Presenting statistical uncertainty on cytogenetic dose estimates

Cytogenetic dosimetry results are customarily accompanied by 95% confidence limits and the derivation of these limits is discussed. Simply presenting values of the best estimate of dose and 95% limits assumes that recipients of dosimetry reports, possibly the patients themselves, have a sufficient grasp of statistics to comprehend the concept of uncertainty. Two further methods have been developed to aid improved understanding for the lay person. The first method, particularly useful for suspected low doses, presents a simple odds ratio indicating that the results favour either zero dose or that suspected; for example, from an unexpected dose recorded on a badge. The second method, described fully here for the first time, uses a graphical illustration of a normalised distribution of dose so that one may visualise the area beneath a curve that represents the probability that the dose lies within various bounds. Marta Szluińska, Health Protection Agency. Crown