Gastrointestinal hemorrhage is one of the most serious presenting complaints a pediatric gastroenterologist may need to diagnose and treat. A proper understanding of various etiologies and available treatment modalities is necessary to assure successful outcome. Multiple review articles have examined these in detail, however, very little is known about the true epidemiology of pediatric gastrointestinal bleeding. The only studies examining this topic were performed in intensive care settings where the incidence of upper gastrointestinal bleeding was found to range from 6% to 25%.
There are many causes of gastrointestinal bleeding ranging from the ones requiring urgent treatment to those that resolve without any therapy. Determining which is occurring is the art of medicine and this chapter will attempt to convey the practice of this art by discussing a diagnostic approach and therapeutic options available to pediatric practitioners.
Melena is the passage of dark, black, tar-like colored stool usually associated with esophageal, gastric, or upper small bowel gastrointestinal hemorrhage.
Hematochezia is the passage of bright red blood or maroon colored material from the rectum usually associated with distal small bowel, or colonic gastrointestinal bleeding.
Hematemesis is the passage of bloody or dark brown (“coffee ground”) appearing material from the mouth usually associated with bleeding in the gastrointestinal tract proximal to the ligament of Treitz.
Initial Presentation and Assessment
The child or adolescent with gastrointestinal (GI) bleeding may present to medical attention in a wide variety of ways and locations including a local emergency room, a pediatrician’s office, a phone call from a parent, or an inpatient hospital ward. Regardless of the location, by asking several brief and pointed questions, the medical provider will be able to assess the need for immediate intervention and subsequently establish a specific diagnosis. These include ( Box 13-1 ):
Describe the location, duration, quantity, and appearance of the bleeding.
What is the physical appearance and vital signs of the patient, if available?
What other medical conditions does the child have?
What medication is the child on?
Open Ended Questions:
Describe the location, quantity, and appearance of the bleeding.
What are the physical appearance and vital signs of the patient (if available)?
What medical conditions does the child have? (Liver disease, inflammatory bowel disease, etc.)
What medication is the child on? (Anticoagulants, NSAIDs, etc.)
Description of onset, location, duration, occurrence
Exposure to raw food, reptiles, travel, or toxins
Foreign body ingestion
Exposure to others with similar such symptoms
Other associated symptoms (sores, pains, rashes, vomiting, swelling, headaches, neck pain, chest pain, diarrhea, fevers, bruising, infections)
Medications (NSAIDS, warfarin, hepatotoxins, antibiotics use)
Review of Systems:
Gastrointestinal disorders (polyps, ulcers, colitis)
Based on the answers to these questions one could approximate the severity of bleeding, where the patient needs to be sent, and what services need to be readied for therapy. For example, large volume bleeding, history of portal hypertension or coagulopathy, ill appearance, or unstable vital signs would prompt emergent care with rapid medical transport. On the other hand, if the answers do not suggest a need for emergent care or if the care plan has already started, further questions will help assess a possible cause. Specific areas that should be addressed include:
Description of onset, location, and duration of bleeding
Exposure to other individuals with similar symptoms or development of similar symptoms in contact
History of travel
Exposure to toxins, animals, or potentially contaminated food or water source
Foreign body ingestion
Ingestion of specific foods or medications
Other associated symptoms (mouth sores, abdominal pain, rashes, vomiting, swelling, headaches, neck pain, chest pain, diarrhea, fevers, bruising).
Each of the answers to the above questions must be taken in appropriate context with the understanding that in times of emergency parents and children may overestimate the amount of blood, duration of the event, or frequency. A good estimate of the amount of blood is the appearance of blood in the toilet. If the water is opacified (not transparent), it is likely significant. If the child is old enough to provide a reliable history, additional questions should be asked of him/her and if needed without parental presence. One should also remember that having one positive historical event does not forgo the possibility of another one that could be more serious. For example, a child who ingests blueberries that cause a melenic type stool could also have a bleeding duodenal ulcer that could cause melena.
