1 INTRODUCTION

Suicide is listed among the top 20 causes of death worldwide, accounting for 800 000 deaths each year (World Health Organization, 2019). In 2016, the global age-standardized suicide rate was 10.5 per 100 000 population. However, collective rates higher than 15 per 100 000 were recorded for countries from the South East Asian and Western Pacific regions.

Mental and substance use disorders substantially contribute to the global burden of disease, accounting for 7% of DALYs (Disability-adjusted life years = years of life lost, due to premature death + years lived with disability) and 19% of all years lived with disability (Rehm & Shield, 2019). Indeed, in 2016, approximately 16% of the world's population was affected by mental or substance use disorders, corresponding to more than one billion people. Despite a general decrease in the burden of all other diseases, since 1990, mental and substance use disorders showed a 4.3% increase in the age-standardized rates of DALYs. Depressive disorders (3627 per 100 000), anxiety disorders (3715 per 100 000), drug (844 per 100 000), and alcohol use disorders (1358 per 100 000) were among those with the highest prevalence and accounted for the highest levels of burden, calculated as DALYs. However, gender differences exist. Depressive, bipolar, anxiety, and eating disorders are more common in women, while men have higher rates in all other disorders. The only exception is schizophrenia with an almost identical prevalence in both sexes.

Despite this sizeable impact of mental and substance use disorders on health, their global burden is thought to be underestimated (Vigo, Thornicroft, & Atun, 2016). This is especially so when considering the additional global burden due to suicide. Ferrari et al. (2014) estimated that two-thirds of all suicide DALYs are attributable to mental and substance use disorders. This emphasizes the importance of early identification and treatment of mental and substance use disorders disorders in global suicide prevention.

An increased risk of suicide has been reported for psychiatric patients (Chesney, Goodwin, & Fazel, 2014; Nordentoft et al., 2013; Too et al., 2019), and an underlying psychiatric disorder is a major risk factor for completed suicide (Wasserman et al., 2012). In several parts of the world, an underlying psychiatric disorder is reported in up to 90% of people who die from suicide. While this proportion is considerably lower in low- and middle-income countries (Vijayakumar, 2004), in a meta-analysis by Knipe et al. (2019), a mood disorder was identified in 25% of people who died by suicide.

Despite this strong relationship, it should be noted that most people with psychiatric disorders do not die by suicide. The presence of a psychiatric illness is not a sufficient cause to explain suicidal behavior, yet it is one of the most important risk factors interplaying in the suicidal process (Wasserman & Sokolowski, 2016). To explain the relationship between psychiatric disorders and suicide, Mishara and Chagnon (2016) proposed a model in which different mechanisms are combined with several critical factors: the presence of a crisis situation; the availability of an acceptable means; a lack of help or social support; and poor coping skills. A first mechanism suggests that suicide and psychiatric disorders may share a common etiology, such as the genetic predisposition to depression and impulsivity or the experience of early negative life events. A recent review (Sokolowski & Wasserman, 2020) identified 106 suicidal behavior candidate genes with a high level of evidence: forty of which were considered specific for suicidal behavior as not implicated in the genetics of other neuropsychiatric disorders. In another suggested mechanism, some conditions, such as substance use disorders, are thought to develop as an attempt to avoid or diminish suicidal impulses. A third mechanism proposes that suicide can be the consequence of cognitive distortions that characterize mental disorders, such as depressive delusions, psychotic command hallucinations, or extreme hopelessness. Furthermore, living with a mental disorder may generate negative experiences, such as social exclusion, unemployment, relational problems, stigma, and discrimination, that in turn contribute to suicide. Finally, suicide can be considered as a iatrogenic result of inadequate, inappropriate or incomplete treatment of mental disorders.

2 METHODS

This article uses a narrative review method to discuss prevalence and risk factors for suicidal thoughts and behavior among psychiatric patients. A narrative review methodology is also used to examine evidence-based practices in suicide assessment and treatment.

Literature searches were conducted using PubMed and Google Scholar. Articles were included if they were published in English in peer-reviewed journals. Cross-sectional, cohort, case–control, or interventional studies were considered. Whenever available, large international surveys, systematic-review, and meta-analyses were preferentially included in the narrative review.

