Observational study of suicide in Switzerland: comparison between psychiatric in- and outpatients

DOI: https://doi.org/10.4414/SMW.2022.w30140

Introduction

Suicide is a major public health issue: the World Health Organization (WHO) estimated that 800,000 people died by suicide worldwide in 2019, which makes it one of the leading causes of premature death [1]. In Switzerland, suicide is among the first causes of years of potential life lost [2]. Between 60% and 98% of the people who die by suicide suffer from a mental illness [3] and about 20% who subsequently die from suicide have contact with mental health services within one month and 32% within a year before suicide [4]. Psychiatric patients and psychiatric institutions are thus a major target for suicide prevention.

Suicide in psychiatric inpatients and related risk factors have been the subject of many studies. Rates of suicide for psychiatric inpatients range from 200 to 920 per 100,000 admissions [5–13] and their incidence per 100,000 admissions to general hospitals ranges from 1.7 to 20.9 [13]. Compared with people who die by suicide in the general population, an adjusted ratio for suicide in the preceding year of 44.3 for psychiatric inpatients was reported [14]. Chronic mental illness, a family history of suicide, suicidal ideation, recent bereavement and delusions were identified as predictors for inpatient suicide [8].

On the other hand, suicide in psychiatric outpatients has been less investigated [15, 16], although these patients may have different profiles and need targeted preventive interventions. We did not find any study comparing suicide rates or suicide circumstances between psychiatric in- and outpatients.

Gender difference in suicide is a frequently investigated topic in the general population [17]. Regarding psychiatric patients, gender-related differences were incidentally reported by some studies: male patients who died by suicide are more likely to be single [15] and suffer from schizophrenia [18,19], female outpatients die more by drug intoxication [15] and are more likely to suffer from an affective disorder [20]. However, gender differences among psychiatric in- and outpatients who die by suicide have not been investigated.

Evidence is therefore lacking on sociodemographic characteristics and gender specificities of psychiatric in- and outpatients dying by suicide. We aimed to gain more information on those questions by analysing data from a committee dedicated to review all suicides in the public institutions of the Canton of Vaud, Switzerland.

Material and methods

Study design and study setting

All critical incidents involving patients from the Psychiatric Department of the Lausanne University Hospital, which includes suicide and other deaths, major agitation and prolonged restraint (excluding suicide attempts, which are not evaluated by this committee), are reviewed by the Committee of Clinical Practice Review (CCPR). Its primary goal is to identify possible problematic issues related to the organisation of the clinic and to provide a feedback including a report, both specific to the case (e.g., meeting with clinicians involved in a given situation and restitution to the team) and general (e.g., a yearly newsletter summarising some illustrative situations in an anonymous format addressed to all clinicians of the department) [21].

We performed an observational study of all cases of patients who died by suicide reviewed by the CCPR between 2007 and 2019 (without exclusion criteria).

Data collection methods and data flow

An ad hoc evaluation form was developed and implemented in the CCPR from 2016 to 2019. The same form was retrospectively filled in for the reviews provided by the CCPR between 2007 and 2016.

The selection of data for the form was based on variables identified in the literature on existing self-harm monitoring systems [22–24] and on our clinical experience. The data included sociodemographic (e.g., age, gender, marital status, nationality, socio-economic situation, education, professional activity) and clinical information (e.g., main psychiatric diagnosis, treatment, drug use, past history of suicide attempt or hospitalisation for suicide attempt, method of suicide, recent and past significant life events, last contact with a psychiatric clinician and evaluation of suicidality formalised or not by a structured evaluation of suicidality).

Ethical considerations

The project was conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. The local ethics committee on human research (CER-VD) approved the project (no. 2017-01932). The articles of law concerning research on deceased persons (LRH, art. 36 and 37) [25] require consent of the research participant given during that person's lifetime or, in the absence of a document attesting to the consent or refusal of the deceased, from a relative or a person designated by the deceased during his or her lifetime. The CER-VD considered, however, that it would not be ethical to contact the relatives, which could revive painful experiences and reveal confidential data about the deceased.

Statistical analyses

Because of the descriptive nature of our analysis, power calculations and sample size were not predetermined. Comparisons between groups were performed using the Wilcoxon rank sum test [26] for continuous variables and Fisher's exact test [27] for categorical variables. To compare the differences observed in some variables (such as method of suicide) between in- and outpatients, a series of serially adjusted logistic regressions were performed (first adjusted for age and gender, then socioeconomic situation and finally the hospitalisation status). Confidence intervals (CIs) were calculated using the profile likelihood which provides reliable confidence intervals for estimated parameters [28]. All statistical analyses were performed using R (4.1.1) Environment for statistical computing [29]. Tables were produced using Stargazer [30] and gtsummary packages in R [31]. All statistical tests were two-tailed and significance was determined at the 0.05 level.

