Avaliação das estatísticas utilizadas nos controles de qualidade de equipamentos para radiodiagnóstico

Abstract

The objective of this study was to verify and analyze the adequacy of the statistics used in methodologies on quality control tests for the estimation of accuracy and precision of the tube voltage and exposure time variables. The methodologies recommended in the Portaria Federal 453/98, in the Guia Radiodiagnóstico Médico – Segurança e Desempenho de Equipamentos published by Anvisa, the Protocolo Español de Control de Calidad em Radiodiagnóstico and Protocolos de Control de Calidad em Radiodiagnóstico. The adaptation of the statistics was based on the definitions of the International Vocabulary of Metrology and on statistical concepts. In addition, they were compared between each other considering the percentage of non-conformity (or rejection) of equipment presented by the applied methodology. In this last evaluation, simulated data were used with the objective of obtaining empirical distributions. Differences were found between the statistics and tolerances indicated for the precision evaluation, performed by means of reproducibility and/or repetitiveness tests and, in the accuracy evaluations, in the tension of the tube and exposure time. Despite being in accordance with the statistical and metrological definitions, the different statistics presented conducted distinct percentages of non-conformity in the equipment for the same type of evaluation, pointing to statistics with higher sensitivity in the detection of discrepant values, which leads to the disapproval of a radiodiagnostic equipment. Different tolerances also influenced the disapproval percentage and led to different results for the same equipment. Thus, different conformity reports might be obtained of an equipment, leading to inconsistent results.