Being Pro-Active in Meeting the Needs of Suicide-Bereaved Survivors: results from an audit in Montreal

Background: Suicide is a major public health concern. In 2017, the suicide rate in Canada was 11 per 100 000 inhabitants. According to the literature, 1 in 5 people have experienced a death by suicide during their lifetime. The aim of this study was to describe the met and unmet needs of suicide-bereaved survivors and to formulate postvention recommendations. Methods: Further to an exploratory mixed-method audit of 39 suicides that occurred in Montreal (Canada) in 2016, suicide-bereaved survivors (n = 29) participated in semi-structured interviews and completed instruments to discuss and assess potential pathological grief, depression (PHQ-9), and anxiety (GAD-7), as well as health and social services utilization. A panel then reviewed each case and formulated recommendations. The mean age of participants was 57.7 years and 23 were women. Results: Although help was offered initially, in most cases by a health professional or service provider (16/29), 22 survivors would have liked to be contacted by telephone in the rst two months post suicide. Four categories of individual unmet needs (medical/pharmacological, information, support, and outreach) and one collective unmet need (suicide pre/postvention training and delivery) emerged. Conclusions: Although provincial services have been developed and offered to suicide-bereaved survivors in the past decade, many dwindled over time and none has been applied systematically. Recommendations for different stakeholders (Ministry of Health and Social Services, coroners, NGOs, and representatives of suicide-bereaved survivors) outlined in this study could be an interesting rst step to help develop a provincial suicide pre/postvention strategy. Trial registration: The research protocol was approved by the Scientic Committee of the Montreal Mental Health Institute (November 2018) and the publication of results was approved by its Ethics Research Committee (February 1, 2019), project number 2019-1647.

According to McKinnon and Chonody (2014), proactive help should be offered by rst responders, coroners and other professionals that come into contact with SBS very early on. The absence of a formal process to connect SBS with support programs is in fact the principal unmet need of SBS (18).
While bereaved individuals know little about bereavement programs and how to access them, researchers know very little about the content and effectiveness of these programs (19). Some evidence of bene ts has emerged from a few intervention studies (19)(20)(21). For example, Andriessen et al. (2019) carried out a systematic review of controlled studies to assess evidence of the effectiveness of interventions for SBS and simultaneously appraise research quality. Their results showed that interventions targeting support, therapy and education seemed to work best when they included the social environment of the bereaved and when the therapy sessions were led by professionals. In another systematic review, Linde et al. (2017) found that bereavement groups had positive effects on uncomplicated grief and that cognitivebehavioral programs had positive effects on individuals at risk for suicide. However, as pointed out by Andriessen et al. (2019) and Linde et al. (2017), the overall quality of research in the eld of SBS postvention remains very weak so that what little evidence there is of the effectiveness of these interventions is neither strong nor reliable. For example, the results of a Canadian systematic review by Szumilas and Kutcher (2011) showed that only 16 of 49 studies of suicide postvention programs met their inclusion criteria regarding quality and evidence of effectiveness. Moreover, outcomes and measures varied widely from one study to another and most studies were conducted with women. To our surprise, none of the studies addressed interventions for individuals over the age of 65. Given the absence of any evidence-based suicide postvention program, further research is required into the exact form and structure of these programs. Furthermore, if these programs are proved ineffective, it could be because they were not based on the actual needs of bereaved adults to begin with. As shown by Wilson and Marshall (2010), less than half of individuals bereaved by suicide who expressed needing help with their grief process actually received help and only 40% of these were satis ed with the help received. In a study exploring the needs of SBS in Ontario (Canada), Gall, Henneberry, and Eyre (2014) described and compared the perspectives of SBS and of mental health workers and found that they complemented each other and helped identify best practices for SBS postvention. Pitman et al. (2018), too, demonstrated the importance of addressing support needs from the perspective of SBS. Finally, Séguin and Castelli-Dransart (2006) proposed that the help offered SBS follow a speci c progression as a function of their needs: family support in the rst weeks, support group if necessary, and therapy for di cult or pathological bereavement.
The above studies and ndings argue in favor of approaching SBS needs proactively (16,18) and of developing speci c programs to address these needs. The majority of SBS will not develop pathological bereavement or other problems. However, it is important to recognize those who do and to bear in mind that this may occur even 18 to 24 months post suicide. Moreover, it is reassuring to know that programs exist that have proved effective in situations of complicated bereavement (22,23).
In Quebec, suicide prevention centers often propose help in the form of support groups but SBS are not systematically referred to these centers. This is why it is important to assess social and health service needs and to understand why such needs go unmet. Grieving di culties, service utilization, and unmet needs should be assessed systematically.
Against this background, we undertook an exploratory cross-sectional and retrospective mixed-methods study to describe the met and unmet needs of SBS two years after the event and to formulate speci c suicide postvention recommendations over this two-year period.

