Gastrointestinal Bleeding

Sabina Mir

Douglas S. Fishman

KEY POINTS

Patients in a critical care unit are at high risk of gastrointestinal bleeding from stress-related mucosal ulcers and may benefit from prophylactic antisecretory therapy.

Initiate broad-spectrum antibiotic coverage and medical reduction of portal hypertension early in the course of acute variceal bleeding.

A clear nasogastric aspirate does not exclude a bleeding source proximal to the ligament of Treitz.

Investigate infectious etiologies in acute bloody diarrhea: Escherichia coli O157:H7, Shigella, Salmonella, Yersinia, Campylobacter, and Clostridium difficile.

INTRODUCTION AND EPIDEMIOLOGY

Gastrointestinal (GI) bleeding is a common condition in the pediatric age group with a higher incidence in children admitted in the intensive care unit (ICU). The reported incidence of GI bleeding in the pediatric ICU (PICU) is 6%-25% (1,2,3). Fortunately, the risk of life-threatening bleeding in children admitted in the PICU with upper GI bleeding is only 0.4% (3). Critically ill patients have an increased risk of GI bleeding secondary to multiple comorbidities (high blood pressure, mechanical ventilation (2), disseminated intravascular coagulopathy (DIC), medications, circulatory shock, trauma, and septicemia).

In 2009, 23,383 children aged 11-15 years were admitted in US hospitals with a diagnosis of GI bleeding. The highest mortality rates associated with GI bleeding were in cases with intestinal perforation (8.7%) and esophageal perforation (8.4%). Children with GI bleed had higher rates of comorbidities (12.3% vs. 2.3%; p<.001) (4).

In adult patients, non variceal upper GI bleeding is associated with costs ranging from $3180 to 8890 per admission and a mortality of 3.5% (5). The severity of upper GI bleeding may range from self-resolving benign conditions to progressive and life-threatening bleeding that prompts immediate intervention. This chapter focuses on the initial approach, diagnostic investigation, and management (medical and therapeutic) of GI bleeding in the PICU.

ANATOMICAL LOCATION

GI bleeding is broadly divided into upper and lower GI tract, depending on the anatomical origin. The ligament of Treitz, located at the junction of the end of the duodenum and beginning of the jejunum, provides the division between the upper and lower GI tracts; upper bleeding originates proximal to the ligament, and lower bleeding originates distally. Upper GI bleeding is further characterized into variceal and nonvariceal bleeding. It is essential to exclude other, non-GI, sites where bleeding may originate, such as from the nose or lungs, as different management is required.

DEFINITIONS AND CLINICAL PRESENTATION

Hematemesis is vomiting fresh blood and indicates a rapidly bleeding lesion.

Coffee ground emesis is vomit with a dark brown color due to the effects of gastric acid on the blood.

Melena is a dark-, black-, or tar-colored stool, secondary to bleeding from above the ileocecal valve or, if colonic transit time is slow, from proximal large bowel. A small volume of blood in the stomach can cause melena for 3-5 days and may not indicate ongoing bleeding.

Hematochezia is the passage of bright red blood per rectum or maroon-colored stools, resulting from a colonic source or massive upper GI bleeding.

Obscure GI bleeding (OGIB) is blood loss that occurs with a negative upper endoscopy and colonoscopy or obvious source from basic imaging.

ETIOLOGY

There are numerous causes of GI bleeding in the pediatric population, varying with the age of the child. Tables 99.1, 99.2, 99.3 summarize the etiology of GI bleeding in children (6).

DIAGNOSTIC EVALUATION

A detailed history, including medications that may potentiate GI bleeding (e.g., anti platelet agents and non steroidal anti-inflammatory drugs [NSAIDS]), and physical examination are important in elucidating the cause and source of bleeding. Physical findings characteristic of underlying disease can help establish the diagnosis (Table 99.4). There are a number of interventional options available for diagnostic evaluation and intervention; they are described in the following section and the next, and summarized in Figure 99.1.