Results

Description of the population

Data of 153 suicides were analysed. Sociodemographic characteristics of the sample are presented in table 1. The majority of the patients were male (65%, n = 99) and of Swiss nationality (71%, n = 109). The sample shows an almost equal distribution of patients aged 18–35 years (31.4%, n = 48), 35–50 (30.7%, n = 47) and 50–66 (28.1%, n = 43), but not for patients above 66 years 7.8% (n = 12). Regarding marital status, 48% (n = 71) of patients were single, 20.3% (n = 30) in a partnership (married of cohabiting) and 31.4 % were either divorced (18%), separated (10%) or widowed (3.4%). Sixty-five percent of the patients (n = 67) were reported as facing a difficult socioeconomic situation, 35% as non-problematic (n = 36) (missing data, n = 50). An overwhelming majority (82.3%, n = 93) had low/medium education (they did not attend high school, college or university).

Mood disorders were the most frequent main diagnoses (39%, n = 56), followed by schizophrenia and other psychotic disorders (32%, n = 46), and personality disorders (15%, n = 21). Hanging was the most commonly used method of suicide (28%, n = 37), followed by jumping from a height (21%, n = 27) and by transportation (collision with a train, for example) (15 %, n = 20). Three quarters (76%, n = 81) of the patients had at least one previous suicide attempt (85% of the outpatients and 74% of the inpatients). Most patients mentioned suicidal thoughts at their last consultation (66%, n = 92). Therapists frequently used a formalised scale to evaluate the suicide potential (46%, n = 62).

We analysed significant events recorded in the reports of the CCPR (tables 1, 2, 4, 8, 9) for cases mostly from 2016 to 2019 (family conflict, change/transition of treating physician/caregiver, school or professional breakdown, asked for/ or were refused a disability insurance, intimate partner relationship conflict). Data for the years 2007 to 2016 were missing which explains the small numbers.

Comparison between in- and outpatients

Almost all of the outpatients (97.2%, n = 70) and inpatients (n = 70) had at least one past hospitalisation. Most of the sociodemographic and clinical characteristics showed no significant differences (table 2) between the two groups. Outpatients (19%, n = 11) more often than inpatients (1.6%, n = 1) used intoxication by medication as a suicide method (Fisher’s exact test p <0.001). About three quarters (72%, n = 49) of the outpatients had a last personal contact with clinicians less than a week before suicide (38.8 % of those less than 24 hours), as did 95.6% (n = 65) of the inpatients (90.8% of those less than 24 hours) (Fisher’s exact test p <0.001 for the last contact. See also table 3 for the logistic regressions). About 60% (n = 34) of the inpatients had a formalised suicidality evaluation during the last personal contact with clinicians versus 37% (s = 26) of the outpatients (Fisher’s exact test p <0.044 for the evaluation of suicide).

Comparison between genders

Table 4 shows the characteristics of female and male patients.

Female patients had more often been previously admitted into a psychiatric hospital (four or more hospitalisations for 61% of females, n = 31 versus 39%, n = 32 for males) (Fisher’s exact test p <0.014 for the number of past hospitalisations (see table 5 for the logistic regressions: OR  0.773, 95% CI 0.623–0.959). They had a higher rate of past self-harm (more than three in 35%, n = 13 versus 17%, n = 12 for males) (Fisher’s exact test p <0.046 for the previous suicide attempt). Females used hanging less often (17%, n = 7) than males (33%, n = 29) (Fisher’s exact test p <0.91) (see table 6 for the logistic regressions) and significantly more often jumping from a height (34%, n = 14 versus 15%, n = 13 for males) (Fisher’s exact test p <0.019) (see table 7 for the logistic regressions: OR 0.817, 95% CI 0.682–0.977). Main psychiatric diagnoses also showed a significant difference, with personality disorders being more frequent in females (26%, n = 13) than in males (8.5%, n = 8) (Fisher’s exact test p <0.018 for the diagnosis).

Discussion

We will first compare the sample of in- and outpatients and compare our results with data on suicide in the general population [2], then examine gender differences and finally discuss suicidal risk assessment at the last contact before death.

