A new knowledge-based risk control method for risk sensitive devices

Medical engineering is always closely linked to the well-being of the human.This close relation can strike out at two directions: although medical devices are intentionally designed to support diagnosis and therapy, they can also cause serious adverse events and harm patients, users and third parties.Therefore, according to ISO 14971, risk management -including risk identification, risk evaluation, risk control and market surveillance -is an important and inevitable chapter in medical device development.Unfortunately, the risk control process, which implies selection and application of countermeasures (generally through inherent, protective or descriptive safety measures), is not yet supported systematically and methodically.Therefore the Chair of Medical Engineering at the RWTH Aachen University has developed a methodological approach to generate appropriate countermeasures for given risks, helping to mitigate previously identified technical and human-induced errors or hazards in products and processes.The methodology uses a knowledge-base, reorganizing prior experience, from by now fourteen risk analyses of medical systems, comprising research and industrial risk assessments.Case-tailored categories from error-taxonomies allow the user to hark back to his antecessors' knowledge in a user-friendly manner.The methods' basic structure is built on the Theory of Inventive Problem Solving (TRIZ) and can be fed with further data in the future.Purely technical and system-inherent, as well as Human-Machine-Interaction errors, have been organized in thirteen error categories, filing 61 individual failure modes, which represent the former (root) causes and failures from the analyzed risk analysis data base.The different possible combinations of cause and failure are displayed