Search results

AbstractThis article describes a system dynamics analysis of the 1992 Westray mine disaster in Nova Scotia, Canada. It examines the causal structure of the Westray system, including relationships that could have led to conditions that caused the fatal explosion at the mine. The value of simulation is its ability to capture a "mental model" of the safety system, which can stimulate discussion among safety experts as to the systemic causes of a disaster. By taking into account feedback loops and non‐linear relationships, which is not possible with conventional root‐cause analysis, a dynamic model of the system provides insights into the complex web of causes that can lead to disaster and valuable lessons for organizational learning. Copyright © 2003 John Wiley & Sons, Ltd.

Inappropriate administration of radiation for cancer treatment can result in severe consequences such as premature death or appreciably impaired quality of life. There has been little study of vulnerable treatment process components and their contribution to the risk of radiation treatment (RT). In this article, we describe the application of probabilistic fault tree methods to assess the probability of radiation misadministration to patients at a large cancer treatment center. We conducted a systematic analysis of the RT process that identified four process domains: Assessment, Preparation, Treatment, and Follow‐up. For the Preparation domain, we analyzed possible incident scenarios via fault trees. For each task, we also identified existing quality control measures. To populate the fault trees we used subjective probabilities from experts and compared results with incident report data. Both the fault tree and the incident report analysis revealed simulation tasks to be most prone to incidents, and the treatment prescription task to be least prone to incidents. The probability of a Preparation domain incident was estimated to be in the range of 0.1–0.7% based on incident reports, which is comparable to the mean value of 0.4% from the fault tree analysis using probabilities from the expert elicitation exercise. In conclusion, an analysis of part of the RT system using a fault tree populated with subjective probabilities from experts was useful in identifying vulnerable components of the system, and provided quantitative data for risk management.