Our study showed that psychiatric outpatients dying by suicide do not significantly differ from inpatients, suggesting that suicide prevention should address these two populations in the same way. However, we also observed that 97.2% of the outpatients had at least one previous psychiatric hospitalisation and that about 50% of both in- and outpatients had four or more past hospitalisations. Moreover, a large proportion of our sample had a history of attempting suicide (85% of the outpatients and 74% of the inpatients). Both populations thus had a serious psychiatric history and had already received psychiatric care prior to suicide.

Among risk factors, we looked at significant recent events and found a high rate of change of treating physician/caregiver, twice as high in the outpatient setting as the inpatient setting (57% vs 21%) (see table 8 for the logistic regressions: OR  0.685, 95% CI 0.485–0.969). This is particularly relevant because of the high frequency of caregiver rotation, especially among residents in training in public psychiatry. Training requirements imply that patients have to change therapists regularly, at least once a year and sometimes every 6 months. We found that 46% of the outpatients with a change in caretaker who died by suicide did so 2 months or less after the institutional turnover. Michaud et al. found that those type of transitions may contribute to patients' feeling of “loneliness” and/or “abandonment”, which may revive past experiences and increase suicide risks [21]. Our sample is too small to verify this hypothesis, but the data nevertheless suggest that teams should be especially cautious at times of changes of caregivers.

We also found a certain proportion of family conflicts (36.9%) among our population of patients who died by suicide, a factor that could be systematically investigated by clinicians (see table 9 for the logistic regressions).

Compared with the general population, our psychiatric population had higher rates of previous hospitalisations. Pirkis and Burgess found that 41% of people in the general population who died by suicide were hospitalised in the year before their death [32] and Brown et al. reported that 67% of outpatients who died by suicide had a history of psychiatric hospitalisation [16]. Parra-Uribe et al. observed that 60.4% of people died at their first attempt [33], which is a much higher proportion than in our sample (only about 15% of the outpatients and 26% of the inpatients). Powell et al. compared psychiatric inpatients who died by suicide with a general population who died by suicide and found that the rates of previous self-harm (not including acts leading to or during index admission) were 54% and 26%, respectively [8]. Although this high proportion of past suicide attempts among psychiatric patients is obviously partly related to the fact that attempting suicide leads to being treated, clinicians should bear in mind that suicide seems strongly connected with past suicide attempts in psychiatric patients and that patients without a suicide attempt history rarely die by suicide.

Several descriptive findings deserve attention. Hanging, jumping from a height and suicide by transportation were the most frequent methods, as reported in the literature for inpatients [6, 11, 13, 34]. In the Swiss population (2017), hanging and jumping from a height were also the most frequent methods [2], with suicide by firearm coming third (19%) [2]. In our sample only 5.3% of the patients used the latter method. To understand this discrepancy between suicide in the general population and our sample it is useful to look at the context of firearms in Switzerland. Switzerland has an army consisting of a militia and soldiers store their gun at home. Consequently, Swiss strategies for preventing suicide systematically address the issue of access to firearms at home [35]. This example shows the importance of restricting access to lethal means, as Sabe et al. observed [36], as well as the importance of continuing to raise awareness about these issues among caregivers. In our sample, this point was also highlighted by the fact that rates of intoxication by medication were 11 times higher in outpatients than in inpatients and that restricting access to medication seems to work in inpatient settings and could probably be improved in the outpatient setting (e.g., daily or weekly pharmacy-controlled allocation).

Comparison between genders showed a male/female ratio of 1.9 in our sample. This ratio is in line with existing data of psychiatric patients [8, 9, 19, 37, 38], but notably differs from the one found in the Swiss general population, i.e., 3 [2]. Martelli et al. also found this discrepancy in their literature review [13], as did Frei et al. [39], but neither of them commented on it. A first hypothesis could be that females are overrepresented in psychiatric patients. Indeed, more males than females fail to obtain care [40, 41], because they may “deny illness, suppress negative feelings and refuse to admit depressive symptoms or delay seeking help” [17]. For males, perceiving and accepting a need for help can be interpreted as infringing traditional role expectations [42]; females, on the other hand, are more inclined to seek help and to be treated [43]. These gender differences highlight the need for developing preventive interventions targeting males, such as the sharing of experiences among suicide survivors [44], and environmental and educational approaches to help males express their feelings [17]. Nevertheless, the proportion of female patients in our department over the last 11 years ranged from 49% to 52% in inpatients and from 45% to 47% in outpatients. This hypothesis may thus be valid for the general but not for our psychiatric population. The fact that we have a male/female ratio close to 1 in the overall psychiatric patient population of our department, which is similar to the ratio that Ostertag et al. found in patients who attempt suicide [24], whereas the ratio is close to 2 in our sample of patients who died by suicide could be explained by the fact that males choose more fatal methods than females [41].

