DSA in acute gastrointestinal hemorrhage: clinical and in vitro studies

Selective intraarterial digital subtraction angiography (DSA) was used to examine 37 patients with acute gastrointestinal (GI) tract bleeding. Conventional screen-film angiography was used as an adjunct to DSA when a larger field of view was needed (five patients) and when bowel motion prevented the acquisition of adequate image quality with DSA (two patients). Conventional angiography was also performed in all cases in which there were negative DSA examinations. DSA reduced the mean examination time considerably (20% reduction overall), especially for cases involving embolization therapy (35% reduction). DSA was especially valuable in the upper GI tract, where it was used to rapidly locate and/or assist in the embolization of bleeding sites in 19 of 20 patients with positive angiograms. There were 12 true-negative DSA examinations and one false-negative examination due to the limited field of view (9 inches [22.9 cm]). Bowel and respiratory motion were not important problems in the upper GI tract. In the lower GI tract, the usefulness of DSA was severely limited by the small field of view and the misregistration artifact caused by bowel motion. In an in vitro study, DSA and conventional angiography were compared as to their ability to depict several rates of extravasation of contrast material in a model of GI bleeding. DSA tended to be more sensitive for the detection of simulated extravasation (P less than .07).