Hematemesis in the newborn period can be alarming for parents and doctors. Let’s look specifically at bloody emesis only in the newborn period – age less than 8 weeks. Obviously the initial questions should focus on the quantity and character of the vomiting. Was it spontaneous? After feeds? Streaks of red versus coffee grounds? Also, you’ll want to know if the infant is breast or bottle feeding as well as whether or not there is any history of familial milk protein intolerance or allergy. Babies born at home might not have received Vitamin K, and thus are risk for hemorrhagic disease of the newborn.

Hematemesis, defined as the vomiting of blood, can range from fresh and bright red to old and dark (due to the effect of gastric acidity) with the appearance of “coffee grounds.” If the baby looks ill, focus on the ABCs, IV access and resuscitation. If you are unsure whether or not it is actually blood you can do a gastric occult blood test. Significant bleeding may merit the placement of an NG tube for a diagnostic saline lavage if the patient is stable enough. Again, this maneuver is not warranted for “streaks” of red blood, or just a singe episode. Lavage can help you confirm the rate and level. It does not halt or control hemorrhage – and a clear aspirate doesn’t rule it out either. 50ml of saline per lavage is appropriate for infants, whereas 100-200ml is for older children. Again, the rate of bleeding depends on what you see:

Red flecks or coffee grounds – LOW rate of bleeding

Bright red blood – FAST rate of bleeding

The rate is thought to be especially fast if the blood doesn’t clear with repeated lavage for 5-10 minutes. You really don’t need to perform lavage for greater than 10 minutes. Afterwards leave the tube to low wall suction to reduce the irritative effect of gastric blood. This is not something that I have done frequently. This is because I have access to Pediatric Gastroenterologists in the hospital.

Differential diagnosis

Especially in the well appearing newborn hematemesis is often from swallowed maternal blood. In the first few days of life this could be from delivery, and later on from the mother’s breasts in the nursing infant. It is important to examine the mother’s nipples for cracked/bleeding skin. The best way to look for active nipple bleeding is to let the baby feed for a few minutes, then pull them away. If you see blood, reassurance is all that is needed. You could have the lab perform the Apt test if available. This involves mixing the bloody sample sodium hydroxide. Fetal hemoglobin won’t denature and will remain pink. Adult hemoglobin will become brown. Of course, if the infant hasn’t vomited that much you won’t have a large enough sample.

Blood can come from the nose if there has been frequent suctioning. Gastroesophageal reflux can result in hematemesis, but it is rare and generally seen only in babies that have more significant disease and/or vomit forcefully. Speaking of forceful vomiting, Mallory-Weiss tears can occur in babies too; they follow the acute onset of vigorous vomiting or retching at any age. These tears occur at the gastroesophageal junction due to a combination of mechanical factors (e.g., retching) and gastric acidity. The general assumption that a Mallory-Weiss tear is the cause of hematemesis is something I’ve heard countless times while precepting. This is a dangerous assumptoin to make in a newborn.

Another cause that is likely underreported is prolapse gastropathy. Per Pohl et al. prolapse gastropathy is “thought to be caused by forceful emesis when the stomach prolapses through the lower esophageal sphincter. Repeated prolapse of the stomach wall results in mucosal injury. The disrupted mucosal barrier has an increased susceptibility to gastric acid that causes vascular permeability and bleeding.”

Another condition that can present with hematemesis is hypertrophic pyloric stenosis. Those that vomit blood are generally pretty ill appearing by the time they hit your ED, but not always. Diagnosis begins with a high index of suspicion and is confirmed with ultrasonography.  Babies in shock/sepsis can develop stress ulcers and ischemic breakdown of the stomach. I won’t belabor the point that GI hemorrhage can worsen shock and thus NG/OG tubes can be used for more than just air decompression of the stomach in the intubated baby. Critically ill children of any age are at risk for developing stress-related peptic ulcer disease. Such ulcers occur with life-threatening illnesses, including shock, respiratory failure, hypoglycemia, dehydration, burns (Curling’s ulcer), intracranial lesions or trauma (Cushing’s ulcer), renal failure, and vasculitis. These ulcers may develop within minutes to hours after the initial insult and primarily result from ischemia. Hematemesis, hematochezia, melena, and/ or perforation of a viscus may accompany stress-associated ulcers. Hematemesis secondary to gastroesophageal reflux and esophagitis is uncommon but should be considered in patients who are severely symptomatic with vomiting or aspiration. Finally, the baby with perotinitis/acute abdomen could have necrotizing enterocolitis. This can occur in term newborns, but it’s pretty rare. It is more common in preemies.


As I noted earlier this begins with the ABCs. Get IV access and resuscitate the infant in shock. Blood transfusions are a tricky proposition in the newborn – and it is safe to start with 10ml/kg to 20ml/kg NS boluses. Get a CBC, Type and Screen, and an electrolyte panel at least, along with liver panel, coags and lipase and know that I-STATs can underestimate severe anemia. If you do need to transfuse you should also have already discussed with gastroenterology