In the hospital setting, upper gastrointestinal (GI) bleeding remains the most common reason for consulting a gastroenterologist. The care of patients with upper GI bleeding has recently changed, including the approach to initial management, timing of endoscopy, use of blood products, and available techniques for endoscopic therapy. I am going to provide an overview of my current approach to the management of upper GI bleeding with an emphasis on new concepts and techniques.
The key to initial management remains adequate resuscitation concurrent with the initial assessment. Two large bore intravenous lines (≥ 18 gauge) should be placed. Initial assessment should help evaluate and stratify patients into low- and high-risk categories. Use of nasogastric tubes is no longer recommended. However, current guidelines recommend the use of risk stratification scores to determine patient risk.
The most widely used pre-endoscopic scores are the Glasgow–Blatchford bleeding score (GBS) and the albumin (< 3.0 mg/ dL), international normalized ratio (> 1.5; INR), mental status, systolic blood pressure (< 90 mm Hg), and older age (> 65 years; AIMS65) score. The GBS is a weighted tool that includes eight categories and results in a score between zero and 23. Patients with a low-risk GBS rating have scores of zero to one and may be cared for as outpatients, provided that they are adequately followed-up, including undergoing endoscopy within three days. Patients with a GBS of seven or more should be considered to be at increased risk, and those with a GBS between 10 and 12 are at high risk for rebleeding and death. The AIMS65 score is a five-item, nonweighted scale in which a score of at least two is considered to be high risk; increasing scores are associated with increased mortality. I calculate both of these scores as part of my initial assessment and combine the results of these scores with my clinical judgment to determine triage and timing of endoscopy for each individual patient.
Recent studies have shown that the optimal timing of endoscopy associated with best outcomes is between six and 24 hours in most patients.
Current guidelines recommend performing endoscopy early (≤ 24 hours of presentation) for patients hospitalized with upper GI bleeding. Performing endoscopy within 24 hours results in overall improved patient outcomes. However, some low-risk patients do not require early endoscopy, such as those with a GBS of zero to one. In contrast, patients with high-risk, upper GI bleeding may benefit from emergency endoscopy within eight hours, especially to stop active bleeding. However, increased adverse events may occur with emergency endoscopy, especially during off hours (e.g., in middle of the night), possibly due to under-resuscitation, inadequate support staff (e.g., nurses, technicians), the unavailability of specialists (e.g., surgeons, angiographers), or decreased physician performance. Recent studies have shown that the optimal timing of endoscopy associated with best outcomes is between six and 24 hours in most patients. I perform emergent endoscopy in patients for whom I suspect have ongoing active bleeding, but only after adequate resuscitation.
The use of blood products has changed with the overall decreased use of blood products and better outcomes. The goal for hemoglobin transfusion for packed red blood cells is 7.0 g/dL unless active coronary disease is present, as this has been shown to reduce mortality compared with higher thresholds. Patients with mildly to moderately elevated INRs should undergo endoscopy, because endoscopic therapy is effective up to an INR of 2.5. However, if INR reversal is needed, use of prothrombin complex concentrates provides effective low volume and rapid reversal. Platelet transfusions are reasonable with a platelet count of less than 50,000/μL. However, in patients on dual antiplatelet therapy who have normal platelet counts, platelets transfusions have not been shown to improve outcomes.
Several important new advances in endoscopic therapy are now available. The FDA just approved a hemostatic spray (Hemospray®, Cook Medical, Bloomington, IN) for treatment of active GI bleeding. Hemostatic sprays have been shown to be 90 percent effective in stopping active GI bleeding, including bleeding caused by peptic ulcers as well as in complex situations such as bleeding due to malignancies. In high-risk lesions, a second modality such as hemoclips may be needed to prevent rebleeding. Large, over-the-scope clips have been shown to be effective at controlling bleeding, both as first-line therapy for large vessels and for rebleeding. Use of Doppler probes may decrease the risk of rebleeding in high-risk lesions, and their use to assess end points of endoscopic therapy (cessation of blood flow) has been shown to decrease rebleeding. In my practice, I selectively use Doppler probes in high-risk lesions; I use over-the-scope clips to treat large caliber arteries; and have started to use Hemospray.
The incorporation into clinical practice of recent, novel concepts and techniques for the management of upper GI bleeding will lead to better patient outcomes, including reduced morbidity and mortality.
Dr. Saltzman is an ACG Trustee.