Along with a directed history of the current events, a directed social and family history should be performed. These may direct the physician to specific conditions that could be presenting for the first time. Additionally, various foods and medications can cause stool or vomitus to falsely appear to contain blood ( Box 13-2 ).
Red: candy, fruit punch, beets, laxatives, phenytoin, rifampin
Black: bismuth, activated charcoal, iron, spinach, blueberry, licorice
Foods and medications that may cause abnormal color of stool: Red: candy, fruit punch, beets, laxatives, phenytoin, rifampin Black: bismuth, activated charcoal, iron, spinach, blueberry, licorice
Once the practitioner takes an appropriate history he/she should be able to place the child into one of two pathways: emergent and nonemergent. This should subsequently be substantiated by a physical exam and laboratory evaluation. The physical exam should also be similarly focused to direct the possible sick patient to appropriate care quickly. It should primarily assess a directed ABC approach of General A ssessment for Well or Ill, Evidence of Brisk B leeding (blood in the oral cavity or on rectal exam), and vital signs directed at C irculatory adequacy (heart rate, orthostatic vital signs, pulse pressure, blood pressure, urine output). The balance of the physical exam will analyze possible causes of bleeding such as splenomegaly for portal hypertension, lip and buccal hyperpigmentation for Peutz Jeghers syndrome, emaciation and perianal disease for Crohn’s disease, or abdominal distention and ill appearance for intussusception. A directed physical exam with appropriate areas of interest is outlined below.
Vital Signs: orthostasis, pulse pressure, instability, urine output
General: appearance (well or ill), fever, mental status
Head, eyes, ears, nose, and throat: trauma, scleral injection, petechiae, lip pigmentation, epistaxis, erythema or burns to posterior pharynx, bleeding
Chest/Cardiovascular: tachycardia, murmur, capillary refill
Abdomen: tenderness, splenomegaly, hepatomegaly, caput medusa, distention, ascites
Genitourinary: fistula, swelling
Rectal: gross blood, melena, tags, tenderness, fissure, fistula, swelling, warmth
Dermatological: pallor, jaundice, rash, arteriovenous malformations (AVM), bruising, petechiae
Complete blood count
Prothrombin time/international normalized ratio/partial thromboplastin time
Complete metabolic profile (electrolytes, liver function tests)
Type and screen
If massive loss: Type and cross and fibrinogen
Stool Culture with E. coli O157:H7 assay (Shigella, Salmonella, Yersinia, Campylobacter), C. difficile toxins A&B, and ova and parasite examination if indicated by history
Erythrocyte sedimentation rate, C-reactive protein, gamma-glutamyl transpeptidase if indicated by history
Heme occult and gastro occult testing
Serial CBCs, electrolyte studies, and fibrinogen depending rate of bleeding
Once the physical exam is complete, a targeted set of diagnostic laboratory studies should be requested. Similar to the history with a primary and secondary assessment, the laboratory studies should be ordered in a proper sequence. Unless the bleeding is obvious, testing for the presence of blood should usually be confirmed with the appropriate test kit. Except in cases where upper GI bleeding appears to be minor, consideration should be given to investigation with nasogastric lavage to assess the volume and for evidence of continued bleeding. This is performed with a double lumen vented pediatric-sized nasogastric tube. The tube is placed after adequate lubrication and topical anesthesia nasally and advanced into the stomach. Placement confirmation may differ from institution to institution, but commonly used methods include pH testing or air bubble technique. After the placement, the tube is irrigated with 50 cc aliquots of room temperature sterile water until the irrigant is clear. If one is unable to lavage, the tube might be positioned against the stomach wall or clogged by the presence of a clot or gastric contents. Repositioning of the tube and/or flushing it should resolve this. The lavage is usually reported as amount of fluid necessary until the tube aspirate is clear. Confirmation of the material as blood should be performed with appropriate gastric testing equipment. Presence of dark (“coffee-ground”) material suggests blood that has been exposed to oxygen for some time and indicates that the acute bleeding may have ceased, while bright red blood that does not readily clear indicates active bleeding. The absence of blood suggests bleeding originating distal to the ligament of Treitz, the tube may lodged in a large clot, or the lack of bleeding. However in studies in adults, there is a significant false negative rate of NG lavage in acute upper GI bleeding. Some centers prefer not to utilize nasogastric lavage because of a theoretical risk of dislodging a clot in the stomach or over a varix, although this is unlikely, or if they plan to perform urgent endoscopy to evaluate the bleeding source and the lavage will not provide additional useful information. Nasogastric lavage is often very helpful to the practitioner when making clinical decisions about the necessity of urgent medical interventions, severity and location of bleeding, as well as ongoing monitoring of bleeding, and in assuring good visualization during endoscopy.
