Early roentgen diagnosis in massive bleeding from the upper gastrointestinal tract: I. Clinical evaluation of safety and reliability of the method in 123 patients

Abstract

D ESPITE the value of roentgenology in the diagnosis of gastrointestinal tract lesions it has not been widely used in the past for the emergency diagnosis of the cause of massive upper gastrointestinal hemorrhage, largely because of the fear of precipitating renewed bleeding. Hampton, Schatzki and others,l-s however, permitted carefully restricted examination of selected patients and contradicted the prevalent conservatism. At the same time they emphasized that emergency roentgen diagnosis cannot be used indiscriminately. There is need for further clarification of its indications, contraindications and complications, as well as a critical evaluation of its diagnostic reliability. These are the objects of the present study. Emergency roentgen examination of the gastrointestinal tract of 123 severely bleeding patients was performed. The examination was earlier and more complete in almost all cases than that recommended by other workers. The patients were divided into two groups: Group A consisted of the first fifty-tw-o patients examined. These were subjected to special attention and precautions not regularly available to ward patients. They were attended by a large group of roentgenologists, surgeons, ward physicians and others. Decisions regarding management were made in consultation by the several specialists. Certain patients were carefully exeluded from examination. Patients were regularly accompanied to the x-ray examining room by a member of the clinical group. One resident roentgenologist (T. P. C.) performed almost all of the examinations. After return to the ward the patient was again followed closely by various members of the team. Group B consisted of the last seventy-one patients examined. These patients were studied with the object of determining the accuracy and safety of the emergency roentgen method when applied without the special supervision afforded patients in Group A. They were managed almost entirely by the resident staff. Selection of patients was made by the ward physician. They were not regularly accompanied to the x-ray department by a clinician. Eight resident roentgenologists participated in the emergency roentgen examinations; one roentgenologist (S. E. H.) performed forty of the examinations. Age and Sex. The ages of the 123 patients ranged from thirteen to eighty-eight, averaging fifty-four years. One hundred one of the patients were male, twenty-two female. Severity of Bleeding. All patients manifested gross hematemesis or tarry stools or both, and were considered medical emergencies at the time of admission. Many were admitted in shock. Patients with occult bleeding were not included in the study. One indication of the degree of blood loss suffered may be gained from the * From the Thorndike Memorial Laboratory, Second and Fourth Medical Services (Harvard) and the Department of Roentgenology, Boston City Hospital, and the Departments of Medicine and Roentgenology, Harvard Medical School, Boston, Mass. This study was supported in part by a grant from the U. S. Public Health Service.