A first search was conducted in order to identify epidemiological studies, systematic reviews, and meta-analyses estimating the prevalence and the risk for suicide in different psychiatric disorders. Main psychiatric conditions associated with an increased suicide risk were therefore recognized, and additional literature searches were performed to identify specific risk and protective factors for suicide in these subgroups of psychiatric patients. “The European Psychiatric Association (EPA) guidance on suicide treatment and prevention” (Wasserman et al., 2012) was used as framework for the review of assessment and treatment interventions and literature searches were conducted to take in account progresses in this field.

3 EPIDEMIOLOGY OF SUICIDE AMONG PSYCHIATRIC PATIENTS

Psychiatric disorders pose an increased risk for premature mortality from both natural and unnatural causes of death (Harris & Barraclough, 1998; Lawrence, Kisely, & Pais, 2010). In a study conducted in Denmark, Finland, and Sweden, psychiatric patients showed two to threefold increased mortality due to diseases and medical conditions and an increased risk for mortality from external causes. Compared to the general population, this resulted in an approximately 15 years shorter life expectancy for females and 20 years shorter for males (Nordentoft et al., 2013). Considering suicide, an increased risk of premature mortality from nine- to 37-fold for men and from 13- to 77-fold for women was described, with the highest risk associated with affective disorders and the lowest with substance use.

A meta-review confirmed the increased risk for all-cause mortality in mental disorders and a substantial reduction of life expectancy (7–24 years) compared to the general population (Chesney et al., 2014). Considering suicide mortality, the highest risk was found in borderline personality disorder (standardized mortality ratio [SMR] = 45.1, 95% confidence interval [CI]: 29.0–61.3), depression (SMR = 19.7, 95% CI: 12.2–32.0), bipolar disorder (SMR = 17.1, 95% CI: 9.8–29.5), opioid use (SMR = 13.5, 95% CI: 10.5–17.2), and schizophrenia (SMR = 12.9, 95% CI: 0.7–174.3), as well as anorexia nervosa (SMR = 31, 95% CI: 21.0–44.0) and alcohol use disorder (SMR = 16.4, 95% CI: 10.7–25.2) in women.

A meta-analysis of thirteen record linkage studies, utilizing linked data of mental health service use and suicide deaths, reported an eightfold increased risk of suicide in individuals with mental disorders compared with those without such disorders (risk ratio [RR] = 7.5, 95% CI: 6.6–8.6) (Too et al., 2019). The risk was highest in psychotic disorders (RR = 13.2, 95% CI: 8.6–20.3), followed by mood disorders (RR = 12.3, 95% CI: 8.9–17.1) and personality disorders (RR = 8.1, 95% CI: 4.6–14.2). High risk was also found in substance use disorder (RRpooled = 4.4, 95% CI: 2.9–6.8) and anxiety disorders (RRpooled = 4.1, 95% CI: 2.4–6.9).

Comorbidity between different disorders is frequent and is associated with a high suicide risk (Cavanagh, Carson, Sharpe, & Lawrie, 2003). In a large U.S. survey, increased odds ratios for suicide attempts in individuals with multiple psychiatric disorders, compared with participants with no disorders were found (Nock, Hwang, Sampson, & Kessler, 2010). The odds ratios (ORs) were 3.7 for any one disorder, 6.8 for two, 12.1 for three and up through 29.0 for six or more disorders. In a Swedish cohort study, the long-term risk of suicide was highest in men with comorbid depression and alcohol use disorder (OR = 25.11) (Holmstrand, Bogren, Mattisson, & Bradvik, 2015).

Systematic reviews and meta-analyses of psychological autopsy studies estimated that at least one psychiatric diagnosis is found in 87.3–98% of suicides, with mood disorders (30.2–43.2%), substance-related disorders (17.6–25.7%), personality disorders (13–16.2%), and psychotic disorders (9.2%–14%) being the most common diagnoses (Arsenault-Lapierre, Kim, & Turecki, 2004; Bertolote, Fleischmann, De Leo, & Wasserman, 2004; Cavanagh et al., 2003). The extent of the contribution and the causal relation between psychiatric disorders and suicide is debated (Hjelmeland & Knizek, 2017; Pridmore, 2015): psychological autopsy studies may introduce a recall bias and so cause an overestimation of the presence of psychopathological symptoms in suicide decedents (Too et al., 2019). Nevertheless, record linkage studies estimated that up 21% of suicide may be prevented through the prevention of psychiatric disorders and confirmed their importance as an important modifiable risk factor for suicide (Too et al., 2019).

