Qualität von Befunderstellung, Beurteilung und Selbsteinschätzung bei der Auswertung von Radiographien des Thorax in einer Ebene in Abhängigkeit von der Verwendung eines anatomischen Betrachtungsschemas und der Auswerterexpertise

PURPOSE: To objectify effects of an anatomical viewing scheme on the respective correctness of (a) findings, (b) interpretations, and (c) self-assessments of readers in chest radiographs acquired in one plane and the assessment of other influencing factors. MATERIALS AND METHODS: In all, 20 radiologists with 3-60 months of full-time radiography experience evaluated 12 chest radiographs of varying difficulty: once with and once without using an anatomical viewing scheme with at least 1 month in between (n= 480). In consensus of 3 radiological experts (a) and (b) were determined by means of a current computed tomography. The self-assessment (c) of readers was queried. RESULTS: (a) Findings were either missed or not described in 21%. Another 20% were recognized, but incorrectly described, (b) 62% of interpretations and 31% of derived clinical consequences were wrong and (c) in 39% of items the readers overestimated themselves. Experts were faster and better than novices, but for the scheme usage no further significant differences were detected (p> 0.5, respectively). The most pronounced effect was found in comparison with the routine report produced by the joint evaluation of novices and experts being clearly superior even to the expert study results (a), (b) and (c) alone (p< 0.001, respectively). CONCLUSION: Reporting of chest Xrays acquired in one plane was often incomplete or even wrong, and the evaluators overestimated themselves, which was not influenced by the use of the anatomical viewing scheme. Since errors between the evaluators sometimes differed greatly, duplicate evaluation of the radiographs by two different radiologists, which is already the case in many training facilities, may possibly be advisable in general.