A second hypothesis could be that females are less adequately or differently cared for compared with males and therefore more at risk of suicide once they enter the healthcare system. In our sample, females had a higher rate of past hospitalisations (four or more: 61% vs 39% for males) (see table 5 for the logistic regressions). This seems consistent with evidence showing that repeated admission correlate with an increased suicide risk, even more significantly in females [20]. Clinicians should thus pay close attention to female patients with repeated psychiatric hospitalisations. Furthermore, gender differences in allocated care exist in other areas [45] and should be studied for suicidal patients. Indeed, the expression of mental illness can show gender differences for certain psychiatric disorders such as depression [46].

A third hypothesis would be that this discrepancy is explained by differences between the general and the psychiatric populations in the prevalence of certain risk factors, unequally distributed between males and females (e.g., level of education [37]).

Our study also revealed gender differences with regard to suicide methods. Females used jumping from a height twice as often as males, which contrasts with the Swiss population (where 29.6% of females used jumping versus 25.2% in males). Earle et al. found a similar ratio in an outpatient sample (45.2% of the females versus 25.9% of the males jumped from a height) but did not comment on it. On the other hand, data on suicide attempts in French-speaking Swiss regions showed no significant difference in methods [24]. Moreover, in our sample, hanging was used twice as often by males than females, which also differs from the Swiss population, in which the ratio is around 1:1 (31.1% and 34.3%) [2]. Hanging and jumping from a height therefore show more gender differences in the psychiatric population than in the general population, a finding which should be further investigated.

Finally, our results provide insights on suicide risk assessment. We found a high rate of suicidal ideation reported during the last clinical contact in our population (65% of the inpatients and 70% of the outpatients). These percentages are higher than those reported in the literature. Earle et al. observed that 27% of their cohort of outpatients who died by suicide had expressed suicidal ideation at the last interview [15]; Bush et al. reported 22% for inpatients [47]. Healthcare system characteristics might explain these differences. It is possible that the possibility to set up frequent consultations for outpatients in our region helps patients to establish a trusting relationship with their therapist. This hypothesis is confirmed by the observation that 72% of the outpatients had a last contact less than a week before suicide, almost 40% of these in the last 24 hours (28% of the outpatients). Other studies found longer delays between suicide and last clinical contact. Laanani et al. found that 94% of the outpatients who died by suicide consulted in the year before suicide death (including all types of medical consultations and emergency room visits), one third of them in the last week and 8.5% on the last day [48]. Appleby et al. obtained similar results in community patients with 32.2% of suicides occurring less than a week after a last contact with mental health services and 8.1% within 24 hours [38]. Earle et al. found that 76% of the outpatients had a last contact within the past two weeks [15] and John et al. observed that 2.9% of the hospital outpatients had contact within the last week [49]. Our psychiatric patients had thus a closer contact with mental health care than people who died by suicide in the general population. In comparison with these other studies, our care system may be different in terms of resources with the possibility, for example, of more intensive follow-up and shorter time between consultations, which could explain this difference. However, this finding also demonstrates that mental health caregivers are able to identify people at risk for suicide and establish a closer follow-up or refer them to psychiatric inpatient care. The fact that a high proportion of our sample (almost 100%) had a history of hospitalisation also supports this hypothesis. Unfortunately, close follow-up and the opportunity to express suicidal ideation, although important to prevent suicide [36], is not always successful. Furthermore, our results show that therapists should not be falsely reassured by a very recent contact with a patient.