TABLE 99.1 CAUSES OF NON VARICEAL UPPER GASTROINTESTINAL BLEEDING IN CHILDREN

▪ NEWBORN	▪ INFANT	▪ CHILD-ADOLESCENT
Swallowed maternal blood	Esophagitis	Esophagitis
Maternal breast milk irritation	Maternal breast milk irritation	H. pylori (gastritis)
Stress gastritis, ulcers	Duodenitis	Salicylates, NSAIDS
Vitamin K deficiency	Peptic ulcer disease	Mallory-Weiss tear
Vascular malformation	Mallory-Weiss tear	Duodenal Crohn
Milk protein sensitivity	Gastric esophageal duplication	Vascular malformation
Necrotizing enterocolitis	Salicylates, NSAIDS	Dieulafoy lesion
Hemophilia	Foreign body	Coagulopathy (DIC, ITP)
Gastric esophageal duplication	Coagulopathy (DIC, ITP)
Maternal ITP	Vascular malformation Dieulafoy lesion
DIC, disseminated intravascular coagulopathy; ITP, idiopathic thrombocytopenic purpura; NSAIDS, non steroidal anti-inflammatory drugs.

TABLE 99.2 CAUSES OF VARICEAL GASTROINTESTINAL BLEEDING IN CHILDREN

▪ PREHEPATIC	▪ POSTHEPATIC	▪ INTRAHEPATIC
Portal vein thrombosis	Budd-Chiari syndrome	Biliary atresia
Splenic vein thrombosis	Congestive heart failure	Autoimmune hepatitis
Arteriovenous fistula	Inferior vena cava obstruction	Metabolic liver disease (Wilson disease, hemachromatosis)
	Constrictive pericarditis	Toxins Sinusoidal obstruction syndrome (veno-occlusive disease) Infectious hepatitis

The Nasogastric Tube

An assessment of active bleeding (and gastric decompression) can be determined by nasogastric (NG) tube placement, although the role of tube placement is controversial. A meta-analysis of adult patients with GI bleeding revealed that blood or coffee ground material in an NG aspirate has a 44% sensitivity, 95% specificity, and positive likelihood ratio of 9.4 for UGI bleeding (7). In a study of 520 patients that underwent NG aspirate prior to endoscopy for UGI bleed, 15% had a clear aspirate but were found to have an upper GI lesion on endoscopy (8). Hence, a clear NG aspirate does not exclude a

bleeding source proximal to the ligament of Treitz. Conversely, the presence of blood in the NG aspirate generally confirms the diagnosis of UGI bleeding in an otherwise consistent clinical context and provided swallowed blood can be ruled out. On balance, we favor the placement of an NG tube in children in the PICU with suspected UGI bleeding to confirm the diagnosis and improve the conditions for endoscopy. Potential
complications of NG tube insertion should be considered in patients with basil skull fracture, severe facial trauma, clotting disorders, esophageal varices, esophageal tumors, or esophageal surgery.

TABLE 99.3 CAUSES OF LOWER GASTROINTESTINAL BLEEDING IN CHILDREN

▪ NEONATE	▪ INFANT	▪ CHILD-ADOLESCENT
Hemorrhagic disease of newborn	Anal fissure	Anal fissure
Coagulopathy	Infectious colitis	Infectious colitis
Milk protein sensitivity	Intussusception	Polyp
Necrotizing enterocolitis	Meckel diverticulum	Vascular malformation
Volvulus	Volvulus	Inflammatory bowel disease
Vascular malformation	Vascular malformation	Vasculitis (Henoch-Schönlein purpura)
Hirschsprung disease enterocolitis	Intestinal duplication	Intestinal duplication

TABLE 99.4 PHYSICAL EXAMINATION

▪ SYSTEM	▪ SIGN	▪ SUSPECTED DIAGNOSIS
Eyes	Jaundice	Cirrhosis
	Iritis	IBD
Skin	Hemangiomas, telangiectasia	Vascular anomalies
	Petechia, bruises	Cirrhosis, DIC
	Jaundice	Cirrhosis
	Purpura	HSP
	Skin tags, anal fissures, erythema nodosum, pyoderma gangrenosum	IBD
Abdomen	Hepatosplenomegaly, ascites	Cirrhosis
	Distension, rebound tenderness	Bowel perforation
Joints	Arthritis	HSP, IBD
DIC, disseminated intravascular coagulopathy; HSP, Henoch-Schönlein purpura; IBD, inflammatory bowel disease.