Laboratory studies directed to assess blood volume status including a complete blood count (CBC), PT/International Normalized Ratio (INR), and a blood type and screen in case a transfusion is indicated, are usually the first line of tests obtained. The CBC should be interpreted using the mean corpuscular volume (MCV) and mean corpuscular hemoglobin (MCH) to assess the chronicity of the condition, or with appropriate medical history for hemoglobinopathy. Low MCV/MCH levels (microcytic, hypochromic) suggest chronic bleeding or acute on chronic condition, while normal levels indicate an acute blood loss. If the patient has folate or B12 deficiency the values may vary from the expected. These deficiencies often create a macrocytic process with an elevated MCV and MCH. Blood urea nitrogen (BUN) creatinine ratio may also be helpful in determining the source of bleeding. A ratio of >30 has 98% sensitivity and 68% specificity for upper GI bleeding. This occurs due to the rise in serum urea level as a byproduct of amino acid catabolism secondary to digestion of blood in the GI tract. To assess the severity of bleeding, serial CBCs can be obtained. If the history is suggestive of a nonemergent cause of bleeding, such as anal fissure, rectal polyp, or a simple nosebleed, limited laboratory studies are required and a CBC may suffice.
A lower GI bleeding is usually assessed through a combination of the above laboratory studies and the appearance of the rectal effluent or rectal exam. Although hematochezia is usually representative of lower GI bleeding, a brisk upper GI bleeding may cause this finding also.
Vital signs: orthostasis, pulse pressure, urine output
General: appearance (well or ill), fever, mental status
Head, eyes, ears, nose, and throat: trauma, scleral injection, petechiae, lip and buccal mucosa pigmentation, epistaxis, erythema or burns to posterior pharynx, bleeding
Chest/Cardiovascular: tachycardia, murmur, capillary refill
Abdomen: tenderness, splenomegaly, hepatomegaly, caput medusae, distention, ascites
Genitourinary: fistula, swelling
Rectal: gross blood, melena, skin tags, tenderness, fissure, fistula, swelling, warmth
Dermatological: pallor, jaundice, rash, arteriovenous malformations, bruising, petechiae
Prothrombin time/partial thromboplastin time/INR
Complete metabolic profile (electrolytes and liver function tests)
Type and screen
If significant loss: type and cross and fibrinogen level
Stool culture (Shigella, Salmonella, Yersinia, Campylobacter) with Escherichia coli O157:H7 assay, Clostridium difficile toxins A&B, Cryptosporidium and Giardia assay, and ova and parasite smear, if indicated by history
Erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), gamma glutamyl transpeptidase (GGT), if indicated by history
Hemoccult and gastroccult testing
The child with gastrointestinal hemorrhage may present in stable condition or with various degrees of circulatory compromise including shock. While performing initial assessment the patient should be kept NPO (nil per os) in order to decrease the risk of further aggravating the bleeding and in anticipation of diagnostic tests or endoscopy that will require sedation, and intravenous (IV) access should be obtained. If ill appearing, two large bore IV lines should be placed. Depending on the need for initial fluid resuscitation, aliquots of 20 cc/kg of normal saline should be infused quickly (30-60 minutes) as per Paediatric Advanced Life Support (PALS) guidelines for shock. If blood loss is rapid, this volume may need to be replaced with packed red blood cells and appropriate clotting factors/fibrinogen depending on the condition. In case of a planned urgent endoscopic procedure, patients should have ample reserves to tolerate anesthesia. There is no single hemoglobin threshold that determines the need for transfusion in all children. A restricted transfusion threshold of 7 g/dL may be acceptable compared to a more liberal threshold of 9 g/dL in acute upper GI bleeding. The overall goal is for appropriate heart rate and blood pressure, as attempting to intervene with unstable vital signs and a low blood volume may cause more harm than benefit to the patient. Along with volume repletion, all equipment and support staff necessary for successful completion of the procedure should be readily available and when indicated, surgical backup should be considered.
