police officer such as mental or substance use disorders (i.e., depression; Kessler et al. 1999), suicidal ideation (Nugent and Williams 2001), relationship and financial problems (Barron 2010), and susceptibility to post-traumatic stress/disorder(s) (Carlier et al. 2000). Occupational factors encompass those demands considered to be 'part of the job' such as completing shift work, engaging with victims and perpetrators of crime, and interacting with the justice system (Hart et al. 1995). Organisational issues comprise concerns around reduced support, bureaucratic red tape, and lack of career advancement opportunities (Barron 2010). These classes of stressors upon reaching a critical mass are thought to decrease an officer's suicidal threshold, thereby increasing the likelihood of suicidal behaviour (Barron 2010).
More recently, Di Nota et al. (2020), in an effort to move beyond the classification of stressors towards a psychological model that delineates the causal pathways implicated in police suicidal behaviours, argue that increased suicidal risk may be partially explained by both the Interpersonal-Psychological Theory of Suicide (Joiner 2005) and Klonsky and May's (2015) Three-Step Theory. Incorporating this, Di Nota et al. (2020) asserts that progressive suicidal behaviours develop via a tripartite process: (1) suicidal ideation emerges from feelings of pain (e.g. psychological or physical) and hopelessness often triggered by exposure to potentially psychologically traumatic events, (2) the presence of pain and hopelessness coupled with low or absent connectedness exacerbates suicidal ideation leading to suicide planning, and (3) suicide attempt/s are carried out where the officer possesses the necessary dispositional or genetic features, has become habituated to pain, injury fear, and/or death, and has knowledge and access to lethal means. In short, suicidal behaviours are likely to occur where a police officer with suicidal ideation has the impetus and capacity to enact a plan (Di Nota et al. 2020), and therefore fitting with the aforenoted theories.
Given the lack of published systematic reviews concerning why police officers commit suicide, it was deemed timely to undertake a systematic assessment of the existing evidence base. Accordingly, this review aimed to provide a structured examination of the available literature concerning the following research question: What is empirically known regarding why police officers commit suicide? The aims of this systematic review were as follows: (1) to identify potential risk factors (i.e., precipitating factors, motives, and triggers) that influence suicide in police officers; (2) to identify any sex differences in the mode, methodology, or risk factors underpinning suicide in police officers; (3) to identify any differences in risk factors for suicide between sworn and unsworn police officers; and (4) to review any police suicide risk factor differences between developed and underdeveloped jurisdictions.

Method
Following PRISMA guidelines, publications that empirically examined risk factors contributing to suicide in police officers were identified (Moher et al. 2009). Four English language databases were considered: PsycInfo, PsycNet, PubMed, and Web of Science. Search terms were grounded in previous research and selected using the SPIDER search strategy protocol (Cooke et al. 2012). Consistent with the SPIDER protocol, the following was achieved: (1) identification of the relevant sample, phenomenon of interest, design, evaluation, and research type; (2) key terms highlighted; and (3) alternative phrases generated, to develop a coherent search strategy. The research team then reviewed the appropriateness of the search strategy and identified potential further terms. Final search terms included officer*, suicide, and factor*. Additional articles were obtained via manual searching of relevant publications and reference lists.

Inclusion and Exclusion Criteria
The SPIDER search strategy protocol was used to identify inclusion and exclusion criteria (Table 1). Studies were considered eligible if they (a) reported information concerning risk factors for completed suicide solely in civilian police officers, (b) presented findings derived from any empirical research method, and (c) were available in English. Time limits were placed on searches to allow a contemporary landscape of issues to be canvassed (e.g. January 1990 -present; Violanti et al. 2004).
Studies were excluded if they (a) did not report any information regarding risk factors for completed suicide in civilian police officers, (b) studied the defined outcome (i.e., completed suicide) in non-civilian police samples (e.g. military law enforcement) or in mixed samples (e.g. civilian police officers and firefighters), (c) did not contain primary data, (d) did not satisfy an empirical research methodology (i.e., conference paper, editorial, book chapter), and (e) were not available in English.

