Participatory model building for suicide prevention in Canada

In 2015, 2.5% of the Canadian population 15 years and older reported having thoughts of suicide, 0.8% had made a plan, and 0.4% had attempted suicide in the previous 12 months [1]. Suicide is a behaviour that results from a complex interplay of factors, including biological, psychological, social, cultural, and environmental factors, among others [2, 3]. Multiple contributory factors that may lead to suicide are as diverse and multi-faceted as economic crises [4], access to means [5], and media reporting of suicide [6], and may be influenced by intra-individual factors such as pain and illness [7, 8]; hopelessness and guilt [9]; and problem solving deficits [10]. All of these contributing factors may vary by age, gender and other sociodemographic characteristics [1]. In response to this complex public health issue in Canada, An Act Respecting a Federal Framework for Suicide Prevention became law in Canada in December 2012 and was followed by the Federal Framework on Suicide Prevention in 2016 [11]. The objectives of “The Framework” are to reduce stigma and raise public awareness, connect Canadians to information and resources, and accelerate the use of research and innovation in suicide prevention [12].

Given the complexity of suicide, a range of prevention initiatives aimed at reducing risk among the general population (e.g. universal prevention such as installing barriers on bridges), subpopulations at heightened risk (e.g. selective prevention such as training first responders), and among individuals at high risk of suicide (e.g. indicated prevention such as treatment of mental illness) [2, 13] are generally used, although the strength of the evidence for particular prevention strategies may vary [2, 14]. In a number of areas of public health, conceptual models have been used to map out complex issues to bring to light the interconnectedness of prevention efforts. For instance, Allender et al. [15] developed a conceptual model of childhood obesity to create a multi-faceted understanding of how factors as diverse as pet ownership, marketing of processed foods, and income inequality interact dynamically to influence obesity levels. The development of these conceptual models provides an opportunity for knowledge to emerge and for consensus to be built while generating a tool that may be used to organize and communicate the complex factors and their interrelationships that influence an outcome of interest. These models can be used to identify leverage points and also important gaps in prevention, services, or policy.

One approach to developing a conceptual model of a complex public health issue is through a participatory model building workshop. Participatory model building has been used since the late 1960s to investigate complex issues as diverse as water resource management [16], sustainable development [17], and the relationship between primary health care and population health [18]. Rather than relying solely on evidence derived from traditional scientific research, this approach allows for a greater breadth and complexity of knowledge to be incorporated, which is important for the identification of the practical leverage points that are necessary to address complex issues [19] These workshops provide an opportunity for participants to develop shared understandings of a problem of interest, uncovering hidden assumptions, revealing differences in the language that people are using, and ultimately leading to intervention approaches that may be more effective and more readily implemented [20‐23].

Participatory model building workshops provide an opportunity for stakeholder engagement and to capture the insights of those who deal with the problem of focus on a day-to-day basis [24], leading to the development of models that may more accurately reflect the real world [25]. A failure to consider real world complexities may result in the implementation or continuation of interventions that fail to achieve the desired outcomes or even worsen problems [26]. An emerging application of participatory model building is for quantitative system dynamics modelling, which may be used to develop more accurate simulations of complex systems [26‐28].

The objective of this paper is to describe the process of implementing a participatory model building workshop to develop a new conceptual model for suicide prevention in Canada. The aims of the model building workshop were to (1) generate a conceptual model that expanded beyond conventional models of suicide, by incorporating upstream risk and protective factors that act at the level of individuals, families, communities, and societies, and (2) identify the prevention levers that may operate upstream and across different intersecting societal structures and systems in the Canadian context. The resulting model could help inform future public health efforts around suicide prevention, guide future research, and inform micro-simulations and modelling efforts aimed at quantifying the impact of population-based interventions [29].