In a large U.S. survey, a prior psychiatric condition was found in 66% of participants with serious suicidal ideation, in 77.5% of those with a suicidal plan and in approximately 80% of suicide attempters (Nock et al., 2010). The prevalence of a history of mental disorders was higher among people making planned than among those making unplanned suicide attempts (83.4 vs. 74.1%), leading the authors to hypothesize that other factors, such as stressful life events, may have a greater influence on unplanned attempts. Furthermore, this study showed that different disorders had a different impact on suicidal ideation and attempts. Depression predicted suicide ideation, but it was not a significant predictor of suicide plans or attempts among those with ideation. Instead, suicide plans and attempts among ideators were predicted by disorders characterized by severe anxiety/agitation (e.g., posttraumatic stress disorder) and poor impulse control (e.g., conduct and substance use disorders).

3.1 Mood and anxiety disorders

A more than 20-fold higher standardized mortality ratio (SMR) for suicide in unipolar depressed patients than in the general population has been reported (Lepine & Briley, 2011) and suicide accounts for approximately 15% of deaths in bipolar disorder (Miller & Bauer, 2014). A meta-analysis on Chinese patients with major depression estimated a pooled lifetime prevalence of 53.1% for suicidal ideation, 17.5% for suicide plan and 23.7% for suicide attempt (Dong et al., 2018).

Given the quite extensive prevalence of depression and its strong association with suicidality, preventing suicide in people with depression is extremely important. Nevertheless, there is little knowledge about the specific risk and protective factors for suicide in this group.

Isometsa (2014) discussed in a systematic review risk factors for both completed and attempted suicide in depressive and bipolar disorders. Suicidal behavior in mood disorders was found to be strongly associated over time with the course of the illness. Most of these behaviors take place during major depressive and mixed mood episodes. For this reason, the author concluded that reducing the duration of high-risk states may be an effective measure in reducing the overall risk for suicidal acts. Other identified risk factors were the comorbidity with substance use or cluster B personality disorder, hopelessness, impulsive-aggressive traits, childhood and recent adverse life experiences, and poor perceived social support.

In a systematic review of 19 cohort and case–control studies, a previous suicide attempt (OR = 4.84, 95% CI: 3.26–7.20) was the strongest risk factor for suicide in people with major depression (Hawton, Casanas, Haw, & Saunders, 2013). Male gender and a family history of psychiatric disorder also conferred an increased risk, while a family history of suicide showed a nonsignificant increased trend. Concerning clinical characteristics, more severe depressive psychopathology (OR = 2.20, 95% CI: 1.05–4.60), and hopelessness (OR = 2.20, 95% CI: 1.49–3.23) were also associated with increased suicide risk, while suicidal ideation (OR = 2.39, 95% CI: 0.80–7.11) did not reach significant results. Comorbidity with alcohol and drugs use (OR = 2.17, 95% CI: 1.77–2.66), anxiety symptoms (OR = 1.59, 95% CI: 1.03–2.45), and personality disorders (OR = 4.95, 95% CI: 1.99–12.33) were associated with increased risk of suicide.

A similar prevalence of suicidal behavior has been reported in bipolar I and II disorders (Novick, Swartz, & Frank, 2010). The international Society for Bipolar Disorders Task Force on Suicide conducted a meta-analysis on risk factors for suicide attempts and suicide deaths (Schaffer et al., 2015). As reported for suicide in the general population, the results indicated that suicide attempts in bipolar disorders were associated with female gender (OR = 1.54, 95% CI: 1.44–1.66), while suicide deaths were more common in men (OR = 1.83, 95% CI: 1.41–2.39). First-degree family history of suicide conferred an increased risk for both suicide attempts (OR = 1.69, 95% CI: 1.25–2.27) and deaths (OR = 2.91, 95% CI: 1.54–5.48). Other risk factors for suicide attempts in bipolar patients were younger age at illness onset (−2.99 years, 95% CI: −3.78–-2.20), depressive polarity of both first illness episode (OR = 1.92, 95% CI: 1.39–2.65) and current or most recent episode (OR = 5.99, 95% CI: 1.75–20.5), comorbid cluster B/borderline personality disorder (OR = 2.51, 95% CI: 1.91–3.3), anxiety disorder (OR = 1.81, 95% CI: 1.66–1.97), or substance use disorder (OR = 1.81, 95% CI: 1.31–2.50), especially alcohol use.

