Bleeding scans in patients with lower gastrointestinal bleeding of undetermined cause: Are they useful?

Bleeding scans using technetium-99m labeled red blood cells are routinely used for the detection of lower gastrointestinal bleeding (LGIB). However, there has been controversy about its usefulness in the detection and localization of LGIB. Methods: We reviewed the records of 89 patients with clinical evidence of acute lower gastrointestinal bleed of undetermined source who had bleeding scans. The following parameters were evaluated: 1) presence of a positive or negative scan 2) results of other diagnostic tests such as arteriography, endoscopy and surgery 3) the ability of the scans to accurately localize the source of bleed using arteriography, endoscopy or surgery as the gold standard. Results: There were 89 patients (53 males, 36 females, mean age 60.2 ± 21.8 years) who had 102 bleeding scans, with 45 (44.1%) patients with positive scans and 57 (55.9%) patients with negative scans. Of the 45 patients with positive scans, 39 (86.7%) were confirmed to be positive by finding a bleeding source at either arteriography, endoscopy or surgery. However, of the patients with negative scans, 19 (33.3%) were found to have a bleeding source by either arteriography, endoscopy or surgery. The bleeding scans had a sensitivity of 67.2%, specificity of 86.4%, positive predictive value of 86.7% and negative predictive value of 66.7%. Furthermore in 31 of 39 patients (79%) the bleeding scans accurately localized the source of bleeding as determined by arteriography, endoscopy or surgery. 3 patients who had an upper gastrointestinal source for the LGIB were all inaccurately localized by the scans. Conclusions: Bleeding scans are not very sensitive for the detection of LGIB and may be misleading if the source of bleeding is in the upper gastrointestinal tract. However when the bleeding scans are positive they are accurate in predicting the site of bleeding.