RESUMO
This paper summarizes the experience from 1969 to 1972 of a mixed team of radiologists and endoscopists at the Northern Marseilles University Hospital Centre. The patients were unselected and the team were not given all the information. After recalling the clinical and radiological conditions, together with the endoscopic conditions of the examination, the authors define the criteria which they chose to orient patients towards fiber endoscopy, i.e. either frankly abnormal radiology, or a radiological examination classed as normal in 563 cases, but which the team considered doubtful. They then report their results. Under these conditions, the results seem to be less unfavourable for radiology than would have been thought, for only 7% of the results were made by endoscopy alone (96 cases out of 1,500). In both benign and malignant ulcer, fiber endoscopy simply confirmed the radiological diagnosis and, in 54% of all the examinations, fiber endoscopy confirmed the normal character of radiology. However, in deciding whether carcinoma or benign ulcer, i.e. 11% of the cases of carcinoma and 14% of the cases of benign ulcer, radiology completely omitted the diagnosis of the lesion which might seem unacceptable. They conclude by asking radiologists to improve their technique of lesion detection and concentrate their examination on the smallest visible lesion in areas considered diffcult to explore radiologically. They are more and more circumspect about giving a full etiological diagnosis of lesions by radiology, especially in gastric ulceration where fiber endoscopy and direct biopsy seems more and more advisable.