Risk and dose assessment methods in gamma knife QA.

摘要

Traditional methods used in assessing risk in nuclear power plants may be inappropriate to use in assessing medical radiation risks. The typical philosophy used in assessing nuclear reactor risks is machine dominated with only secondary attention paid to the human component, and only after critical machine failure events have been identified. In assessing the risk of a misadministrative radiation dose to patients, the primary source of failures seems to stem overwhelmingly, from the actions of people and only secondarily from machine mode failures. In essence, certain medical misadministrations are dominated by human events not machine failures. Radiological medical devices such as the Leksell Gamma Knife are very simple in design, have few moving parts, and are relatively free from the risks of wear when compared with a nuclear power plant. Since there are major technical differences between a gamma knife and a nuclear power plant, one must select a particular risk assessment method which is sensitive to these system differences and tailored to the unique medical aspects of the phenomena under study. These differences also generate major shifts in the philosophy and assumptions which drive the risk assessment (Machine-centered vs Person-centered) method. We were prompted by these basic differences to develop a person-centered approach to risk assessment which would reflect these basic philosophical and technological differences, have the necessary resolution in its metrics, and be highly reliable (repeatable). The risk approach chosen by the Livermore investigative team has been called the ''Relative Risk Profile Method'' and has been described in detail by Banks and Paramore, (1983).