Postgraduate research opportunities Validating hybrid simulation models for policy & decision-making in health systems

The COVID-19 pandemic has highlighted the need to validate mathematical and computational simulation models used for policy analysis and decision-making. We must be able to trust these models’ results as they inform decisions that significantly affect populations. Yet, the credibility of many COVID-19 models has been called into question. Model validation is framed by the model purpose. The proposed studentship focuses on models for policy and decision-making in health systems, and thus, the sufficient level of validity of a model should factor the urgency, magnitude, and scope of the decision the model informs. While the importance of some of these elements is at the purview of the decision-maker, as the technical experts, modellers require a framework and methods to evaluate that level of validity and convey it to the decision-makers.

The modelling research communities—e.g., systems simulations, health economics, and mathematical epidemiology—have developed methods that bring together multiple simulation methods but have only explored the validity of models using single simulation approaches. Hybrid models—that combine simulation methods—use is increasing as the complexities of real-world problems mean that single simulation methods cannot address all the issues that need to be tackled, and they have become prominent for studying many health system contexts. Combining simulation modelling methods introduces intricacies that complicate validation such as how to build trust in the linkages between the different simulation models being used.

While it is tempting to think that combining single simulation validation approaches is sufficient, only utilising these implies the overarching hybrid model and the links between the sub-models are not considered. Systems simulation methodology and complex adaptive system theory suggest the systems’ parts do not necessarily provide insight into how the whole system will behave, and the same is true for hybrid models.

The proposed studentship will develop a framework and methods to validate hybrid models. The application area will build on the department’s ongoing work in areas such as COVID-19 prevention and control in care homes or health systems strengthening in low- and middle-income countries and ensure these methods are providing trustworthy results.

Further information

This work builds on research conducted in the department in recent years, including theoretical and applied research projects that have supported Scottish and UK COVID-19 policy and work with global health organisations and low and middle-income countries' government bodies. In this project, we will similarly work with external partners in these organisations as relevant.