Since the 1990s there have increasingly been calls for epidemiologists and public health practitioners to consider explanatory factors that extend beyond the behaviours and biology of individuals, with the recognition that a wide range of complex factors interact dynamically in a system to produce patterns of ill health in populations [33, 34]. More recently, developing a greater understanding of the macro-level determinants of health and creating methodological innovation capable of “tackling the central questions about how social circumstances shape the health of populations” have been described as two formative next steps in the advancement of epidemiology ([34], p. 845). While data are most readily available for individual-level factors that are more easily quantifiable and most proximate to suicide, such as thoughts of suicide (obtained through cross-sectional surveys) and self-harm related hospitalizations (obtained through hospital administrative databases), the consideration of such factors alone does not adequately capture the complexity of suicide and suicide-related behaviour [35]. The participatory model building workshop was used to capture the understandings of employees who work in suicide prevention or related areas in a federal public health organization, eliciting useful information about upstream risk and protective factors [19]. The addition of factors related to discrimination and oppression to the pre-existing list of risk and protective factors is important, particularly given the higher rates of suicide among some Indigenous populations in Canada [36, 37]. Together, these factors interrelate and operate at multiple levels, and the resulting draft conceptual model is an initial attempt to articulate this. It is recognized that this “enhanced” draft conceptual model is not a true representation of reality, but reflects the participants’ mental models of suicide and suicide prevention. The draft conceptual model can be used as a starting point for discussion, empirical testing, refinement, and adaptation for specific subpopulation groups.

While many of the participants enjoyed an opportunity to share their knowledge, it was challenging to determine the best way to structure a model that is composed of so many diverse factors which interact with each other in a multitude of ways. However, the challenging nature of this was expected, as “visualizing complexity is part of the aim of the exercise” [15]. Ultimately, the workshop planning team worked to consolidate and simplify the drafts of the model developed by workshop participants in order to capture the overarching themes while generating a useful tool to communicate concepts and identify upstream factors (macro-level determinants of health) of importance.

The draft conceptual model developed in this study is the first of its kind that represents a unique Canadian perspective on suicide behaviours and prevention that goes beyond what is currently available in the scientific literature. For instance, many conceptual models of suicide focus on particular subpopulations, such as those who live in rural settings [38] or are members of the military [39], typically emphasizing interactions between predisposing and precipitating factors [3] such as connectedness [40] or physical illness [41]. Our draft model, instead, is a reflection of the understandings of the participants in our model-building workshop, intended to apply broadly to the general population of Canada and incorporates risk and protective factors, based on the social determinants of health, at multiple levels. At the level of the individual, intersecting factors such as gender, religion, disability, and culture which may increase risk or provide protection through a variety of pathways are included [42‐45]. At the level of the family, research has indicated, for example, that family connectedness is a strong predictor of positive mental health outcomes among transgender youth [46], while child maltreatment is associated with adolescent suicidal behaviours [47]. Factors such as suicide contagion [48], school environments [49], and work environments (e.g. among first responders [50]) play out as risk and protective factors at the community level. At the societal level, our draft model incorporates means restriction laws [51], systemic racism [52, 53], and media reporting guidelines [6], among others, as important and intersecting risk and protective factors for suicide. The US Centers for Disease Control and Prevention has also developed a comprehensive model which they termed Socio-Ecological Suicide Prevention Model (SESPM), with accompanying recommendations for the establishment of multi-level social-ecologically based suicide prevention programs [38]. The model presented here is based on the understanding of those who participated in the workshop; as such, a workshop with different participants may have yielded a different model. Discussion and refinement of these models is invited [38].

The results of this study are subject to several limitations. The results of this model building workshop are limited to the perspectives of staff of a federal public health organization that are policy analysts, program staff and surveillance analysts and did not include suicide prevention service providers or other individuals who work at regional or local levels and who may have different knowledge and perspectives. Members of the community were also not included as participants in this workshop. While people with lived experienced were not explicitly included among the participants, it is possible that some workshop participants also brought this perspective to the workshop. The draft model that was developed over the course of the day was limited by time and the energy of the workshop participants. This draft model was developed specifically with regards to the general population and further targeted work and consultation would be required to generate models appropriate for specific populations, particularly those which experience inequalities in suicide risk due to factors including structural violence, marginalization, and historical trauma.

The participatory model building workshop provided an opportunity for staff to share their knowledge of suicide and suicide prevention in Canada through a structured approach. This workshop yielded a draft conceptual model that contextualised upstream factors that can be used to better understand suicide prevention efforts in Canada. The conceptual framework that emerged from this workshop could be used to inform suicide prevention research in the Canadian context. It also presents an excellent opportunity to bridge the gap between policy, programs and research.