| Summary: | Harm from medical errors costs tens of billions of dollars and causes tens of thousands of deaths each year in the United States alone. This is not so surprising considering the lack of opportunity for medical students and consultants alike to practice rare events, to be systematically exposed to common scenarios, or to be objectively assessed. Similarly, there are limited opportunities to test new medical devices or procedures without putting patients at harm. Simulation has provided the airline industry, in particular, with such opportunities and is a contributing factor to the safety of air travel. Simulation in medicine has the same potential but there are few, if any, concrete standards to adhere to. My objectives were to provide a structure for such standards to be set, to develop methods for evaluating the modelled physiology of simulators, and to further the mathematical modelling needed for autonomous, realistic, and extendable simulators. To these ends, I have analysed the key components of simulation and reviewed existing simulators and the modelling which underpins their responses to interventions. A framework for standards was developed with a focus on the physiological modelling of anaesthetic simulators. Methods for evaluating the repeatability and concordance of simulators were explored using simple interventions. I created an extendable database of accurate, complete physiological and interventional time series data from anaesthetic cases. Some of these cases were used to confirm the repeatability and concordance results, and then used to develop more advanced methods for evaluating fidelity. Finally, I used a novel modelling approach to create an integrated model of the human cardio respiratory system encompassing cellular through to systemic physiological processes which produced promising results. It is my hope that the work in this thesis may pave the way for more realistic simulators and a more standardised approach to simulation so that medical errors are reduced.
|
|---|