FIGURE 99.1. Diagnostic algorithm for severe upper GI hemorrhage. Epi, epinephrine.

Laboratory Tests

Hemodynamic changes (e.g., tachycardia, hypotension) are signs of hypovolemic shock and require immediate resuscitation measures prior to further evaluation. Initial laboratory investigations (complete blood count and coagulation profile) should be instituted during resuscitation. Once the child is hemodynamically stable, additional laboratory testing (Table 99.5) to investigate the cause and to direct therapy may include liver panel, metabolic panel, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and stool cultures (E. coli O157:H7, Shigella, Salmonella, Yersinia, Campylobacter, and Clostridium

difficile [C. diff] PCR). Fecal and gastric contents can be tested for the presence of hemoglobin using a guaiac-based test to detect occult bleeding. False-positive guaiac reactions occur with intake of red meat, raw fruits or vegetables, and other foods with peroxidase activity. Inflammatory markers (ESR and CRP) are useful when inflammatory bowel disease is considered.

TABLE 99.5 INVESTIGATION

▪ INVESTIGATION	▪ INDICATION
Laboratory
Complete blood count	Assess severity of bleeding (anemia and thrombocytopenia)
PT/INR and PTT	Coagulopathy (liver dysfunction, DIC)
BUN/Cr	High ratio (>30) suggestive of UGI bleed (in adults)
Stool culture	Infectious colitis (Shigella, Salmonella, Yersenia, Campylobacter, and enteropathic E. coli) Stool PCR for C. diff
ESR/CRP	Infectious and inflammatory disorders
Guaiac-based tests	Detection of hemoglobin in gastric or fecal content
Imaging
X-ray	Foreign body, bowel perforation or obstruction
Abdominal ultrasound	Portal hypertension, vascular malformation, intussusception
Radionuclide scanning	TPT (Meckel scan), detection of heterotopic gastric mucosa
CT/MRI abdomen	Mass, vascular malformation, intestinal duplication
Angiography	In massive UGI bleed (bleeding must be at least 0.5 mL/min to be detected
Endoscopy
Esophagogastroduodenoscopy	For assessing acute UGI bleeding requiring transfusion or unexplained recurrent bleeding
Colonoscopy	Polyps, colitis, vascular malformations
Enteroscopy	Small bowel polyps, ulcers, colitis, vascular malformations
Wireless capsule study	Small bowel polyps, ulcers, colitis, vascular malformations
DIC, disseminated intravascular coagulopathy; UGI, upper gastrointestinal; TPT, technetium 99m pertechnetate.

Radiologic and Nuclear Medicine Imaging

Plain x-ray film may be useful in the setting of suspected GI perforation or intestinal obstruction presenting with bleeding. Ultrasonography with Doppler flow is useful for the evaluation of portal hypertension, liver disease, large vascular anomalies (arteriovenous malformations, hemangiomas), and intussusception. Air contrast or barium contrast enema can be used to confirm the diagnosis and treat colonic intussusception. Regular and angiographic CT and MRI are noninvasive and useful in detecting mass lesions and vascular malformations. These modalities are noninvasive but may require sedation. Technetium 99m pertechnetate (TPT) rapidly binds to gastric mucosa and is useful in detecting ectopic gastric mucosa, particularly Meckel diverticulum. The sensitivity of TPT scintigraphy has been reported to range from 50% to 91% (9,10). Pretreatment with pentagastrin, histamine H₂ blockers, or glucagon is reported to enhance the sensitivity of the Meckel scan (11

Only gold members can continue reading. Log In or Register to continue