How close is close enough?

摘要

This is a non-trivial question when it comes to radiation dosimetry. The determination of absorbed dose or dose equivalent is often more than an academic exercise or a research project. In the medical application of ionising radiation, the difference between a dose delivered to a tumour being too low or too high can result in either a recurrence or complications due to late effects. In this case, the dose has to be just right. And, the degree of just-righted-ness may require the combined standard uncertainty (with no coverage factor) to be just a few percent. If the radiation is being used to sterilise disposable medical products, a similar situation can occur, although the required uncertainty may be somewhat larger. Too low a dose and the bacteria can re-grow. Too high a dose and the product could be damaged by radiation-induced embrittlement. In the field of radiation protection, one might also ask how close a measurement should be to the true or conventional quantity value. The true value of a quantity is not obtainable, but a conventional quantity value that is taken by formal agreement to represent the true value can be established by a calibration laboratory.