The patient is suffering from a gastrointestinal bleeding (GIB). Two large-bore intravenous catheters (≥ 16 gauge) are necessary to initiate resuscitation with crystalloid and colloid fluids or blood products. When peripheral venous access cannot be obtained, a central venous line such as a Cordis catheter (Cordis Corp, Bridgewater, NJ) should be inserted. An arterial line should be inserted for beat-to-beat monitoring of systemic blood pressure. Four units of packed red blood cells (PRBCs) and 4 units of fresh-frozen plasma should be typed and crossed. Bedside assessment of end-organ perfusion and serial laboratory testing (basic metabolic profile, hemoglobin level, coagulation studies) guide transfusions and prevent delayed resuscitation and hypoperfusion. Vasoactive agents are administered during fluid resuscitation to maintain an adequate perfusion pressure.

This patient has a protected airway, but endotracheal intubation should be considered in patients with massive hemoptysis, shock, or obtundation. A nasogastric tube (NGT) should be placed to assess the site and rate of bleeding. If the abdomen is distended or if the patient complains of abdominal pain, a KUB (kidneys, ureters, and bladder) film should be performed and evaluated.

Two large-bore intravenous catheters, an arterial line, and an NGT are placed. The stomach is lavaged, and 300 mL of bright red blood with coffee grounds appearance are observed. The patient’s blood pressure has decreased to 70/55 mm Hg, and his heart rate is 137 bpm. The gastrointestinal medicine service has been consulted.

The differential diagnosis of GIB in critically ill patients is broad (Table 32-1). Observation of bright red blood with coffee grounds appearance on NGT lavage suggests an upper gastrointestinal (UGI) source (originating proximal to the ligament of Treitz). UGI bleeding accounts for approximately 75% of GI bleeds and has a mortality rate of 20% to 30% in hospitalized patients.¹ Intensive care unit (ICU) patients whose lungs are mechanically ventilated for > 48 hours and have coagulopathy, a history of GI ulceration, or bleeding within the last 12 months or two of the following risk factors: sepsis, ICU admission > 7 days, occult bleeding of ≥ 6 days, or daily use of 250 mg of hydrocortisone or an equivalent are at risk for stress-related mucosal damage.² A history of alcoholism, cirrhosis, or portal hypertension is often elicited in patients with variceal bleeding. Although melena is more common in patients with UGI bleeding, melena is also observed in patients with lower GIB.

Patients whose lungs are mechanically ventilated are at risk for mucosal damage and should be receiving medications to prevent stress gastritis. Prophylactic medications to prevent erosive gastritis include proton pump inhibitors, sucralfate, antacids, and histamine-2 receptor blockers. In patients who have a GIB, a proton pump inhibitor infusion should be initiated. Pepsin is inactivated and cannot effectively lyse clots at pH levels > 6. Prescribing proton pump inhibitors with endoscopic hemostasis decreases the risk for rebleeding, the need for surgery, and mortality.^3,4 Eighty milligrams of omeprazole or pantoprazole are bolused intravenously to achieve rapid gastric acid suppression. An infusion is started at 8 mg/h for 72 hours.⁵

Seventy-five to 80% of UGI bleeding ceases with supportive care, but patients who are at high risk for rebleeding or death require endoscopic evaluation and treatment.^6–9 The Glasgow Blatchford Scale (GBS) score aids in predicting which patients are high risk. The GBS is an assessment tool developed to predict a patient’s need for further medical or endoscopic treatment for UGI bleeding (Table 32-2). The GBS gives scores to multiple clinical and laboratory parameters, and 50% of patients with a score > 6 need endoscopic intervention.^10–12 Even without a recent BUN (blood urea nitrogen) test, the patient has a GBS score of 11 and should get an EGD for further evaluation and treatment. EGD identifies the location and type of bleeding in 90% of cases and decreases the mortality from hemorrhage.¹³

EGD also aids in the prognostication of patients with UGI bleeding. Multiple studies have taken into consideration clinical factors and findings on endoscopy to determine the risk of rebleeding and mortality (Table 32-3).^14–16 Risk of rebleeding and mortality increases with the number of indicators present.

A bleeding peptic ulcer is seen with EGD. The lesion is injected, sclerosed, and banded, yet the ulcer continues to bleed. The patient has received 6 units of PRBCs over 2 hours. A norepinephrine infusion is started for hypotension via a central line.