Methods
This study consisted of a secondary analysis of data collected from SBS outreached in the course of a systematic audit of the consecutive suicide cases that occurred in the east end of the island of Montreal Of these, 39 consecutive cases were audited. These had been securely identi ed by the Quebec Coroner's O ce (QCO). The SBS were initially contacted by the QCO by mail. The letter that they received explained the purpose of the audit and solicited parties interested in participating in it. Participation was voluntary, no inducements were offered, and no obligations were imposed. Interested parties were then contacted to schedule an interview. Interviews were conducted at their home, at the research unit or over the telephone, according to the participant's preference and availability. For the purposes of this study, we used only the second part of the interview that focused speci cally on the SBS's experience of the suicide rather than on the deceased.

Data Collection
Data were collected through interviews and two self-report instruments. Aside from socio-demographics, data were also collected on risk/protective factors in and signs of pathological grief, relational proximity to the deceased, emotional experiences post suicide death noti cation and their duration, and general physical and mental health and problems/di culties post suicide. Depression and anxiety were assessed through self-reported data on the Patient Health Questionnaire (PHQ-9; (25,26)) and the Generalized Anxiety Disorder (GAD-7; (27-29)), respectively. During the interview, the audit team member explored whether other SBS might be interested in participating in the study, as more than one SBS could be interviewed regarding the same case. Names provided were contacted and invited to be interviewed.
Unmet needs were identi ed from needs expressed by the SBS themselves during the interview as well as from recommendations formulated by the systematic audit panel, which included a representative for bereaved families and the representative of a suicide prevention non-government organization (NGO).

Data Analysis
JR used content analysis (30) to identify unmet needs and group them in a small number of categories. FL and AL con rmed the ndings. Descriptive statistics compiled included raw numbers, percentages, means and standard deviations. Analyses were carried out on SAS 9.4 software.

Results
For the purposes of our study, 29 survivors bereaved by 25 victims who died by suicide in the east end of the island of Montreal in 2016 agreed to be interviewed in 2018-2019.

Quantitative Results
The mean age of the participants was 57.7 years (SD = 13.1). Of the 29, 23 were women. The average rating for relational proximity to the individual who died by suicide on a scale of 0 (no relationship) to 10 (highest relational proximity) was 8 (SD = 1.9). Other sample characteristics are given in Table 1.
In most cases (51.7%), the SBS were informed of the suicide by police o cers. Upon doing so, some o cers offered help verbally or handed over a suicide prevention center information card, at their discretion. Of the ten SBS who went to general practitioners for help, eight found it useful and four were prescribed medication. Further details regarding death noti cation, offers of psychotherapy/counseling, and utilization of professional, community organization and social network support and services are presented in Table 2.
The vast majority of the SBS did not screen positively for depression or anxiety at time of interview on the PHQ-9 and the GAD-7. The mean scores (SD) on the instruments were 4.56 (4.08) and 2.63 (3.54), respectively. A score of 10 or more is indicative of high risk for major depression and anxiety disorder. Table 3 presents the various emotional experiences reported by the SBS in the two years post suicide, which included shock, sadness and anger. In addition, of the SBS who witnessed the suicide (n = 2) or found the deceased (n = 9), three reported insomnia and six had ashbacks. Moreover, 22 SBS did not feel rejected by the loved one who died by suicide and 17 still wondered why their loved one committed suicide but without the questioning being obsessive in any way. Some SBS also reported developing a somatic condition (n = 8/29) or a psychiatric disorder such as anxiety or depression (n = 3/29). Furthermore, 3/29 increased their alcohol consumption and 5/29 increased their tobacco consumption post suicide. One of these last ve also developed depression.

Qualitative Results
Six SBS who did not receive professional help post suicide felt they might have bene tted from such help. In all, 22 SBS would have liked to receive a call from a professional within 66 days (SD = 65) on average of the suicide. Table 4 summarizes the main needs to emerge from the content analysis of the SBS interview data, indicates who should meet these needs, and presents the recommendations formulated by the audit panel for each need. Some of these needs were considered met by the SBS, others were not. However, either way, the SBS and/or the audit panel deemed that they should inform future postvention for SBS.