Once the child is stabilized and initial testing results return, the practitioner needs to determine the source of bleeding. If it appears that the source of bleeding originated from outside of GI tract such as pulmonary or nasopharyngeal, the appropriate specialist should be consulted. Advanced imaging modalities (CT/MRI) could be ordered to assist in diagnosing these conditions. There are multiple causes of bleeding and no single algorithm will cover all of them, thus a careful individualized approach utilizing a variety of tests is needed. For example, large volume esophageal or gastric bleeding may be easier to identify and necessitate fewer studies than a slow indolent small bowel hemorrhage. Etiologies relating to specific bleeding lesions of the GI tract will be discussed here in more detail including a discussion of various imaging techniques. Infectious and autoimmune causes are covered elsewhere in this text.
Chest and abdominal x-ray-foreign body, constipation, vomiting
Upper gastrointestinal series-dysphagia, odynophagia, drooling, obstruction, vomiting
Barium enema-suspected stricture, intussusception, Hirschsprung’s disease (late onset)
Ultrasound (Doppler recommended for liver disease)-portal hypertension, intussusception, possible IBD
Meckel’s scan-Meckel’s diverticulum
Tagged RBC scan-obscure gastrointestinal bleeding
MR/CT/direct angiography-obscure gastrointestinal bleeding, suspected arteriovenous malformation
Esophagogastroduodenoscopy-hematemesis, melena, hematochezia
Small bowel enteroscopy, capsule endoscopy, balloon enteroscopy,
Spiral enteroscopy-obscure gastrointestinal blood loss
Pediatric gastroenterologists should be skilled in a variety of endoscopic techniques described below. Some of these may require advanced skills for which additional endoscopic training may be required. It is advisable to have an experienced endoscopist available for assistance. The North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the American Society for Gastrointestinal Endoscopy (ASGE) have issued guidelines and competence requirements for individuals performing these procedures. Additionally, a discussion between the endoscopist and family/legal guardian should occur regarding the risks, benefits, alternatives, methods, and indications/purpose for the procedure. One should also consider limited ethical or legal capacity for a parent to refuse care for a life threatening condition. If a parent refuses intervention during the need for life saving therapy, legal assistance, court intervention, or social work consult should be obtained immediately.
Bleeding from the upper GI tract can often be readily discovered by nasogastric lavage, and further localized by endoscopy. Nasogastric lavage has good positive predictive value for high risk endoscopic lesions in patients with acute upper GI bleeding but false negative lavage results may occur. The purpose of the esophagogastroduodenoscopy (EGD) is to confirm the diagnosis and possibly treat the bleeding lesion. If a patient presents with hematemesis, or has blood or heme positive content found on nasogastric lavage, radiographic testing is rarely indicated unless the EGD is negative. Possible imaging modalities prior to performing EGD include an abdominal x-ray to localize a foreign body, ultrasound with Doppler to evaluate the liver and abdominal vasculature, or CT scan to evaluate organomegaly, masses, or vasculature.