Malone et al. (2000) compared 45 inpatients with major depression who had attempted suicide with 39 inpatients with the same diagnosis who had not attempted suicide. No differences emerged for objective severity of psychopathology and quantity of recent life events, but those without a history of suicide attempt showed significantly more reasons for living (i.e., responsibility toward family, fear of social disapproval, moral objections to suicide, survival and coping beliefs, fear of suicide and, for those with offspring, child-related concerns) than those with such a history. Reasons for living were correlated negatively with self-reported hopelessness, depressive symptoms, and suicidal ideation.

Anxiety symptoms represent a major risk factor for suicide among patients with mood disorders. The high comorbidity between mood and anxiety disorders makes this particularly important (Brown, Campbell, Lehman, Grisham, & Mancill, 2001; Hirschfeld, 2001). A meta-analysis has found patients with anxiety disorders showed to be at an increased risk for suicidal ideation (OR = 2.89), suicide attempt (OR = 2.47), and completed suicide (OR = 2.85) when compared with patients without anxiety (Kanwar et al., 2013).

3.2 Alcohol and other substance use disorders

As noted before, comorbidity between alcohol or other substance use and psychiatric disorders increases the risk for suicide, probably through increased impulsivity, hostility, and aggression (Vijayakumar, Kumar, & Vijayakumar, 2011). In a cohort study in Denmark, patients with a psychiatric disorder and comorbid alcohol or other substance use showed a threefold increased risk of completed suicide compared to patients without comorbidity (Ostergaard, Nordentoft, & Hjorthoj, 2017).

A meta-analysis reported sufficient evidence for an increased risk for suicidal ideation (OR = 1.86, 95% CI: 1.38, 2.35), suicide attempt (OR = 3.13, 95% CI: 2.45, 3.81), and completed suicide (OR = 2.59; 95% CI: 1.95, 3.23) in alcohol use disorder (Darvishi, Farhadi, Haghtalab, & Poorolajal, 2015). Furthermore, Borges et al. (2017) conducted a meta-analysis on results from seven studies on acute alcohol use and suicide attempt. The authors showed that, despite the current literature being more focused on alcohol's chronic effects, suicide attempts are more likely to occur in acute use of alcohol (OR = 6.97, 95% CI: 4.77–10.17), particularly at high doses (OR = 37.18, 95% CI: 17.38–79.53). The alcohol use at the time of the attempt can be associated with a lesser degree of planning and a lower suicide intent (Bagge, Conner, Reed, Dawkins, & Murray, 2015).

Norstrom and Rossow (2016) described both individual and social mechanisms responsible for the increased risk for suicide related to alcohol consumption. Indeed, alcohol intoxication impairs cognitive functions, increasing the feeling of despair and sadness, as well as impulsivity and aggression. It also weakens or removes the barriers to self-harm and reduce coping abilities. Furthermore, alcohol use seriously reduces the level of social integration. For example, the study by Khemiri, Jokinen, Runeson, and Jayaram-Lindstrom (2016) shows that the experience of violence in childhood and the expression of violent behavior, coupled with increased impulsivity, were found to be overrepresented in patients with alcohol dependence and associated with an elevated risk for suicidal thoughts and behavior. In a study on patients with alcohol dependence in Taiwan, auditory hallucination (adjusted risk ratio [aRR] = 1.80, p = .04) and attempted suicide (aRR = 7.52, p = .001) were associated with suicide death. On the contrary, financial independence (aRR = 0.11, p = .005) and being married (aRR = 0.16, p = .02) emerged as protective factors (Hung et al., 2015).

