Transparency and Documentation in Simulations of Infectious Disease Outbreaks: Towards Evidence-based Public Health Decisions and Communications


Joakim Ekberg, Toomas Timpka, Magnus Morin, Johan Jenvald, James M. Nyce, Elin A. Gursky, Henrik Eriksson


Computer simulations have emerged as important tools in preparation for outbreaks of infectious disease. To support the collaborative planning and responding to the outbreaks, reports from simulations need to be transparent (accessible) with regard to the underlying parametric settings. This paper presents a design for generation of simulation reports where the background settings used in the simulation models are automatically visualized. We extended the ontology-management system Protégé to tag different settings into categories, and included these in report generation in parallel to the simulation outcomes. The report generator takes advantage of an XSLT specification and collects the documentation for the particular simulation settings into abridged XMLs with summarized results and settings. We conclude that even though inclusion of critical background settings in reports may not increase the accuracy of infectious disease simulations, it can prevent misunderstandings and less than optimal public-health decisions.