Eligibility Screening
Titles and abstracts were screened to determine if they met the inclusion criteria. If it was not possible to determine a paper's eligibility at this stage, it remained on the list and was examined during the full-text stage. To minimise the risk of bias during the eligibility screening stage, a random sample (n = 12) of articles examined for eligibility at the full-text stage were selected for quality assessment by a member outside of the research group. Agreement between the research group and external member, calculated using Cohen's kappa, was 1.00 (p < 0.05), which is considered very good.

Quality Assessment
Two quality assessment tools were selected to appraise the quantitative and qualitative studies included within this review: (1) The JBI Checklist for Analytical Cross-Sectional Studies (i.e., quantitative; Moola et al. 2020), which consists of eight questions designed to probe various quantitative methodological aspects (e.g. sample clearly defined, outcomes validly measured, appropriate statistical analyses), with each question rated as 'yes ', 'no', 'unclear', or 'not applicable', and item ratings used to guide the assessor's judgement regarding article inclusion or exclusion, and (2) The Critical Appraisal Skills Programme (CASP) qualitative checklist (QC), to appraise all qualitative studies that met the inclusion criteria (Critical Appraisal Skills Programme 2018). The tool consists of 10 questions that tap various methodological aspects (aims, methodology, research design, recruitment strategy, data collection, research-participant relationships, ethical issues, data analysis, findings and meaningful), which are given a rating of 'yes', 'can't tell', or 'no'. For the purposes of this review, item ratings for both the CASP and JBI were used to guide the assessor's judgement regarding article inclusion or exclusion, and overall risk of bias.
To minimise the risk of bias during the quality appraisal process, a random sample (n = 5 [10%]) of the included studies were selected for quality assessment by a member outside of the research group. Agreement using Cohen's kappa was 1.00 (p < 0.05) and thus very good.

Data Extraction and Analysis
A data extraction table was designed to collate information relevant to aims (1) to (4) of the systematic review. This included extracting data regarding sample descriptors (i.e., sample size, country, setting, officer rank), sex differences, differences between sworn and unsworn police officers, methodological study type and design, key findings, and main conclusions drawn.
Thematic analysis was used to examine aims (1) to (4). Using Braun and Clarke's (2006) six-step guide to thematic analysis, once familiar with each article, information was organised into codes based on features within the data. This allowed for a deeper understanding of the information and identification of initial patterns in relation to the research aims. Themes were then generated based on commonalities and patterns among the codes. Proposed themes were reviewed in relation to the data and refined, and then named based on their content. Care was taken to ensure that themes were not overly complex and presented an internally consistent narrative that accurately represented the data.

Search Results
In total, 19,004 articles were identified, with manual searching identifying a further eight records. After duplicates were removed, 16,680 articles remained. The specific process of screening and retrieval is indicated in Fig. 1. The final sample included 20 articles, which were examined using the quality assessment tools. No papers were excluded after the quality appraisal stage.

Characteristics of the Included Studies
The majority of studies were conducted in Africa (9), in urban precincts (30%), using a cross-sectional study design (90%; see Table 2 for further information regarding study demographics). The 20 studies included 12,603 participants, with 9476 men and 3042 women, and 85 sex not specified. Of the studies that reported age parameters, participant's age ranged between 18 and 69, with a mean age of 36.45(SD = 3.48). Studies mostly examined risk factors in sworn junior (i.e. corporal, constable, patrol duty; 6031 officers) police officer populations (55%). All 20 articles were considered low risk of bias (see Table 2 for paper demographic and Table 3 for papers included and quality assessment).  Research aim 1: to identify any risk factors (i.e., precipitating factors, motives, and/or triggers) that influence suicide in police officers.
Five superordinate themes were identified via thematic analysis, as follows: Theme 1: problematic substance use close to, or at the time of death (40%; eight articles).
Consensus emerged that substance misuse was related to both suicidal ideation and completed suicide (Di Nota et al. 2020;Meyer et al. 2003;Violanti 2004). In their psychological autopsies, Barron (2010) and Rouse et al. (2015) found that problematic drinking was prevalent for police officers who died by suicide. Police officers were often under the influence of alcohol or other substances close to, or at the time of death, with officers having increased their alcohol intake prior to suicide. Problematic drinking behaviour was also related to coping mechanisms, with rugged individualism considered as stereotypical for police officers, which increased problematic drinking behaviour (Beehr et al. 1995). The more alcohol that was consumed per week, the higher the risk of suicidal ideation, with officers consuming higher levels of alcohol less able to cope and seek support (Pienaar and Rothmann 2005;Rothmann and van Rensburg 2002).