A meta-analysis of 43 studies confirmed the existence of an association between substance use disorder and suicidal ideation (OR = 2.04, 95% CI: 1.59, 2.50), suicide attempt (OR = 2.49, 95% CI: 2.00, 2.98), and suicide death (OR = 1.49, 95% CI: 0.97, 2.00), with opioid substances having a stronger effect than other substances (Poorolajal, Haghtalab, Farhadi, & Darvishi, 2016). Indeed, a substantial increase in opioid-overdose deaths has been reported in United States to which suicide probably strongly contributed (Bohnert & Ilgen, 2019; Oquendo & Volkow, 2018). Furthermore, substance use has been linked to the recurrence of suicide attempts (Monnin et al., 2012; Parra-Uribe et al., 2017).

Fourteen times higher suicide rates are reported among heroin users compared to the general population and they also appear to be at increased risk for suicide attempt (Darke & Ross, 2002). This is likely linked to an extremely high prevalence of common recognized suicide risk factors in this group (e.g., male gender, presence of psychiatric disorders, such as depression and antisocial personality disorder, childhood adversities, and social isolation) (Darke & Ross, 2002).

Less evidence exists about an association between cannabis use and suicidality. Few studies have assessed this association and, while an increased risk for suicide death (OR = 2.56, 95% CI: 1.25–5.27), suicidal ideation (OR = 1.43, 95% CI: 1.13–1.83), and attempt (OR = 2.23, 95% CI: 1.24–4.00) has been found, additional studies are needed to better assess the measure of cannabis exposure and explore the role of confounding factors, such as depression and alcohol use (Borges, Bagge, & Orozco, 2016).

3.3 Schizophrenia

Approximately 5% of patients diagnosed with schizophrenia die by suicide, usually close to illness onset (Hor & Taylor, 2010; Palmer, Pankratz, & Bostwick, 2005). Laursen et al. (2019) calculated the excess mortality rate ratio for suicide to be 9.05 for males and 15.84 for females with schizophrenia. Furthermore, between 25 and 50% of individuals with schizophrenia make a suicide attempt (Meltzer, 2001).

The first years after diagnosis are associated with the highest suicide risk (Brown, Kim, Mitchell, & Inskip, 2010; Robinson et al., 2010; Simon et al., 2018). Those who die by suicide are more often young, male, and with a high level of education or a higher intelligence quotient (Cassidy, Yang, Kapczinski, & Passos, 2018; Hor & Taylor, 2010; Popovic et al., 2014). However, a registry study in China found that death from suicide was more likely in women than men, as it is for suicide in the general population (Phillips, Yang, Li, & Li, 2004). Considering clinical features, depressive symptoms, including hopelessness, previous suicide attempts, and poor adherence to treatment, were associated with suicide (Cassidy et al., 2018; Hawton, Sutton, Haw, Sinclair, & Deeks, 2005; Hor & Taylor, 2010; Popovic et al., 2014). Comorbid substance use also seems to confer an increased risk for suicide and suicide attempts (Cassidy et al., 2018; Hawton et al., 2005; Hor & Taylor, 2010). Agitation or motor restlessness, fear of mental disintegration, and a recent loss were listed as risk factors for suicide in schizophrenic patients in the systematic review by Hawton et al. (2005). Contrasting results are reported for the presence of hallucinations. While in the systematic review by Hawton et al. (2005), hallucinations were associated with a reduced suicide risk, in the later review by Hor and Taylor (2010), auditory hallucinations and delusions were reported among the factors conferring an increased risk of suicide among schizophrenic patients.

3.4 Borderline personality disorder

Suicidal behavior is quite frequent in patients with borderline personality disorder. Indeed, it is estimated that at least three-quarters of them attempt suicide and approximately 10% die by suicide (Black, Blum, Pfohl, & Hale, 2004).

In the previously cited meta-review conducted by Chesney et al. (2014), borderline personality disorder showed the highest suicide risk among psychiatric disorders (SMR = 45.1, 95% CI: 29.0–61.3), even if the authors invited to be cautious since the data came from hospitalized samples, thus showing a severe degree of psychopathology. In a nested case control study on primary care patients diagnosed with a personality disorder in United States, borderline personality disorder patients showed a 37-fold increased risk for suicide (Doyle et al., 2016). Furthermore, a 45-fold increased suicide risk was reported for patients with personality disorders and comorbid alcohol use.

