No Benefit for Nasogastric Tube in Uppper GI Bleeding

May 6, 2014
Jeffrey Hertzberg, MD, MS

New study data suggest that routine placement of such a tube is not helpful in patients with upper GI bleeding.

I have to admit it: When I was in full-time internal medicine practice, I always hated placing a nasogastric (NG) tube. But it was a time-honored ritual for patients with upper GI bleeding. To have omitted it in residency would have been viewed as the ultimate in shirking. I never omitted the NG tube, and I carried the tradition from training into my practice.

Why did I hate it? Because patients very obviously hated it even more. They were typically not sedated (probably for the best, because the NG tube can induce vomiting), so this was very, very uncomfortable. And like many traditional maneuvers in our tool kit, this one appears not to stand up to rigorous scientific testing of its value.

Rockey and colleagues looked at the use of an NG tube in the diagnosis and treatment of upper GI bleeding at Digestive Disease Week 2014 in Chicago. The investigators note that NG tube placement, aspiration, and lavage remain part of the standard of care in the evaluation and treatment of patients with upper GI bleeding. They also note that “its clinical utility remains unproven.” Their study found that the use of this uncomfortable procedure does not predict which patients need esophagogastroduodenoscopy (EGD), suggesting that its use should no longer be standard for patients with upper GI bleeding.

This was a single-center, single-blind, randomized, prospective, noninferiority study in patients with acute upper GI bleeding presenting with hematemesis and/or melena. Patients were randomized to no NG tube placement or NG tube placement with aspiration and lavage. EGD was performed within 24 hours. Physicians completed a validated questionnaire and a Likert Scale questionnaire that required them to predict the need for endoscopic therapy both pre and post NG tube; the primary outcome was the ability of the NG tube to predict the need for endoscopic therapy, using a noninferiority design.

Of 280 patients, 140 were randomized to each arm. The average age of study participants was 50 years; 35% were women. Hematemesis alone or with melena was present in 68% of patients. The groups were evenly matched in terms of pre-endoscopy vital signs, Blatchford score, hemoglobin level, blood urea nitrogen level, and other clinical features.

Physicians accurately predicted the need for endoscopic therapy in 32% and 38% of patients in the control and NG tube arms, respectively (P=0.7). Further, NG tube placement led to a change in the physicians' prediction of a bleeding lesion from “likely to need therapy” to “unlikely to need therapy” or vice versa in 41/140 (29%) of patients and an absolute change in their prediction of the need for endoscopic therapy of more than 20% total in only 21/140 (15%) of patients. The most common lesions identified included gastroduodenal ulcer (30%), esophageal varices (17%), esophagitis (17%), Mallory-Weiss tear (5%), gastritis (5%), and portal hypertensive gastropathy (5%).

Endoscopic therapy was performed in 31% and 34% of patients in the control and NG tube arms, respectively (p=0.7). Pain, nasal bleeding, or NG tube failure occurred in 49/140 (35%) of patients. Rebleeding rates in the control and NG tube groups were 5.0% and 4.3%, respectively (P=0.8); mortality was 2.9% in each group (P=1.0).

The authors concluded that routine placement of an NG tube (with aspiration and lavage) in typical patients with upper tract bleeding did not assist in predicting which patients had lesions that required endoscopic therapy. Further, NG tube placement caused pain, nasal bleeding, or was unsuccessful in more than one-third of patients and had no effect on outcomes. The data suggest that routine placement of an NG tube is not helpful in patients with upper GI bleeding.