Being diagnosed with, or having traits of depression, was considered a common risk factor for being a vulnerable police officer with an increased risk of suicidal ideation (Berg et al. 2003;Chopko et al. 2014;Di Nota et al. 2020). Further, Violanti et al. (2009) suggested that there was a slightly higher incidence of suicidal ideation in women officers, as they reported more depressive symptoms than men. In the psychological autopsy, Barron (2010) found that 40% of the officers who completed suicide used antidepressants or other psychotropic medication, and 31.5% were diagnosed with depression. Many of these officers had communicated their suicidal intent to others. In addition, Njiro et al. (2021) reported that depression and low support were intertwined in predicting suicidal ideation. A small number of police officers had a history of previous suicide attempts, with this more prevalent for women (Berg et al. 2003;Di Nota et al. 2020). Officers who had attempted suicide previously scored higher on current suicidal ideation (Meyer et al. 2003;Pienaar and Rothmann 2005). Further, officers who completed suicide had a history of multiple suicide attempts (Rouse et al. 2015).
There appeared consensus that post-traumatic stress disorder (PTSD) influenced both suicidal ideation and completed suicide in police officers (Rouse et al. 2015;Steyn 2012;Steyn et al. 2013;Violanti 2004). Violanti (2004, p. 281) found that trauma resultant from "person's dying" or the "homicide of fellow officers" in conjunction with increased substance use "elevated the odds of suicidal ideation approximately ten times over those officers who had lower levels of trauma". More particularly, the included studies found that clusters of PTSD symptom criteria (i.e., hyperarousal, intrusive memories, etc.) predicted suicidal ideation differentially in men and women (Steyn 2012;Steyn et al. 2013). In two linked studies, Steyn (2012) and Steyn et al. (2013) found that for women police officers, "intrusive memories and to a lesser extent level of impairment influenced suicidal ideation" (Steyn 2012, p. 237), while among the men in the experimental group, hyperarousal was found to be "a key predictor of suicidal ideation" (Steyn 2012, p. 237). These divergent findings were theorised by the researchers as resulting from a women officers tending to ruminate more (Steyn 2012, p. 237), whereas for men problems focus on impulse control and the regulation of affect (Steyn 2012, p. 237).
Occupational stress refers to stressors faced in the course of police officers' duties, and which directly influences suicidal ideation and completed suicide. There was consensus that occupational stress was a product of burnout (subordinate theme 1) and job dissatisfaction (subordinate theme 2).
Subordinate theme 1 -burnout. Burnout refers to a state of emotional, physical, and mental exhaustion caused by excessive and prolonged job-related stress. Both Berg et al. (2003) and Burke and Mikkelsen (2007) found that burnout components of "emotional exhaustion and depersonalisation (or cynicism) predicted suicidal ideation" via increases in both depression and anxiety (Berg et al. 2003, p. 310). Using the Maslach Burnout Inventory -Human Service Survey (MBI-HSS; Maslach et al. 1996), Wray and Jarrett's (2019, Subordinate theme 2 -job dissatisfaction. Partly stemming from burnout, five independent studies found that a low sense of job satisfaction or job dissatisfaction showed statistically significant associations with suicidal ideation (Berg et al. 2003;Burke and Mikkelsen 2007;Meyer et al. 2003;Quarshie et al. 2021;Rothman and Strijdom 2002). More specifically, Quarshie et al. (2021) observed that officers who reported low levels of job satisfaction were at an increased risk of experiencing suicidal ideation during the previous 12 months. The researchers theorised that this finding should be expected given that "low job satisfaction may be a response to negative work experiences and unfavourable work characteristics, which can undermine the psychological health of police officers" (Quarshie et al. 2021, p. 198).
Two subthemes emerged, the absence of a committed relationship (subordinate theme 1: civil status) and problems within an existing relationship (subordinate theme 2: relationship problems).