Limits of the study and future directions

Regarding the limitations of our study, underreporting of deaths by suicide in our department cannot be excluded, which may have led to a possible bias. In addition, data were collected by means of medical files, which may have contained wrong or incomplete information (e.g., lack of documentation on suicide risk assessment). A weakness of our study is the fact that our retrospective and observational design did not allow measurement of some important factors regarding suicide risk in psychiatric patients. Indeed, several models underline the importance of taking into account vulnerability traits or the association of other factors that can lead to suicide. Examples include the "Narrative crisis model" [50], which explores several stages leading to suicide with initially the presence of vulnerability traits (history, impulsivity, perfectionism...) and then in the context of stress, the emergence of "suicidal narratives" (such as unrealistic life goals or the perception of no future), which lead to the "suicidal crisis" (entrapment, affective disturbance, loss of cognitive control, hyper-arousal). There is also the "interpersonal-psychological theory of suicide" model [51], which describes suicidal ideation as the result of the concurrence of the notions of “thwarted belongingness” (perception to be alone) and “perceived burdensomeness” (perception to be a burden). Furthermore, personality traits (for example, neuroticism) have been also investigated for their association with suicide [52], as well as the study of the effect of affective temperaments, mediated by other variables (such as hopelessness, depressive symptoms, mentalisation…) on suicide risk [53]. Finally, the potential impact of the caregiver on suicidal patients is also of importance, for example with the concept of adverse countertransference (e.g., disinterest, anxiety, overwhelming, rejection, helplessness or distress) [54]. Future studies assessing and comparing attempted and completed suicides could be most important to gain more knowledge of the suicidal process. Given the training course in our department on suicidal risk assessment, a new study could easily focus on this issue. However, to apply comprehensive models, such as those mentioned, might be difficult to realise given that a prospective study design is difficult to impose in these very delicate clinical situations, in which patients and caregivers are much preoccupied by the clinical situation and its evolution. Meanwhile, retrospective studies are a way to build up tendencies and to identify those factors, such as the change in staff we observed in this study, which can already be clinically addressed, for example by the simple heightened awareness of staff during these periods and a closer monitoring of patients.

Conclusions

This study of psychiatric in- and outpatients dying by suicide reveals most of the time a serious psychiatric history, and – compared to the general population – a higher rate of previous suicide attempts and past psychiatric hospitalisations. Surprisingly, almost every outpatient had at least one past psychiatric hospitalisation. This might explain why we did not find significant differences in characteristics between the in- and outpatients, indicating that they should not be addressed differently with regard to preventive interventions. Regarding gender, we found marked differences compared with the general population and between genders. More research is needed to explain those findings, which could help design preventive interventions and identify treatment modalities. Lastly, the high percentage of last contact in the 24 hours before death of outpatients suggests that patients at risk of suicide are efficiently identified, but that there might be room for improvement in the development of targeted preventive interventions.

Clinical points

• Clinicians should bear in mind that suicide seems to be strongly connected with past suicide attempts in particular in psychiatric patients and that patients without a previous suicide attempt history rarely die by suicide.

• Talking about suicidal ideation with patients is necessary, but not sufficient to prevent suicide.

• Clinicians should pay close attention to female patients with repeated psychiatric hospitalisations.

• Therapists should not be falsely reassured by a very recent contact with patients with suicidal ideation.

• Caregivers’ rotations are high-risk times and clinicians should carefully monitor suicidal risk in these periods.

Acknowledgements

We wish to thank Anouk Morier-Genoud, MMed, for her Master thesis about the retrospective cohort in the psychiatric institutions of the Canton of Vaud who realized the first step of our study. We are grateful for the help of Mehdi Gholam, PhD and Enrique Castelao, M.Psych., respectively statistician and data manager at the Center for Psychiatric Epidemiology and Psychopathology of the Lausanne University Hospital. A special thanks to Salome Diehr, Louise Ostertag and Lila Vammacigno from the Service of Liaison Psychiatry for their work in building the database. We would also like to thank Elodie Malbois, PhD, for her helpful comments.

References

1. WHO . | Practice manual for establishing and maintaining surveillance systems for suicide attempts and self-harm [Internet]. WHO. World Health Organization; [cited 2020 May 14]. Available from: http://www.who.int/mental_health/suicide-prevention/attempts_surveillance_systems/en/  

2. Suicide | OBSAN [Internet]. [cited 2020 May 14]. Available from: https://www.obsan.admin.ch/fr/indicateurs/suicide  

3. Bachmann S . Epidemiology of Suicide and the Psychiatric Perspective. Int J Environ Res Public Health. 2018 Jul;15(7):E1425. https://doi.org/10.3390/ijerph15071425  

4. Luoma JB , Martin CE , Pearson JL . Contact with mental health and primary care providers before suicide: a review of the evidence. Am J Psychiatry. 2002 Jun;159(6):909–16. https://doi.org/10.1176/appi.ajp.159.6.909  

