Qualitative modelling of social determinants of health using group model building: the case of debt, poverty, and health

Abstract

Background Social determinants of health (SDoH) are known to have a large impact on health outcomes, but their effects are difficult to make visible. They are part of complex systems of variables largely indirect effects on multiple levels, constituting so-called wicked problems. This study describes a participatory approach using group model building (GMB) with stakeholders, in order to develop a qualitative causal model of the health effects of SDoH, taking poverty and debt in the Dutch city of Utrecht as a case study. Methods With GMB we utilised the perspective of stakeholders who are directly involved in policy and practice regarding poverty, debt, and/or health. This was done using system dynamic modelling, in three interactive sessions lasting three hours each. In these sessions, they constructed a model, resulting in a system of variables with causal relationships and feedback loops. Subsequently, the results of these GMB sessions were compared to scientific literature and reviewed by a panel of researchers with extensive experience in relevant scientific fields. Results The resulting model contains 71 causal relationships between 39 variables, 29 of which are present in feedback loops. The variables of participation in society, stress, shame, social contacts and use of services/provisions appear to hold prominent roles in the model’s mechanisms. Most of the relationships in the model are supported by scientific literature. The researchers reviewing the model in the scientific meeting agreed that the vast majority of relationships would concur with scientific knowledge, but that the model constructed by the stakeholders consists mostly of individual-level factors, while important conditions usually relate to systemic variables. Conclusions Building a model with GMB helps grasp the complex situation of a wicked problem, for which it is unlikely that its interrelationships result in a fully intuitive understanding with linear mechanisms. Using this approach, effects of SDoH can be made visible and the body of evidence expanded. Importantly, it elicits stakeholders’ perspectives on a complex reality and offers a non-arbitrary way of formulating the model structure. This qualitative model is also well suited to serve as conceptual input for a quantitative model, which can be used to test and estimate the relationships.

Applications of Participatory System Dynamics Methods to Public Health

“…but also facilitates equitable and social learning processes among stakeholders. 25,27,28,30,34,[36][37][38][39][40][41] Causal loop diagrams are compelling because they offer the unique capability of conveying complex information effectively and parsimoniously and may also be more engaging and interesting than traditional graphs and charts. 37 The act of engaging in visual storytelling by building CLDs helps stakeholders understand the "potential for nonlinear relationships between contributing factors and unanticipated emergent behavior of complex systems," 38(p2) which allows for increased ability to identify leverage points for intervention that may not otherwise be obvious.…”

“…29,42,61 Subtheme 1.4: Enhanced research and community trust Many articles reported on the ability of PSD methods to enhance relationships between research institutions and community partners. 30,38,40,44 The collaboration, engagement, and group-based participation involved in engaging in PSD strengthen trust and relations between researchers and community partners. 30,38,40,44 This phenomenon itself acts as a feedback loop, as increased trust in research institutions may inspire community partners to wish to engage in future PSD research, which then produces more capacity to sustain change from communitybased interventions.…”

“…30,38,40,44 The collaboration, engagement, and group-based participation involved in engaging in PSD strengthen trust and relations between researchers and community partners. 30,38,40,44 This phenomenon itself acts as a feedback loop, as increased trust in research institutions may inspire community partners to wish to engage in future PSD research, which then produces more capacity to sustain change from communitybased interventions. 38,44 One of the more salient ways in which PSD methods build trust between researchers and stakeholders is by being intentionally responsive to existing power dynamics, as well as power dynamics that may be present during the modeling workshops.…”

“…30,35,60 Subtheme 1.7: Opportunity for community capacity building Using PSD in community context strengthens community capacity for change. 31,35,36,40,44,62,73 Engaging in PSD in a community context strengthens community capacity to apply systems thinking for addressing future problems and finding solutions. 31,36,40,44,67,73 CBSD methods specifically "intentionally builds community capacity over time to apply system dynamics to identified complex problems independently from external researchers or consultants."…”

“…31,35,36,40,44,62,73 Engaging in PSD in a community context strengthens community capacity to apply systems thinking for addressing future problems and finding solutions. 31,36,40,44,67,73 CBSD methods specifically "intentionally builds community capacity over time to apply system dynamics to identified complex problems independently from external researchers or consultants." 44(p787) Not only does engaging in PSD offer communities the opportunity to build greater trust with research institutions, more capacity for future partnerships, and more capacity to sustain change from interventions 77 but participating in PSD also gives communities the tools to create their own interventions, decreasing the need for research support or presence.…”