Discussion
Our ndings show that many SBS needs were not met or failed to be addressed by existing bereavement programs. In the majority of cases, the SBS learned of the suicide from police services. However, one third of the SBS were the ones who found the deceased. Offers of help came from different sources, including police o cers, funeral homes, and general practitioners. However, nearly half of the SBS were offered no help. The multiplicity and diversity of offers indicate that there is no systematic program for SBS and that the help offered is at the discretion of the person offering the help. Under the circumstances, SBS are likely to be treated differently and unevenly. Fortunately, the majority of SBS found support in their social network and even if they went through a hard time initially, they were not a icted by pathological bereavement, depression or anxiety two years post suicide. However, some SBS would have liked to be contacted by telephone by a professional and six SBS who received no help post suicide felt that they might have bene ted from receiving such help. This underscores the importance of being proactive with this vulnerable population from time of suicide death noti cation to at least one year post suicide. The needs and recommendations that emerged from this study underscore just how important it is for SBS programs to be interprofessional and to involve a broad array of stakeholders, including the QCO, law enforcement services, NGOs and the provincial ministry of health and social services.
Though results shed fresh light on the experience of SBS, our study is not without limitations. First, this was a retrospective study prone to memory and reconstruction biases. However, most of the SBS interviewed were able to describe their emotional experiences two years post suicide and, as is often the case, ordeals tend to render memories more vivid. Second, although all interviews were recorded, they were not transcribed. SBS needs were garnered from vignettes constructed by the research team for each case. Finally, the study focused only on SBS mentioned in police reports; other family members, friends and colleagues were neither identi ed nor contacted. However, the research team sought to contact other relatives of the deceased as well by asking direct family members whether anyone among family and friends might want to participate in the study.
Our study also has its strengths. First, the recommendations were formulated by an audit panel of health professionals and SBS, as recommended by Gall et al. (2014). To our knowledge, SBS needs have never been assessed to this extent. Second, we were able to triangulate the data thanks to the various sources of data used. This enhanced the credibility, transferability and dependability of the study results (31). Third, all the SBS were interviewed systematically, including those who did not seek help. This is particularly interesting given that some SBS who did not seek help were able to identify unmet needs and make recommendations for bereavement programs.
The results of our study are consistent with the literature. Indeed, as underlined by Pitman et al. (2018) and McKinnon and Chonody (2014), programs need to be proactive. SBS would like to receive a call because it is sometimes very di cult to seek help. Moreover, help should be offered at different times given that some individuals would prefer being contacted by telephone in the days following the suicide death noti cation, others after four months, and others still after six months. Pitman et al. (2018) reported similar results.

Conclusion
Though various initiatives have been undertaken for SBS in Quebec in the past decade, many have dwindled over time and none has been applied systematically. These initiatives include having police o cers give SBS suicide prevention center information cards at time of suicide death noti cation, having suicide prevention centers systematically call SBS after receiving contact information from the QCO, and offering bereavement services and suicide prevention and postvention training to professionals in schools. To ensure the survival of these strategies and offer better help and support, a provincial program to systematically monitor SBS is necessary. Such a program needs to be proactive and monitor SBS for two years post suicide. Ideally, it should be developed collaboratively by the Quebec Ministry of Health and Social Services, a suicide prevention NGO and the QCO, and should be nanced jointly by the Quebec Ministry of Health and Social Services and the Quebec Ministry of Justice. It could then be monitored and evaluated by a coalition for suicide prevention that would include SBS representatives and researchers. Authors thank all the SBS who participated in the study as well as the QCO, the Quebec Suicide Prevention Association, and the Health and Social Services Centre of east-end Montreal.
Authors contributions AL, CD and FL were major contributors in research conception and design and IA, GF, CB and JR contributed the data for this paper. FL and GF were major contributors in data analysis and interpretation. FL, AL, JR, and MS were involved in writing the manuscript or revising key content critically. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors read and approved the nal manuscript.
Ethical approval and consent to participate

Competing interests
The authors declare that they have no competing interest Consent for publication Not applicable.
Author's information Not applicable.

Funding
Fabienne Ligier was a Quebec Suicide Research Network (RQSHA) postdoc. Gabrielle Fortin was a University of Montreal Faculty of Medicine summer research awardee. There was no speci c funding for the research but for all their work.

Availability of data and materials
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request. No support reported 2 (6.9) No information 1 (3.4) Table 3 Emotional   Offer postvention to healthcare teams.