Economic models – and specifically “early” models that aim to map what may be uncharted waters – are crucial to the work of health economists seeking to provide a guide to decision-makers about health interventions. In a health ecosystem of countless actors and variables, they are key to giving a picture of the potential costs and benefits of different policy and procedural options. But determining what elements and considerations to include in a model and which to leave out is a perennial challenge.
The factors and priorities guiding the development of such models are the central question of a new commentary, published in the International Journal of Health Policy and Management (IJHPM), by PCHSS investigators Andrew Partington and Prof Jonathan Karnon.
Drawing on research by Grutter et al., Partington and Karnon point to the fact that decisions about health policy and heath interventions rarely take an idealised ‘go’ or ‘no-go’ form. Rather, as such decision-making involves a variety of stakeholders, options tend to be revised over time and decisions usually end up being something closer to a choice between “not yet” and “yes, but (with conditions)”.
In light of this, Partington and Karnon highlight the usefulness of participatory, iterative economic modelling, which better takes into account the complex, adaptive nature of the health system, as well as the variety across organisational settings:
The promotion of iterative economic modelling for such support is nothing new, but its application has been limited. However, there appears to be renewed encouragement and receptiveness within health services to foster “learning systems,” whereby the complexities of a service and intervention are explicitly represented, and explored in an iterative process. Such a process facilitates the engagement of diverse stakeholders in the development of interventions through a common language and framework.
By actively eliciting stakeholder engagement throughout the modelling process, health economists can both better define the parameters of their models and build greater understanding of the real probabilities (and uncertainties) of different options and frameworks among those who will be impacted by the decisions. In this regard, they note that “it is not necessarily the model itself that is informative, but rather the modelling process.”