Secondary Logo

Journal Logo

Abstracts: ACCEPTED: GI BLEEDING

Gastrointestinal Bleed: A New Look at an Old Trend

587

Chugh, Rishika MD1; Beauvais, Jacques MD1; Luther, Sanjana MD1; Tokayer, Aaron MD2

Author Information
American Journal of Gastroenterology: October 2017 - Volume 112 - Issue - p S316
  • Free

Introduction: Blood urea nitrogen (BUN) to creatinine ratio is a common surrogate for identifying upper gastrointestinal sources of bleeding. Several ratios (18:1, 30:1, 36:1, 100:1) have been previously studied for this purpose, with variable sensitivities and specificities. To date, there is no consensus on the best BUN to creatinine ratio threshold for determining upper gastrointestinal bleeding. Our goal is to determine how to best utilize these ratios and identify the ratio with the best discriminatory value.

Methods: 135 subjects with presence of gastrointestinal bleeding between 2008 and 2016 were identified utilizing an institutional research database (Clinical Looking Glass) at Montefiore Medical Center. We included individuals with confirmed gastrointestinal bleeding, baseline glomerular filtration rate (GFR) greater than 60 mL/minute/body surface area (BSA), and drop in hemoglobin greater than 1 gram/dL from baseline. Data collected included baseline and change in hemoglobin concentration, creatinine, and BUN, mode of diagnosis (esophagogastroduodenoscopy [EGD], push enteroscopy, capsule endoscopy, colonoscopy, CT angiography, standard angiography, nuclear bleeding scan, laparoscopy, or laparotomy), and baseline clinical characteristics. BUN to creatinine ratios were calculated for day 1 and day 2 of admission and were subsequently used to derive sensitivity and specificity ratios.

Results: Subjects were on average 67 years old, 79% of minority race, and 51% male. 61% of individuals had an upper gastrointestinal source of bleeding, with median hemoglobin of 9.4 g/dL (Interquartile range: 7.7, 11.4), and median change in hemoglobin of 3.4 g/dL (Interquartile range: 2.1, 4.5). On day 1 of admission, sensitivity and specificity, respectively, of identifying an upper gastrointestinal bleed utilizing the BUN to creatinine ratio was 80% and 52% for a ratio ≥ 20, 48% and 83% for a ratio ≥ 30, 32% and 90% for a ratio ≥ 36, 24% and 96% for a ratio ≥ 40, 16% and 98% for a ratio ≥ 50, and 10% and 100% for a ratio ≥ 60. Receiver operating characteristic curves were then created for day 1 and day 2 of admission and revealed that a ratio ≥ 20 best differentiated a lower gastrointestinal bleed from an upper gastrointestinal bleed.

Conclusion: A BUN to creatinine ratio greater than or equal to 20 differentiates an upper gastrointestinal bleed from a lower gastrointestinal bleed with the greatest accuracy. This ratio has a sensitivity of 80% and specificity of 52%.

© The American College of Gastroenterology 2017. All Rights Reserved.