Small bowel bleeding is often suspected based on the absence of upper endoscopic or lavage findings and the appearance of a dark melenic or maroon-colored stool. The color of stool will depend on the location of bleeding and the bowel transit time. In rare cases of very brisk proximal small bowel bleeding with rapid transit time hematochezia might be encountered. In addition to endoscopic techniques, various imaging studies can be utilized to diagnose small bowel bleeding including upper GI small bowel follow through, nuclear scintigraphy, CT enterography and angiography (CTA), ultrasound with Doppler, direct angiography, and MR angiography or venography (MRA/MRV). Nuclear radionucleotide scan may be used to diagnose a rapidly bleeding small bowel lesion. Generally, brisk bleeding is needed for a red blood cell nucleotide scan to be successful with a rate of bleeding of more than 0.05 to 0.5 mL per minute. Some authors suggest 24-hour delayed images be obtained in order to increase its sensitivity. Alternatively, if Meckel’s diverticulum is suspected, a Meckel’s scan using technetium-99 pertechnetate that is taken up by the ectopic gastric tissue if present can be useful. Administration of oral or IV ranitidine for 24 to 48 hours prior to the study, thus up-regulating the H2 receptor in the ectopic gastric tissue, may optimize chance of success. In case of persistent bleeding and negative studies outlined above a CTA or MRA/MRV may offer visualization of a vascular malformation. The yield of such testing is relatively low, but in the correct clinical setting can be helpful. Finally, direct angiography by a trained pediatric interventional radiologist offers a way to visualize and embolize the bleeding vessel.
Newer endoscopic modalities can also be useful to image and intervene in the small bowel. Techniques that have become more recently available include single and double balloon, spiral, and capsule endoscopy (CE). Each of the modalities has its risks and benefits. Capsule endoscopy is more widely implemented and is approved by the FDA for patients 2 years of age and older, although the youngest patients reported to date include a 10-month-old. A majority of patients older than 8 years are able to swallow the capsule, while younger patients and those unable to swallow it need the capsule to be placed endoscopically. The capsule endoscopy may visualize the bleeding lesions, but with its current design offers no therapeutic options. Studies have shown that the diagnostic efficacy of CE for obscure small bowel bleeding ranges between 55% to 81%. During the last 5 years balloon enteroscopy has become more widely practised in pediatrics. The pediatric literature on the use of single and double balloon enteroscopy is growing, while use of spiral enteroscopy in pediatrics remains sparse. In a recent publication of 58 children undergoing 113 double balloon enteroscopy procedures, OGIB was the second most common procedure indication (28%), the diagnostic yield for all cases was 70.7%, and therapeutic endoscopic techniques were applied in 46.5% of cases. This technique requires special training and is time consuming, but offers the opportunity for direct visualization of the entire small bowel in select cases, and therapy. Panendoscopy at the time of surgical exploration is another possible modality, but with the advent of the above technologies has become less commonly utilized.
Hematochezia almost always originates in the colon, except in a very rare case of brisk upper GI bleeding. To evaluate the source of bleeding, and after ruling out infectious causes in the appropriate setting, colonoscopy is the best initial diagnostic choice. If the bleeding lesion is not detected during colonoscopic examination, EGD may be needed. Some centers perform both EGD and colonoscopy taking advantage of the same sedation to rule out an upper tract source of the bleeding in the setting of significant acute hematochezia. If inflammatory bowel disease is suspected, both EGD and colonoscopy are recommended to evaluate the full extent of the disease. When polyps are suspected, colonoscopy is the ideal choice due to its diagnostic ability, but also to provide definitive treatment. Radiological techniques as above may also be used.
Preparation for Procedure
Depending on the procedure of choice there are a variety of preparations that need to be made. The first is to assure that the patient is medically stable, especially for tests requiring sedation, and in the setting of anticipated ongoing and potentially severe bleeding. The second is to make sure that the support staff and all the necessary equipment are available for the procedure. If this is not possible, consideration for patient transfer to an appropriately equipped facility should be made.