Previous suicide attempts and comorbidity with major depressive or substance use disorders increase the risk for suicide in borderline personality disorder. Other risk factors for suicidal behavior in this group of patients are hopelessness, impulsivity, early trauma, and the presence of antisocial traits (Black et al., 2004; Goodman, Roiff, Oakes, & Paris, 2012). Suicide completion in males often occurs with first attempt, while females most often show a path of recurrent attempts and unsuccessful treatment efforts before suicide completion (Goodman et al., 2012). Negative life events, especially related to interpersonal and legal problems and loss, often act as precipitating factors of suicidal behavior in individuals with personality disorders (Foster, 2011; Yen et al., 2005) and in those with comorbid major depressive disorder and borderline personality disorder (Brodsky, Groves, Oquendo, Mann, & Stanley, 2006).

3.5 Previous suicide attempters

A previous suicide attempt is considered to be the strongest predictor of completed suicide (Tidemalm, Langstrom, Lichtenstein, & Runeson, 2008). In a 5-year follow-up study on suicide attempters in Finland, a 54 times higher mortality risk for suicide was found in men and a 77 times higher risk in women (Ostamo & Lonnqvist, 2001). A prospective cohort study of suicide attempters in England found a 34-fold increased risk of suicide (50 times higher in women and 29 times higher in men) in this population (Cooper et al., 2005).

The first year after the attempt appears to be the most critical one (Bostwick, Pabbati, Geske, & McKean, 2016; Cooper et al., 2005); however, the risk for suicide remain high even 15 years later (Hawton, Zahl, & Weatherall, 2003).

Among adolescents, both a previous suicide attempt (OR = 3.3; 95% CI 2.4–4.4) and the use of more lethal methods (OR = 1.5; 95% CI 1.1–2.1) were associated with the repetition of suicidal behavior (Hulten et al., 2001).

Cooper et al. (2005) identified several significant risk factors for completed suicide after suicide attempt: not living with a close relative; avoiding discovery at time of suicide attempt; and current alcohol use. The use of cutting as suicidal mean, previous psychiatric treatment, and somatic health problems were commonly found although not statistically significant as risk factors. Age and gender also seem to play a role, with males and older suicide attempters being at higher risk of completed suicide than female and younger attempters (Bostwick et al., 2016; Hawton et al., 2003).

4 SUICIDE ASSESSMENT AND TREATMENT 4.1 Suicide risk assessment

A comprehensive suicide risk assessment represents the first step in effective suicide prevention among psychiatric patients. The assessment should explore psychiatric, somatic, psychological, social, and, if possible, even neurobiological factors related to suicide risk (Wasserman et al., 2012). Spiritual well-being and religious beliefs should also be considered during the assessment since they may represent an important protective factor against suicidal behavior (Kopacz, 2014; Moreira-Almeida, Sharma, van Rensburg, Verhagen, & Cook, 2016; Rasic et al., 2009). On occasions, religious beliefs and spiritual anguish may add to the risk of suicide and complicate recovery from disorders (Verhagen, Praag, López-Ibor, Cox, & Moussaoui, 2012).

Furthermore, since suicide risk may vary even within a short period of time, the assessment should be repeated over time. This is particularly important if we consider that contact with primary and mental health services prior to suicide is common. A lifetime history of contact with mental health care was found in 53–57% of people who died by suicide, while rates of contact with mental health services were nearly 32% in the year and 19–21% in the month prior to suicide (Luoma, Martin, & Pearson, 2002; Stene-Larsen & Reneflot, 2019). Furthermore, approximately 45% of suicide victims had a contact with primary care services in the month before suicide. These data may also explain why education of general practitioners and other health workers on recognition and treatment of depression may decrease suicide rates (Wasserman et al., 2012; Zalsman et al., 2016).

