RESUMO
The radiological examination frequency, i.e. the number of examinations performed annually, is necessary for estimating the collective effective dose of the population from medical exposures with ionizing radiation. Examination frequency surveys usually collect data from a limited number of radiological facilities participating in the survey. The collected data are then extrapolated to the existing radiological facilities in a country/region. Thus, the number of facilities and the specific facilities to participate, as well as, the extrapolation method used, are significant elements when designing the survey sample and methodology for examinations frequency assessments. This work attempted to simulate the situation when examination frequency data are collected from a limited number of facilities by investigating several "virtual sample" designs and two extrapolation methods. Comparisons between the calculated - by extrapolation - and the actual examination frequency in the country were made, for several scenarios and examination type data sets. The uncertainties were estimated and discussed thoroughly. The findings of this work highlighted the need for appropriate registry of the existing facilities in a country/region, the categorization of facilities considering the medical sector pattern in the country/region, the representativity and homogeneity of the samples used for a survey, as well as, the necessity for quality control of the collected examination frequency data. The results showed that when the aforementioned conditions were fulfilled, the examination frequency could be calculated with reasonable accuracy, based on data collected from a limited number of facilities. The paper also provides suggestions and tips for the collection and analysis of examination frequency data.