Subordinate theme 1 -civil status. There was a consensus that officers who reported higher rates of suicidal ideation were either single, separated, or divorced (Berg et al. 2003;Burke and Mikkelsen 2007;Violanti et al. 2009). Thus, being married or in a committed relationship was a protective factor (Quarshie et al. 2021). However, Barron (2010) found that almost half or the police officers who committed suicide were married with children, suggesting that occupational relationships were poor, and thus, the personal relationships were not sufficiently protective.
Subordinate theme 2 -relationship problems. Many of the police officers who committed suicide experienced family conflicts (Barron 2010). This was suggested to indicate that difficulties in their personal relationships could result in the inability to cope with other difficulties (Berg et al. 2003). Beehr et al. (1995) reported that coping mechanisms were related to divorce potential, with rugged individualism increasing this potential. Additionally, some police officers were either the victim or alleged perpetrator of domestic abuse, which included the use of weapons or threat of weapon use (Rouse et al. 2015).
Research aim 2: to identify sex differences in the mode, methodology, or risk factors underpinning completed suicide in police officers.
The review comprised 9476 men and 3042 women. Of the 20 included articles, four studies (20%) specifically compared and contrasted suicide risk factors between sex. No sex differences were identified with regards to problematic substance misuse, the presence of depression and previous suicide attempts, or absence of a stable intimate relationship. As previously stated, and with regards to PTSD, women police officers showed a higher incidence and prevalence of intrusive memories, while men showed elevated levels of hyperarousal (Steyn 2012;Steyn et al. 2013). Women police officers also reported significantly higher levels of occupational stress and emotional exhaustion than men, with this sex disparity arguably a result of less peer support and supervision offered to women within the police force (Wray and Jarrett 2019).
Research aim 3: to identify any differences in risk factors for completed suicide between sworn and unsworn police officers.
There was a total of 9451 sworn officers, 350 unsworn, and 3717 unknowns. Although the different risk factors between unsworn and sworn officers were not detailed, there was evidence to suggest some differences. Sworn officers had higher levels of substance use and mental disorders (depression, PTSD, and stress), and civilians had a higher rate of PTSD (Di Nota et al. 2020). Therefore, most of the articles combined sworn and unsworn as a group, without distinguishing the differences between the groups. Thus, the results were more likely to be related to sworn officers.
Research aim 4: to review any police suicide risk factor differences between developed and underdeveloped jurisdictions.
In underdeveloped countries, educational levels for officers with high suicidal ideation were on average grade 12 or below. These officers also had lower ranks (Pienaar and Rothmann 2005). The wages in these countries were low (Steyn et al. 2013), which resulted in job dissatisfaction (Rothmann and Strijdom 2002). Thus, some police officers moonlighted, which resulted in increased suicidal ideation, which can be a result of emotional and physical exhaustion from working multiple jobs (Quarshie et al. 2021). Developed countries frequently have specific educational degrees for police officers, and the wages are higher. In developed countries, job dissatisfaction was related to anxiety and depression (Berg et al. 2003), and burnout from the job (Burke and Mikkelsen 2007) rather than low wages. Demographics are presented in Table 2.

Discussion
This systematic review is thought to represent the first known effort to qualitatively explore the factors implicated in the suicidal behaviours of law enforcement officers. Five superordinate risk factors were revealed: problematic substance use close to, or at the time of death; presence of depression and previous suicide attempts; differences in trauma response; exposure to excessive and prolonged job-related stress, including dissatisfaction; and absence of a stable intimate relationship). When co-occurring or comorbid, these factors appeared associated with an increased likelihood of suicidal behaviours.