Experienced endoscopists achieve hemostasis in 90% of patients with UGI bleeding.^17–19 In patients who continue to bleed, arterial embolization by interventional radiologists or surgical intervention may be necessary. Surgery for acute GIB carries a mortality rate of 20% to 35%. Therefore, noninvasive intervention is often preferred prior to surgical intervention in unstable patients.^20,21

Interventional radiology should be consulted for an angiogram and possible embolization in this patient. Transcatheter arteriograms are successful in localizing the bleeding vessel in 75% of cases.²² Treatment with hemostatic therapy carries a long-term success rate of up to 65% with a low rate of complications.²³ There are few studies directly comparing efficacy, morbidity, and mortality of transcatheter interventions vs surgery, but transcatheter treatment for UGI bleeding does not increase mortality rate.^24,25

The interventional radiologist evaluates the patient and embolizes the left gastric artery without complication. The patient continues to require vasopressor therapy and returns to the neurologic ICU (NeuroICU). Twelve units of PRBCs have been administered over the past 24 hours.

The proton pump inhibitor infusion should be continued for 72 hours to prevent upper GI rebleeding. Hematocrit levels should be evaluated every 4 hours for 24 hours. The patient’s coagulation factors should be evaluated. The patient may develop a dilutional coagulopathy secondary to administration of PRBCs without the administration of plasma. A troponin level should be ordered to assess myocardial damage in the setting of his protracted anemia and hemodynamic instability. Urine output and serum creatinine are followed closely. Hypotension places this patient at risk for acute kidney injury (AKI) (Figure 32-1).

A bladder pressure of 15 mm Hg suggests intra-abdominal hypertension (IAH) (Table 32-4). IAH is elevated intra-abdominal pressure (IAP). The normal IAP is 0 to 5 mm Hg, but it varies with body habitus and different disease states (Table 32-5). There are many risk factors for IAH (Table 32-6), including include sepsis, large-volume resuscitation, abdominal masses, pancreatitis, and abdominal surgery. After aggressive volume resuscitation in patients with a capillary leak syndrome, abdominal and bowel wall edema occurs, increasing abdominal pressure. IAH increases patient morbidity and mortality rates.^26–29

There are several ways to measure IAP. Physical examination alone has not been shown to be sensitive in detecting IAH.^30,31 Direct measurement of IAP is possible, but requires placement of either a needle, catheter, or other pressure-monitoring device directly into the abdominal cavity. Direct measurement is complex and invasive. Measuring gastric or rectal pressures can make indirect estimations of IAP, but current standard of care is via bladder pressure measurement. Studies have shown bladder pressure measurements to be a cost-effective, safe, and accurate instrument for identifying IAH and guiding therapy.^32–36 Measurements are made by attaching a pressure-monitoring system to the injection port of a Foley catheter (Table 32-7). The Abdominal Compartment Syndrome World Congress currently recommends an intravesicular volume of 25 mL.³⁷ Previous recommendations suggested the use of 50 to 100 mL of normal saline, but studies have shown that larger injection volumes can lead to falsely elevated IAP.^38–40 Bladder pressures should be measured regularly in patients with or at risk for IAH to aid in diagnosis and management. The optimal frequency of bladder pressure monitoring has not been determined, but every 4 to 6 hours while the patient is critically ill should be sufficient.

The pathophysiology of IAH is complex (Figure 32-2). An increase in IAP limits venous return and decreases left ventricular preload. Abdominal pressure increases systemic vascular resistance, resulting in increased afterload. Decreased preload and increased afterload result in decreased cardiac output. The cardiovascular system initially compensates with an elevation in the heart rate to maintain cardiac output. Abdominal pressure is transmitted to the thoracic cavity, which compresses the lungs and increases airway pressures. These physiologic changes can result in hypotension, which can lead to mesenteric ischemia and AKI and hypoxemia. Unfortunately, aggressive treatment often leads to worsening IAH and continued clinical deterioration (Figure 32-3).

An IAP of 15 mm Hg can affect ICP, CPP, and cerebral blood flow. An increased intrathoracic pressure elevates central venous pressures and decreases venous drainage from cerebral vessels. Cerebral venous congestion leads to elevated ICP, resulting in intracranial hypertension (ICP > 20 mm Hg). The pressure inside the cranium is directly proportional to the volume of its contents: blood, cerebrospinal fluid, and brain. Without adequate venous drainage, the volume inside the cranium increases. To compensate, cerebrospinal fluid shifts into the spinal column, but this is only a temporary solution. Eventually, the ICP will rise and stay elevated unless treated. Given her recent craniotomy and tumor removal, this patient needs to be evaluated for elevated ICP.