The preparation for EGD mostly consists of adequate NPO time. This may differ depending on institutions’ sedation policy, but is approximately 2 to 3 hours for liquids and 6 hours for solids depending on the type of fluids ingested and patient’s age. The NG lavage can be very helpful in assisting with evacuation of blood and other gastric content to assure adequate endoscopic visualization of the mucosa. In the great majority of cases general anesthesia is administered by a dedicated anesthesia team providing the necessary level of sedation and analgesia, thus allowing the endoscopist to solely focus on the procedure at hand, especially for prolonged and therapeutic procedures. Endotracheal intubation is frequently utilized to prevent aspiration.
The preparation for colonoscopic examination requires not only the appropriate NPO time, but also a bowel purgative, which is of the utmost importance in order to assure proper mucosal visualization. If the endoscopist attempts to perform a colonoscopy without proper cleansing this will make a thorough exam very difficult and increase chance of complication. The bowel preparation for colonoscopy may need to be adjusted depending on the type of bleeding encountered and the endoscopist’s preference. For example, for patients with history of hematochezia and diarrhea secondary to presumed colitis, standard bowel preparation should suffice, while a patient with melena or maroon-colored stool may require a more aggressive approach with large volume (40 cc/kg/h) colon lavage, or until rectal effluent is clear. A variety of bowel preparations are available and a detailed review is available elsewhere.
There is an ongoing debate regarding video capsule endoscopy preparation. Some advocate colonoscopy-type preparation with or without simethicone, while others elect a prep similar to that for an EGD. A clinical report on pediatric bowel preparation, including literature review and preparation recommendations from the NASPGHAN Endoscopy & Procedures Committee is pending publication and will be a valuable resource.
Therapy: ( Box 13-7 )
In order to determine the best approach for treatment of GI bleeding one needs to gather information from all available sources including a thorough history and physical exam, as well as appropriate interpretation of results from various diagnostic tests. Once the diagnosis is suspected or made, the choice of therapy with the best possible chance of success and the least possibility of adverse events is determined.
IV fluids (isotonic)
Blood products (packed red blood cells, fresh frozen plasma)
Pressors (dopamine, etc)
Proton pump inhibitors (omeprazole, lansoprazole, pantoprazole)
Prokinetics (metoclopramide, erythromycin)
Somatostatin analog (octreotide)
Injection (sclerosant, epinephrine, normal saline, hypertonic saline)
Coagulation (MPEC, monopolar, heater probe, laser, argon plasma)
Variceal injection and ligation
Hemospray™ powder if available
In patients with upper GI bleeding, intravenous proton pump inhibitor (PPI) therapy is frequently utilized. A Cochrane Review of adult studies evaluating pre-endoscopy PPI use showed decreased rates of finding high risk bleeding stigmata and need for endoscopic therapy in those given PPI. However, no randomized studies exist in the pediatric literature to discuss this practice. Theoretically, the nonacidic gastric pH can help stabilize clot formation and IV use can achieve this faster than oral PPI administration. Use of pre-endoscopy intravenous prokinetic therapy, with either metoclopramide or erythromycin, has been reviewed in adults, demonstrating a decreasing need for repeat endoscopy to determine the site of bleeding. No pediatric studies of this practice exist either, and while not standard of care in pediatrics or adults, may be considered when there is a high likelihood of active gastric bleeding.
For patients with variceal bleeding, pediatric literature review suggests octreotide to be a safe agent, although it can be used for nonvariceal bleeding in addition. Octreotide should be used with caution, especially in patients with known cardiac disease since it affects not only splanchnic circulation, but other vasculature as well. Additionally, there is a risk of bowel ischemia and close monitoring for signs of bowel injury are necessary. The dose of octreotide starts at 1 mcg/kg intravenous bolus followed by 1 to 2 mcg/kg/hr continuous infusion with slow 25% titration to dose and then taper off once bleeding is stabilized.
