Modelling and simulating the performance of user behaviour in serious contexts

Abstract

Real-time experiments on healthcare procedural improvement can be infeasible due to the domain’s criticality and sensitivity. For instance, high morbidity rates and escalated patient treatment duration can, in some circumstances, be associated with medical resources exhaustion. Thus, formal methods can be an answer to lower the effects of experimenting within these healthcare domains as such an approach may be effective in deriving new insights and proposing further recommendations to the investigated domain. Specifically, performance modelling formalisms provide a rich theoretical foundation for dynamic systems, which are affected by an extensive collection of interventions, and supported by the existing formalisms tool sets. Hence, investigating healthcare system contexts involves several complex challenges. These challenges range from data collection methods and data analysis formalisms to optimising medical outcomes. This optimisation is beneficial to behaviour analysts and medical administrators. The current thesis contributes to addressing these challenges in many different ways: (i) By presenting an improved web-based version of a sketch simulation that collects the clinician behaviour during massive bleeding scenarios. This unconventional data collection method is proposed to minimise the need to observe the interventions in person where such treatment of these medical cases are performed; (ii) The modelling of two medical scenarios using different modelling formalisms for analysis and evaluation purposes, these modelling formalisms are Performance Evaluation Process Algebra (PEPA), Collective Adaptive Resource-sharing Markovian Agents (CARMA), and Stochastic Petri nets (SPN); (iii) A proposed tool to enhance the log analysis process. Doing so required the implementation of a trace-driven simulation tool. The tool simulates a clinical behaviour that has been recorded using a sketch simulation version. (iv) Proposing different suggestions to improve the medical outcomes and to effectively reduce the cost of health resources.