Detection of upper gastrointestinal blood with fecal occult blood tests

Objective: Although fecal occult blood (FOB) tests have most often been used to detect occult bleeding from the lower gastrointestinal (GI) tract, their utility in detecting occult blood loss from the upper GI tract is less well understood. The aims of this study were to determine whether small amounts of blood from the upper GI tract can be detected by currently available FOB tests and, if so, to correlate FOB tests with semiquantitative GI blood. Methods: Groups of 10 healthy volunteers without a history of GI disease drank 5, 10, or 20 ml of their own blood mixed with tomato juice for 5 or 3 consecutive days. Standard dietary and medication restrictions were observed. Consecutive stools were tested for 2 days before, as well as 4 days after, blood ingestion. Each stool was simultaneously tested for FOB with HemoQuant (HQ), Hemoccult II (HO II), Hemoccult II SENSA (SENSA), HemeSelect (HS), and FlexSure OBT (FS). Results: The mean age and hemoglobin concentration of the study population were 29.3 ± 0.5 yr and 14.3 ± 0.3 g/dl, respectively. No subject noted GI symptoms during blood ingestion. Fecal blood levels (measured by HQ) were elevated within 2 days after initiation of blood ingestion and remained elevated until 2–3 days after cessation of blood ingestion. Mean fecal blood levels peaked at 2.1, 7.9, 8.0, and 13.5 (mg hemoglobin/g stool) in groups ingesting 5 ml/5 days, 10 ml/3 days, 10 ml/5 days, and 20 ml/3 days, respectively. The proportion of positive tests during and immediately after the period of blood ingestion was greatest in the 20 ml/3 day group; 16% of HO II samples were positive as were 64% of SENSA and 67% of HQ samples. SENSA was more sensitive than HO II in all blood ingestion groups. At least one positive SENSA test was present in 50% of subjects ingesting 10 ml of blood (each 3 and 5 day groups) and in all subjects ingesting 20 ml/day. Immunochemical tests did not detect upper GI blood in any blood ingestion group. Conclusion: Inasmuch as many upper GI tract lesions have been reported to bleed small quantities of blood such as that studied here, and this amount of blood is readily detected with widely used guaiac-based FOB tests inlcuding Hemoccult II SENSA, the data emphasize that caution is warranted before attributing positive guaiac tests only to sites in the lower GI tract. The data raise the possibility that a combination of a highly sensitive guaiac-based FOB test plus an immunochemical could help differentiate occult upper from lower GI bleeding.