Once the patient is stabilized, medical therapies are instituted, and necessary diagnostic tests are obtained, EGD or colonoscopy is often necessary. It is advisable to choose age-appropriate, largest possible instrument size which will optimize the use of various instruments that might be necessary for therapy, as well as enable proper flushing of blood and debris when needed. The endoscopists should be familiar with the size of available endoscopes and accessories; a document is now published detailing this information. In some cases the size of the endoscope’s working channel may dictate the therapeutic approach. For example, not all the equipment will fit through the neonatal-size endoscope. Injection needles and argon plasma coagulation probes are the most readily available hemostatic devices for use in a neonatal endoscope with a 2.0 mm working channel. One bipolar thermocoagulation probe fits this size but may not be available at all centers. Endoscopic clips and variceal band ligation are not available in this size and require utilization of a standard size endoscope. Upper endoscopy and colonoscopy are discussed in more detail in Chapters 60 and 61 , respectively.
Endoscopic modalities to achieve hemostasis include heat, electricity, argon plasma, injectable vasopressors, laser, and mechanical tamponade. An electrical current applied with a multipolar probe (MPEC) causes coagulation of the bleeding site which may be used in conjunction with pressure, known as coaptive coagulation. The settings of the electrical generating unit for cautery varies by manufacturer and should be set according to its instructions. Usual settings for the MPEC probe range from 15 to 25 W. With small lesions the probe is placed directly on the blood vessel or lesion. Otherwise, cautery is performed around the bleeding site in pulses of 2 to 3 seconds until hemostasis is achieved. A final cautery at the center of the lesion is then performed. This technique can also be performed to prevent further bleeding in a high risk lesion such as visible vessel, sentinel clot, or a cherry red spot. A large bleeding artery may require three to six applications of cautery using similar circumferential applications followed by center application. Contraindications to performing these techniques include hemodynamically unstable patient, significant arterial bleeding that obscures the visual field, esophageal variceal bleeding, and arteriovenous malformations >1 cm in size.
Another technique that can be used alone or in conjunction with the above is injection therapy. Various agents, when injected at the bleeding site will cause vasoconstriction or sclerosis. An example is 1:10,000 epinephrine solution injected in 0.5 to 1 mL aliquots into the bleeding lesion. Hypertonic saline/epinephrine, thrombin, and ethanol can also be used. Each agent has specific doses, formulations, and methods of administration and the endoscopist should be familiar with those before their use.
Techniques becoming more commonly used in pediatrics include endoscopic clips and argon plasma coagulation (APC). Hemoclips offer the ability to stop bleeding quickly without associated tissue injury, using continuous tamponade on areas where the bleeding site is readily seen. In nonvariceal bleeding the efficacy in hemostasis has been described to range from 84% to 100%. Hemoclips are most useful if the site of bleeding can be approached directly or diagonally. A clip is more difficult to apply tangentially, such as at the posterior wall of the duodenal bulb. The endoscopist should be familiar with the particular hemoclip model prior to use as the ability to rotate or grasp and reopen before final deployment is not universal. APC delivers a noncontact thermal coagulation via argon gas. Multiple advantages of APC compared to MPEC probes include limited depth of tissue injury, ability to cover a greater mucosal surface area with coagulation, lack of coagulum adherence to the probe, and the smaller probe size is compatible with a neonatal endoscope working channel. Finally, an emerging and promising technique not yet studied in children is the hemostatic spray, which is a nonthermal modality that forms a gel when mixed with moisture to create an adhesive barrier over a bleeding lesion. Unfortunately these products are not yet FDA approved in the United States. If bleeding cannot be stopped or is very significant, surgery to repair the lesion should be considered. Images of the effects of these modalities are shown in sequence of application ( Figures 13-1, 13-2, 13-3, 13-4 ).