Concordant with the presumption of risk factor primacy (Violanti 2003(Violanti , 2004, 40% of the included studies identified problematic substance use or substance misuse (i.e., alcohol and drugs) as both pervasive and directly linked to completed suicide. More specifically, the findings indicated an increasing trajectory of substance use in the days and hours preceding the suicidal event (Barron 2010;Rouse et al. 2015;Violanti 2004). Though concerning, such observations were not unexpected given that substance use has long been characterised as problematic among the police force (Richmond et al. 1998), with officers often regarded as 'symptomatic' drinkers, who use alcohol and other substances for the relief of psychological strains (Pienaar and Rothmann 2005). Indeed, it has been previously found that the organisational culture actually reinforces the use of substances as both a social and psychological device effective in the amelioration of job-related stress (Violanti 2003(Violanti , 2004. This clearly raises challenges in relation to reinforcing unhelpful perceptions concerning coping and representing a potential barrier to treatment responsivity or those who require interventional support. Mental health problems, more particularly depressive disorders and previous suicide attempts, were identified by the majority of policing studies (50%) as both comorbid with substance use disorders and predictive of later completed suicide (Barron 2010;Di Nota et al. 2020;Meyer et al. 2003;Violanti 2004). These findings represent expected risk predictors given that depression and suicidal ideation/attempts are considered incremental contributors to the overall risk burden in like occupational groups (e.g. emergency care workers, medical practitioners; Barron 2010;Berg et al. 2003). But while expected, rates of depression and suicidal ideation commensurate with non-working and institutionalised populations (Kessler et al. 1999) are curious given that police agencies should represent and consist of a healthy and psychologically tested working population (Violanti 2004). Equally interesting, across the aggregated samples, women officers reported higher mean depression scores compared to men (Violanti 2009). While some research has focussed on identifying causes of depression, most studies have been conducted with predominately male samples (Pierino 2007), and therefore have little probative value relevant to discerning the causal pathways implicated in the development of mood disorders in women officers.
While little support was found within the review for the argument that police officers were exposed to a singular traumatic critical incident which directly contributed to suicidality (Rouse et al. 2015), PTSD symptomatology when found to co-occur with substance use and mental health problems (i.e., depression and suicidal ideation) significantly increased overall suicide risk (Violanti 2004). Beyond elucidating the comorbidities that exist with other risk indicators and the resultant cumulative burden, two trauma specific findings emerged from this review. First, results showed that intrusive memories in women officers and hyperarousal in men were implicated in the later development of suicidal behaviours. Though this was argued by Steyn (2012) as stemming from women's tendency towards rumination and male problems with impulse control and affect regulation (Selye 1956), these arguments appear overly simplistic and should be interpreted cautiously. Second, it was found by Violanti (2004) that officers with increased police service showed less susceptibility to trauma-related stress and suicidality. Interestingly, this finding falls foul of the current and predominant position which holds that persons exposed to multiple traumatic episodes, as is commonly observed in veteran law enforcement officials, are more likely to exhibit PTSD symptomatology (Violanti 2004). Despite being potentially anomalous, Violanti's (2004) observations may also suggest experienced officers have developed more appropriate and varied coping mechanisms to adaptively and more effectively process work-related events, thereby decreasing overall symptomatologic risk. When considered holistically, these findings generally indicate that police officers face an initial period of trauma vulnerability before developing the necessary coping and cognitive skills that could act to insulate and protect the individual from further harm.
Apart from those risk predictors considered internal or specific to the constitutional makeup of the individual officer, occupational stressors, particularly burnout and job dissatisfaction, were found to significantly contribute to suicidal behaviours. Such results were not surprising given that both burnout and job dissatisfaction have been found to be positively related in other closed occupational groups (i.e., emergency care workers, medical practitioners; Stehman et al. 2019). Of note, findings from the review found that burnout components of emotional exhaustion and depersonalisation (or cynicism) only predicted suicidal tendencies via depression and anxiety (Berg et al. 2003;Burke and Mikkelsen 2007;Wray and Jarrett 2019). Thus, mood disorders seemingly mediate the relationship between wider occupational stressors and later suicidal behaviours. The findings serve to highlight both the incremental burden attributable to burnout and job dissatisfaction, and the comorbid relationships that exist across levels of stressors (i.e., interaction between personal and occupational risk predictors).
Contradictory findings also emerged with respect to the influence of relationship dynamics on overall suicidal burden. Indeed, while marriage or committed relationships were argued by some researchers to serve as a protective buffer against suicidal ideation and later behaviours (Berg et al. 2003;Burke and Mikkelsen 2007;Quarshie et al. 2021;Violanti et al. 2009), equally, problems within partnerships (e.g. divorce) were indicated by other authors as constituting a motive for suicidality (Barron 2010;Beehr et al. 1995;Rouse et al. 2015). These findings suggest that the mere act of being in a civil relationship is insufficient to protect against suicidal behaviours but rather the quality of the intimate partnership is determinative of the protective function afforded (Barron 2010).