Considering the increased risk for suicide posed by the comorbidity between mood disorders and alcohol and substance use, the suicidal risk should be routinely assessed in individuals receiving addiction treatment (Yuodelis-Flores & Ries, 2015). From a public health perspective, policies aimed at reducing the harmful use of alcohol may have a protective effect on suicide (Värnik, Kõlves, Väli, Tooding, & Wasserman, 2007; Wasserman & Värnik, 1998; Wasserman, Varnik, & Eklund, 1994; Xuan et al., 2016).

4.2 Levels of psychiatric treatment

Particular attention should be paid during and after hospitalization. Compared to the general population, individuals admitted to psychiatric inpatient care had more than forty-fold higher risk of suicide (Hjorthoj, Madsen, Agerbo, & Nordentoft, 2014). The early phase of admission represents the period at highest-risk (Bowers, Banda, & Nijman, 2010). Large, Smith, Sharma, Nielssen, and Singh (2011) conducted a meta-analysis to identify possible factors associated with in-patient suicide. A diagnosis of schizophrenia and depressed mood appeared to confer a particular high risk. Other risk factors were a history of deliberate self-harm, hopelessness, feelings of guilt or inadequacy, depressed mood, suicidal ideas, and a family history of suicide. High post-discharge suicide rates are also consistently reported and identify the first days and weeks after discharge as the most critical ones (Madsen & Nordentoft, 2013; Olfson et al., 2016; Tseng, Chang, Liao, & Yeh, 2019).

4.3 Pharmacological treatment

Given the strong relationship existing between mood disorders and suicide, antidepressants and mood stabilizers have received great attention. A protective effect of antidepressant use on suicide has been reported (Brent, 2016). However, in the early phase of treatment, SSRIs might increase suicidal thoughts, but not actual suicide (Cipriani, Geddes, Furukawa, & Barbui, 2007). Furthermore, they proved to be effective in reducing suicidal thoughts and behaviors in adult and geriatric patients (Barbui, Esposito, & Cipriani, 2009; Gibbons, Brown, Hur, Davis, & Mann, 2012). Since suicidal individuals often experience strong anxiety and agitation, the use of antidepressants with a sedative profile or the use of a sedating or sleep-inducing comedication may be indicated (Möller, 2016). The effectiveness of lithium in reducing the suicide risk in mood disorders has been extensively proven (Del Matto et al., 2020; Smith & Cipriani, 2017). Anticonvulsant mood stabilizers (e.g., carbamazepine, valproate, lamotrigine) have also been shown to decrease the suicide risk in patients with bipolar disorders (Gibbons, Hur, Brown, & Mann, 2009; Miller & Black, 2020).

Antisuicidal effects in patients with schizophrenia are reported for second-generation antipsychotics (Aguilar & Siris, 2007; Barak, Mirecki, Knobler, Natan, & Aizenberg, 2004; Ringback Weitoft et al., 2014), especially for clozapine (Asenjo Lobos et al., 2010).

Among depressed patients, suicidal ideation rapidly decreases after the infusion of ketamine, with effects emerging within an hour and lasting up to a week. However, further research is needed to prove its effectiveness not only on suicidal thoughts but also behavior, as well as in longer follow-ups (Lengvenyte, Olie, & Courtet, 2019; Wilkinson et al., 2018).

4.4 Psychological treatment and brief interventions

Evidence exists for the effectiveness of psychotherapeutic interventions. In particular, cognitive behavioral therapy shows promising results in reducing suicidal thoughts and behavior, in both adults and youth (Leavey & Hawkins, 2017; Robinson, Hetrick, & Martin, 2011; Tarrier, Taylor, & Gooding, 2008). Dialectical behavior therapy has also shown effectiveness in reducing suicidal behavior and reattempts, with particular effect in females with borderline personality disorder (DeCou, Comtois, & Landes, 2019).

Crisis response planning—aimed at identifying warning signs of suicide, promoting coping skills, and increasing family and social support and access to professional help—has been found effective in reducing suicidal ideation and behavior (Bryan et al., 2017; Stanley & Brown, 2012). The brief intervention and contact (BIC) implemented in the WHO Multisite Intervention Study on Suicidal Behaviors (SUPRE-MISS) comprises, in addition to treatment as usual, a 1-hr long motivational for treatment psychoeducation session and nine follow-up contacts in the following 18 months. The SUPRE-MISS randomized controlled trial demonstrated a significant decrease in suicide deaths compared with treatment as usual (Fleischmann et al., 2008).