5. Walsh G , Sara G , Ryan CJ , Large M . Meta-analysis of suicide rates among psychiatric in-patients. Acta Psychiatr Scand. 2015 Mar;131(3):174–84. https://doi.org/10.1111/acps.12383  

6. Hunt IM , Bickley H , Windfuhr K , Shaw J , Appleby L , Kapur N . Suicide in recently admitted psychiatric in-patients: a case-control study. J Affect Disord. 2013 Jan;144(1-2):123–8. https://doi.org/10.1016/j.jad.2012.06.019  

7. Kapur N , Hunt IM , Windfuhr K , Rodway C , Webb R , Rahman MS , et al.  Psychiatric in-patient care and suicide in England, 1997 to 2008: a longitudinal study. Psychol Med. 2013 Jan;43(1):61–71. https://doi.org/10.1017/S0033291712000864  

8. Powell J , Geddes J , Deeks J , Goldacre M , Hawton K . Suicide in psychiatric hospital in-patients. Risk factors and their predictive power. Br J Psychiatry. 2000 Mar;176(3):266–72. https://doi.org/10.1192/bjp.176.3.266  

9. Madsen T , Agerbo E , Mortensen PB , Nordentoft M . Predictors of psychiatric inpatient suicide: a national prospective register-based study. J Clin Psychiatry. 2012 Feb;73(2):144–51. https://doi.org/10.4088/JCP.10m06473  

10. Madsen T , Nordentoft M . Changes in inpatient and postdischarge suicide rates in a nationwide cohort of Danish psychiatric inpatients, 1998-2005. J Clin Psychiatry. 2013 Dec;74(12):e1190–4. https://doi.org/10.4088/JCP.13m08656  

11. Hunt IM , Kapur N , Webb R , Robinson J , Burns J , Turnbull P , et al.  Suicide in current psychiatric in-patients: a case-control study The National Confidential Inquiry into Suicide and Homicide. Psychol Med. 2007 Jun;37(6):831–7. https://doi.org/10.1017/S0033291707000104  

12. Large M , Smith G , Sharma S , Nielssen O , Singh SP . Systematic review and meta-analysis of the clinical factors associated with the suicide of psychiatric in-patients. Acta Psychiatr Scand. 2011 Jul;124(1):18–29. https://doi.org/10.1111/j.1600-0447.2010.01672.x  

13. Martelli C , Awad H , Hardy P . Le suicide dans les établissements de santé: données épidémiologiques et prévention. Encephale. 2010 Jun;36 Suppl 2:D83–91. https://doi.org/10.1016/j.encep.2009.06.011  

14. Hjorthøj CR , Madsen T , Agerbo E , Nordentoft M . Risk of suicide according to level of psychiatric treatment: a nationwide nested case-control study. Soc Psychiatry Psychiatr Epidemiol. 2014 Sep;49(9):1357–65. https://doi.org/10.1007/s00127-014-0860-x  

15. Earle KA , Forquer SL , Volo AM , McDonnell PM . Characteristics of outpatient suicides. Hosp Community Psychiatry. 1994 Feb;45(2):123–6.  

16. Brown GK , Beck AT , Steer RA , Grisham JR . Risk factors for suicide in psychiatric outpatients: a 20-year prospective study. J Consult Clin Psychol. 2000 Jun;68(3):371–7. https://doi.org/10.1037/0022-006X.68.3.371  

17. Barrigón ML , Cegla-Schvartzman F . Sex, Gender, and Suicidal Behavior. Curr Top Behav Neurosci. 2020;46:89–115. https://doi.org/10.1007/7854_2020_165  

18. Lester D . Sex differences in completed suicide by schizophrenic patients: a meta-analysis. Suicide Life Threat Behav. 2006 Feb;36(1):50–6. https://doi.org/10.1521/suli.2006.36.1.50  

19. Sani G , Tondo L , Koukopoulos A , Reginaldi D , Kotzalidis GD , Koukopoulos AE , et al.  Suicide in a large population of former psychiatric inpatients. Psychiatry Clin Neurosci. 2011 Apr;65(3):286–95. https://doi.org/10.1111/j.1440-1819.2011.02205.x  

20. Qin P , Nordentoft M . Suicide risk in relation to psychiatric hospitalization: evidence based on longitudinal registers. Arch Gen Psychiatry. 2005 Apr;62(4):427–32. https://doi.org/10.1001/archpsyc.62.4.427  

21. Michaud L , Stiefel F , Moreau D , Dorogi Y , Morier-Genoud A , Bourquin C . Completed Suicides in Psychiatric Patients: Identifying Health Care-Related Factors through Clinical Practice Reviews. Arch Suicide Res. 2019;1–15.  

