Nasogastric Intubation: How Necessary Is It?

Amal Mattu, MD, FAAEM, FACEP


October 03, 2007

"You Wanna Do What?!" Modern Indications for Nasogastric Intubation

Witting MD
J Emerg Med. 2007;33:61-64

Nasogastric intubation (NGI) is widely regarded as one of the most painful procedures performed in emergency medicine. The nose is very highly innervated and a very uncomfortable part of the body to manipulate. Stimulation of the posterior pharynx often causes gagging and vomiting. Unlike many other procedures in emergency medicine, NGI is usually performed without procedural sedation because of the concerns of vomiting and potential aspiration if the patient is sedated. Aside from pain and discomfort, complications can occur as well, most notably nosebleeds. Although NGI has been used in medicine for almost 100 years, very little evidence actually exists supporting its use. On the contrary, and as demonstrated through this excellent analysis by Dr. Michael Witting published in The Journal of Emergency Medicine, recent evidence has suggested that this procedure is only of limited benefit especially in light of its potential complications and the pain associated with it.

Specifically, Dr. Witting looked at the use of NGI in patients with gastrointestinal (GI) bleeding, for gastric decompression, for the administration of specific treatments, and for the demonstration of anatomic position. He concluded with tips on how to limit pain and facilitate passage through the nasogastric tube. Here is a summary of the main takeaway points of Dr. Witting's analysis.

One of the most common apparent indications for NGI is the management of patients with GI bleeding. Theoretically, NGI can help localize the bleeding to the upper vs lower GI tract; it can gauge the severity and persistence of upper GI bleeding; and it clears blood from the stomach to facilitate upper endoscopy. The discussion on NGI for GI bleeding can be divided by whether or not patients have hematemesis.

Patients with hematemesis. NGI plays a limited role in localizing the site of the bleed in patients with hematemesis because most, if not all, have an upper GI source. The definitive procedure for localization of the bleeding source and for treatment is esophagoduodenoscopy (EGD). NGI can help in determining the quantity and persistence of bleeding, which may help in determining how rapidly EGD should be performed and the need for emergent transfusion. The presence of frank blood during nasogastric suctioning has a positive likelihood ration of 2.0 for the presence of high-risk lesions. However, there are other ways of determining ongoing blood loss, including examining the vomitus (bright red blood indicates brisk bleeding, whereas coffee-ground vomitus indicates that blood has spent some time in the stomach, ie, milder bleeding), serial hematocrit testing, and hemodynamic monitoring. Dr. Witting summarized, "Thus, although NGI can provide some useful information about ongoing blood loss, it is unclear what it adds to other information available at the bedside, and especially whether the information provided is worth the discomfort to the patient."

Traditional teaching also holds that nasogastric lavage can clear blood from the stomach and therefore facilitate EGD visualization. However, it is questionable whether gastric blood can actually be cleared properly through the small lumen of standard nasogastric tubes. Additionally, the lavage can be performed at the time of the EGD, which is a crucial point because the delay of the EGD caused by performing the lavage far in advance may be harmful. Of interest, the use of intravenous erythromycin, which speeds gastric emptying, has been shown to be effective in clearing stomach contents for EGD, whether or not NGI has been performed.

Thus, NGI has unproven benefits in patients with hematemesis.

Patients without hematemesis. A positive nasogastric lavage provides strong evidence of an upper GI source of bleeding (positive likelihood ration, 11), but because nasogastric tubes rarely pass the pyloric sphincter, a negative lavage is limited in ruling out an upper GI source (negative likelihood ration, 0.6). In fact, most NGIs performed in the absence of hematemesis give negative results. Stool color is probably the most important consideration in these patients. Black, tarry stool indicates a slower upper GI source, whereas bright red blood in the stool indicates either a distal source (85%) or a brisk upper GI source (15%). Although it may seem that NGI would be helpful to distinguish between these latter 2 possibilities, in fact it is rarely so. Patients without hematemesis but with a brisk upper GI source of bleeding almost always have a duodenal ulcer, a lesion that NGI rarely detects. Therefore, the nasogastric lavage is negative with both of these latter groups.

On the other hand, 3 factors -- black stool color, a blood urea nitrogen/creatinine ration ≥ 30, and age < 50 years -- independently predict an upper GI source in patients without hematemesis.

NGI has little utility in patients without any of these factors: Only 5% with none of these risk factors will have an upper GI source.

Gastric decompression is vitally important to prevent perforation due to bowel distension in patients who have pronounced proximal small bowel obstruction or gastric outlet obstruction. Gastric decompression is also useful prior to diagnostic peritoneal lavage. NGI can improve ventilation during or after aggressive bag-valve-mask ventilation, which commonly causes gastric distension due to air.

On the other hand, NGI has limited utility (and limited supporting evidence) in patients with mild small bowel obstruction, colonic obstruction with a competent ileocecal valve, and paralytic ileus. In these cases, NGI may help with accumulating upper GI secretions, which can result in vomiting, and aspirated air. However, in the absence of significant proximal distension and/or pain, the risks and discomfort of the procedure should be weighed against the limited benefits.

The use of NGI to decrease vomiting and provide bowel rest in patients with pancreatitis was commonplace for many years. Recent studies have demonstrated that NGI provides no benefit beyond fasting alone, and it can be deferred unless the patient has intractable vomiting despite antiemetics.

According to Dr. Witting, "NGI can be useful for delivering medications or contrast when patients are unwilling or unable to drink the necessary quantity." Suicidal patients who require whole bowel irrigation for removal of ingested substances, for example, may benefit from NGI for administration of the large volumes of polyethylene glycol. "Caution should be exercised in delivering treatment via NGI to obtunded patients" because of the risk for aspiration if the stomach is overfilled.

NGI and visualization of the nasogastric tube on x-ray may indirectly identify aortic aneurysms and dissection and diaphragmatic hernia. However, more advanced imaging technology, such as computerized tomography, is rapidly available in most emergency departments and far more accurate.

Given the pain and discomfort associated with NGI, healthcare providers should take steps to limit pain and facilitate passage on the first attempt. Atomization or nebulization of 4% lidocaine provides some topical anesthesia to the mucous membranes. The addition of a vasoconstrictor (unless contraindicated) to the lidocaine can also enlarge the nasal passages and reduce the chances of epistaxis. Warming the tube in one's hand or in warm water softens it. Only the tip should be warmed, however, to avoid making the tube too flexible, which can cause coiling in the distal pharynx. Finally, describing the procedure to patients in advance can improve their tolerance level for the discomfort associated with it.

NGI has been a routine part of the medical management of various conditions for many years. However, the actual evidence supporting NGI in these conditions is limited. A more selective approach to NGI is prudent, and when it is indicated, steps should be taken to minimize discomfort and maximize chances for successful passage on the first attempt.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.