5 CONCLUSIONS

The early identification and treatment of psychiatric disorders constitutes one of the most important suicide prevention strategies, and both pharmacological and psychological interventions have been proven effective (Wasserman et al., 2012; Zalsman et al., 2016). Nevertheless, psychiatric disorders are often underrecognized and undertreated, especially in primary care settings (Lecrubier, 2007; Mitchell, Vaze, & Rao, 2009; Smith et al., 2011). In a study examining treatment of major depressive disorder in 21 countries, minimally adequate treatment was reported for one in five individuals in high-income and one in 27 in low-/lower-middle-income countries (Thornicroft et al., 2017). The widest treatment gap has been reported for alcohol use and dependence (78%), but large gaps were found also for depression (56%), bipolar disorder (50%), and schizophrenia (32%) (Kohn, Saxena, Levav, & Saraceno, 2004). Increasing the number of psychiatrists and other mental health professionals, increasing the involvement of a range of appropriately trained nonspecialist providers, and the active involvement of people affected by mental disorders were identified as promising strategies to reduce the treatment gap in mental health care (Patel et al., 2010).

A psychiatric diagnosis represents an important suicide risk factor. The acute phases of these disorders may act as precipitating factors of suicidal crises (Wasserman & Sokolowski, 2016). However, since most individuals with psychiatric disorders do not die by suicide, it is important to understand which characteristics, including the genetic ones (Sokolowski & Wasserman, 2020), may differentiate those at high suicide risk. Common recognized risk factors for suicide, such as a personal and family history of suicide attempts, adverse childhood experiences, stressful life events, and feeling of hopelessness, are often identified among psychiatric patients who are overrepresented populations with suicidal behaviors. Additionally, the severity of psychopathology, also reflected in the comorbidity between different disorders and in different levels of psychiatric treatment, confers a significantly increased risk (Cavanagh et al., 2003; Hawton et al., 2013; Hjorthoj et al., 2014). There is evidence that high impulsive and aggressive traits as well as severe anxiety/agitation contribute strongly to suicidal behavior (Isometsa, 2014; Nock et al., 2010). However to better understand why people engage in suicidal behavior, and so predict and prevent those acts, further research on decision-making processes is needed (Gvion, Levi-Belz, Hadlaczky, & Apter, 2015; Isometsa, 2014).

Suicide research is mostly focused on risk factors, while much less is known about factors that may exert a protective role; even if almost 10 years ago Brent (2011) advocated for a hybrid intervention that treats a current disorder but also promotes long-term resilience. For example, social support is recognized as an important protective factor in individuals with psychiatric and substance use disorders (Kleiman & Liu, 2013; Otsuka et al., 2019; Xie et al., 2018; You, Van Orden, & Conner, 2011) and it is associated with increased professional help-seeking for suicidality (Han, Batterham, Calear, & Randall, 2018). Increasing sleep quality, promoting physical activity, healthy diet, and reading literature have been reported as protective strategies for mental health promotion and suicide prevention (Bishop, Walsh, Ashrafioun, Lavigne, & Pigeon, 2020; Davis, 2009; Kasahara-Kiritani et al., 2015; Li, Zhang, & McKeown, 2009; Vancampfort et al., 2018). Spiritual care is also advocated as an important component of the psychiatric treatment (Culliford, 2002). However, it remains unaddressed in many settings despite promising research in the field and attempts at integration (Heidari, Borujeni, & Rafiei, 2019; Kopacz, 2013; Moreira-Almeida et al., 2016; Verhagen et al., 2012).

Although further knowledge of risk and protective factors and on the effectiveness of different prevention strategies is still required, suicide and suicide attempts can be prevented by adequate treatment of people experiencing psychiatric disorders. This often requires a combination of pharmacological, psychological, and social interventions (Wasserman et al., 2012); and desirably spiritual care as needed. In this sense, initiatives to increase the access to mental health care, such as the WHO Mental Health Gap Action Programme (mhGAP) (World Health Organization, 2008), are pivotal for suicide prevention.

CONFLICT OF INTEREST

The authors have no conflicts of interest to disclose.