22. Flavio M , Martin E , Pascal B , Stephanie C , Gabriela S , Merle K , et al.  Suicide attempts in the county of Basel: results from the WHO/EURO Multicentre Study on Suicidal Behaviour. Swiss Med Wkly. 2013 May;143:w13759.  

23. Hawton K , Bergen H , Cooper J , Turnbull P , Waters K , Ness J , et al.  Suicide following self-harm: findings from the Multicentre Study of self-harm in England, 2000-2012. J Affect Disord. 2015 Apr;175:147–51. https://doi.org/10.1016/j.jad.2014.12.062  

24. Ostertag L , Golay P , Dorogi Y , Brovelli S , Bertran M , Cromec I , et al. The implementation and first insights of the French-speaking Swiss programme for monitoring self-harm. Swiss Med Wkly. 2019 28;149:w20016. 

25. RS 810.30 - Loi fédérale du 30 septembre 2011 relative à la recherche sur l’être humain (Loi relative à la recherche sur l’être humain, LRH) [Internet]. [cited 2021 Oct 17]. Available from: https://www.fedlex.admin.ch/eli/cc/2013/617/fr  

26. Hollander M , Wolf DA , Chicken E . Nonparametric Statistical Methods, 3rd Edition | Wiley. 3th ed. 2013. 848 p. 

27. Agresti A . An Introduction to Categorical Data Analysis, 3rd Edition | Wiley. 3rd ed. 2018. 400 p. 

28. Held L, Sabanés Bové D. Applied Statistical Inference [Internet]. 2014 [cited 2021 Nov 12]. 376 p. Available from: https://www.springerprofessional.de/applied-statistical-inference/4092334  

29. R: The R Project for Statistical Computing [Internet]. [cited 2021 Nov 12]. Available from: https://www.r-project.org/  

30. Hlavac M . stargazer: Well-Formatted Regression and Summary Statistics Tables [Internet]. 2018 [cited 2021 Nov 12]. Available from: https://CRAN.R-project.org/package=stargazer  

31. Sjoberg DD , Whiting K , Curry M , Lavery JA , Larmarange J . Reproducible Summary Tables with the gtsummary Package. R J. 2021;13(1):570–80. https://doi.org/10.32614/RJ-2021-053  

32. Pirkis J , Burgess P . Suicide and recency of health care contacts. A systematic review. Br J Psychiatry. 1998 Dec;173(6):462–74. https://doi.org/10.1192/bjp.173.6.462  

33. Parra-Uribe I , Blasco-Fontecilla H , Garcia-Parés G , Martínez-Naval L , Valero-Coppin O , Cebrià-Meca A , et al. Risk of re-attempts and suicide death after a suicide attempt: A survival analysis. BMC Psychiatry. 2017 04;17(1):163. 

34. Meehan J , Kapur N , Hunt IM , Turnbull P , Robinson J , Bickley H , et al.  Suicide in mental health in-patients and within 3 months of discharge. National clinical survey. Br J Psychiatry. 2006 Feb;188(2):129–34. https://doi.org/10.1192/bjp.188.2.129  

35. Saillant S , Michaud L , Besson J , Dorogi Y . [Sentinel program: example of suicide prevention for an agricultural population in Switzerland]. Encephale. 2020 Aug;46(4):258–63. https://doi.org/10.1016/j.encep.2019.10.008  

36. Sabe M , Kaiser S , Niveau G . Suicide in psychiatry and medical liability: A case series. Int J Law Psychiatry. 2021 Jan-Feb;74:101671. https://doi.org/10.1016/j.ijlp.2020.101671  

37. Agerbo E . High income, employment, postgraduate education, and marriage: a suicidal cocktail among psychiatric patients. Arch Gen Psychiatry. 2007 Dec;64(12):1377–84. https://doi.org/10.1001/archpsyc.64.12.1377  

38. Appleby L , Shaw J , Amos T , McDonnell R , Harris C , McCann K , et al.  Suicide within 12 months of contact with mental health services: national clinical survey. BMJ. 1999 May;318(7193):1235–9. https://doi.org/10.1136/bmj.318.7193.1235  

39. Frei A , Bucher T , Walter M , Ajdacic-Gross V . Suicides in the Canton of Lucerne over 5 years: subjects with and without psychiatric history and diagnosis. Swiss Med Wkly. 2013 Mar;143:w13779. https://doi.org/10.4414/smw.2013.13779  