Rather than elucidating a clear framework by which to understand the development of suicidal behaviours in policing cohorts, the findings showed that individual (i.e. substance misuse, depression and previous suicide attempts, PTSD, relationship dynamic) and occupational (i.e., burnout and job dissatisfaction) risk factors, when taken in concert or amalgamated, create a 'critical mass' sufficient to lower the suicide threshold of the individual officer, thereby providing adequate motivation to engage in the target suicidal behaviour. Accounting for individual factors, previous engagement in suicidal behaviour, negative experiences, and a developed capacity to engage in suicidal behaviour all share parallels with the Interpersonal-Psychological Theory of Suicide (Joiner 2005) and Klonsky and May's (2015) Three-Step Theory. It would also appear that risk indicators specifically attributable to the organisation, such as lack of support, shift work, or access to lethal means did not emerge within the data. This perhaps highlights a need to develop models of understanding that are workplace specific and can take advantage of more contemporary research, while also extending to consider protective factors. What is clear from this review is that no one risk predictor, in and of itself, was sufficient to cause suicidal behaviours, but rather the multiplicative effect of "immersion in a vocational culture that stigmatises help-seeking, aggrandises [substance] use, and exposes an already compromised individual to ongoing, cumulative, and extreme stress" (Rouse et al. 2015, p. 102) necessitates cause for suicidal concern. Implications There are several implications that can be drawn from the review. First, studies noted that modern policing agencies were characterised by an organisational culture that appears to aggrandise substance use, while limiting or minimising the importance of mental health support and general wellbeing (Rouse et al. 2015). This is particularly concerning given that findings elucidated a direct and strong relationship between comorbid substance use and mental health disorders, and later suicidality (Barron 2010;Di Nota et al. 2020;Meyer et al. 2003;Violanti 2004). For that reason, suicide prevention would be bolstered by an organisational cultural shift that prioritises and places greater importance on mental welfare over the use of rugged individualism and maladaptive coping mechanisms. Second, the review findings showed that organisational stressors, such as burnout and job dissatisfaction, clearly contributed to overall suicide risk (Berg et al. 2003;Burke and Mikkelsen 2007;Wray and Jarrett 2019). Hence, agencies would benefit from more advanced, comprehensive, and tailored internal guidelines that mitigate risk by promoting both meaningful occupational engagement and task manageability. Third, given that comorbid major mental disorders (i.e. depression, suicidal ideation, PTSD) and substance use principally contributed to later suicidal behaviours, interventions could consider integrated and hybrid modalities that address a range of challenges, which may include preparing an officer to engage in therapy by managing any feature that could present as a barrier to responsivity (e.g. alcohol use).

Limitations and Future Directions
The review is not without limitations. The searching of relevant databases and studies was potentially limited by publication and language biases. The heterogeneity of the included outcomes also precluded an extensive and direct comparison across all findings. In addition, while it was the intention of the researchers to comprehensively examine the evidence base to evidence any age, sex, and occupational (i.e., sworn versus unsworn) risk factor differences, only limited inferences may be drawn, given the disparities between sex and sworn/unsworn numbers. Finally, the majority of eligible studies employed a prospective, cross-sectional survey design, which apart from being fallible to subject biases (i.e. social desirability) offers limited predictive power in establishing cause and effect temporal relationships (Wang and Cheng 2020). The use of psychological autopsies in the studies of Barron (2010) and Rouse et al. (2015) are promising however, and further research would benefit from the improved adoption of such methodologies, given that it provides a more holistic and accurate outlook of those risk factors known to contribute to suicidal behaviours. It is also worth noting that no studies were identified that examined protective factors and which met the eligibility criteria for review. Future researchers are thus encouraged to study both risk and protective factors in tandem, as noted earlier.
Informed Consent Not applicable. This study comprised of a systematic literature review/analysis of existing research.

Conflict of Interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.