40. Oliffe JL , Phillips MJ . Men, depression and masculinities: A review and recommendations. J Men’s Health. 2008 Sep;5(3):194–202. https://doi.org/10.1016/j.jomh.2008.03.016  

41. Liu KY , Chen EY , Cheung AS , Yip PS . Psychiatric history modifies the gender ratio of suicide: an East and West comparison. Soc Psychiatry Psychiatr Epidemiol. 2009 Feb;44(2):130–4. https://doi.org/10.1007/s00127-008-0413-2  

42. Möller-Leimkühler AM . The gender gap in suicide and premature death or: why are men so vulnerable? Eur Arch Psychiatry Clin Neurosci. 2003 Feb;253(1):1–8. https://doi.org/10.1007/s00406-003-0397-6  

43. Gilligan C . In a Different Voice: Psychological Theory and Women’s Development. Reprint édition. Cambridge, Mass.: Harvard University Press; 2016. 216 p. 

44. Seidler ZE , Dawes AJ , Rice SM , Oliffe JL , Dhillon HM . The role of masculinity in men’s help-seeking for depression: A systematic review. Clin Psychol Rev. 2016 Nov;49:106–18. https://doi.org/10.1016/j.cpr.2016.09.002  

45. Chen EH , Shofer FS , Dean AJ , Hollander JE , Baxt WG , Robey JL , et al.  Gender disparity in analgesic treatment of emergency department patients with acute abdominal pain. Acad Emerg Med. 2008 May;15(5):414–8. https://doi.org/10.1111/j.1553-2712.2008.00100.x  

46. Judd F , Armstrong S , Kulkarni J . Gender-sensitive mental health care. Australas Psychiatry. 2009 Apr;17(2):105–11. https://doi.org/10.1080/10398560802596108  

47. Busch KA , Fawcett J , Jacobs DG . Clinical correlates of inpatient suicide. J Clin Psychiatry. 2003 Jan;64(1):14–9. https://doi.org/10.4088/JCP.v64n0105  

48. Laanani M , Imbaud C , Tuppin P , Poulalhon C , Jollant F , Coste J , et al.  Contacts with Health Services During the Year Prior to Suicide Death and Prevalent Conditions A Nationwide Study. J Affect Disord. 2020 Sep;274:174–82. https://doi.org/10.1016/j.jad.2020.05.071  

49. John A , DelPozo-Banos M , Gunnell D , Dennis M , Scourfield J , Ford DV , et al.  Contacts with primary and secondary healthcare prior to suicide: case-control whole-population-based study using person-level linked routine data in Wales, UK, 2000-2017. Br J Psychiatry. 2020 Dec;217(6):717–24. https://doi.org/10.1192/bjp.2020.137  

50. Galynker I . The Suicidal Crisis: Clinical Guide to the Assessment of Imminent Suicide Risk [Internet]. The Suicidal Crisis. Oxford University Press; [cited 2021 Oct 17]. Available from: https://oxfordmedicine.com/view/10.1093/med/9780190260859.001.0001/med-9780190260859  

51. Baertschi M , Costanza A , Richard-Lepouriel H , Pompili M , Sarasin F , Weber K , et al.  The application of the interpersonal-psychological theory of suicide to a sample of Swiss patients attending a psychiatric emergency department for a non-lethal suicidal event. J Affect Disord. 2017 Mar;210:323–31. https://doi.org/10.1016/j.jad.2016.12.049  

52. Stefa-Missagli S , Unterrainer HF , Giupponi G , Holasek SJ , Kapfhammer HP , Conca A , et al.  Suicide and Personality Traits: A Multicenter Study of Austrian and Italian Psychiatric Patients and Students. Suicide Life Threat Behav. 2020 Feb;50(1):220–32. https://doi.org/10.1111/sltb.12579  

53. Erbuto D , Innamorati M , Lamis DA , Berardelli I , Forte A , De Pisa E , et al.  Mediators in the Association Between Affective Temperaments and Suicide Risk Among Psychiatric Inpatients. Psychiatry. 2018;81(3):240–57. https://doi.org/10.1080/00332747.2018.1480251  

54. Michaud L , Greenway KT , Corbeil S , Bourquin C , Richard-Devantoy S . Countertransference towards suicidal patients: a systematic review. Curr Psychol. 2021 Feb. https://doi.org/10.1007/s12144-021-01424-0